Zomato.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 488,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package stopwords to /home/dhawal/nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n",
      "[nltk_data] Downloading package punkt to /home/dhawal/nltk_data...\n",
      "[nltk_data]   Package punkt is already up-to-date!\n",
      "[nltk_data] Downloading package averaged_perceptron_tagger to\n",
      "[nltk_data]     /home/dhawal/nltk_data...\n",
      "[nltk_data]   Package averaged_perceptron_tagger is already up-to-\n",
      "[nltk_data]       date!\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from collections import Counter\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.model_selection import train_test_split\n",
    "from nltk.corpus import stopwords \n",
    "from nltk.tokenize import word_tokenize\n",
    "import nltk\n",
    "nltk.download('stopwords')\n",
    "import seaborn as sns\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import copy\n",
    "from textblob import TextBlob\n",
    "nltk.download('punkt')\n",
    "nltk.download('averaged_perceptron_tagger')\n",
    "from sklearn.utils import resample\n",
    "from wordcloud import WordCloud\n",
    "import re\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "re.compile('<title>(.*)</title>')\n",
    "from sklearn.linear_model import LogisticRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 489,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"merge_zomato.csv\")\n",
    "# col=df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 490,
   "metadata": {},
   "outputs": [],
   "source": [
    "df\n",
    "df=df[['text','stars','oneword']]\n",
    "# df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 491,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df=df.drop(['Unnamed: 10','Unnamed: 11','review_id','date'],axis=1)\n",
    "\n",
    "# df=df.drop(['user_id','business_id','type'],axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## UPSAMPLE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 492,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<nltk.stem.snowball.SnowballStemmer at 0x7f14d4384e48>"
      ]
     },
     "execution_count": 492,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from nltk.stem.snowball import SnowballStemmer\n",
    "stemmer = SnowballStemmer(\"english\")\n",
    "stemmer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Coversion in lower case ###"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 493,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['text'] = df['text'].str.lower()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Removal of stop words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 494,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk.corpus import stopwords\n",
    "stop = stopwords.words('english')\n",
    "\n",
    "stop=stop[0:144]\n",
    "stop.append('back')\n",
    "stop.append('n\\'t')\n",
    "stop.append('dice')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 495,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['text']\n",
    "df['text'] = df['text'].str.replace('[^\\w\\s#@/:%.,_-]', '', flags=re.UNICODE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 496,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 497,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>stars</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5605</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5606</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5607</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5608</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5609</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5610</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5611</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5612</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5613</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5614</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5615</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5616</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5617</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5618</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5619</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5620</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5621</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5622</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5623</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5624</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5625</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5626</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5627</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5628</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5629</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5630</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5631</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5632</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5633</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5634</th>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5635 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      stars\n",
       "0       5.0\n",
       "1       1.0\n",
       "2       5.0\n",
       "3       5.0\n",
       "4       5.0\n",
       "5       1.0\n",
       "6       5.0\n",
       "7       5.0\n",
       "8       3.0\n",
       "9       3.0\n",
       "10      5.0\n",
       "11      5.0\n",
       "12      5.0\n",
       "13      5.0\n",
       "14      3.0\n",
       "15      5.0\n",
       "16      5.0\n",
       "17      5.0\n",
       "18      3.0\n",
       "19      5.0\n",
       "20      5.0\n",
       "21      5.0\n",
       "22      5.0\n",
       "23      3.0\n",
       "24      5.0\n",
       "25      5.0\n",
       "26      5.0\n",
       "27      3.0\n",
       "28      5.0\n",
       "29      5.0\n",
       "...     ...\n",
       "5605    1.0\n",
       "5606    4.0\n",
       "5607    5.0\n",
       "5608    5.0\n",
       "5609    5.0\n",
       "5610    5.0\n",
       "5611    5.0\n",
       "5612    4.0\n",
       "5613    5.0\n",
       "5614    3.0\n",
       "5615    4.0\n",
       "5616    4.0\n",
       "5617    1.0\n",
       "5618    4.0\n",
       "5619    1.0\n",
       "5620    4.0\n",
       "5621    4.0\n",
       "5622    5.0\n",
       "5623    1.0\n",
       "5624    3.0\n",
       "5625    5.0\n",
       "5626    5.0\n",
       "5627    1.0\n",
       "5628    4.0\n",
       "5629    1.0\n",
       "5630    1.0\n",
       "5631    5.0\n",
       "5632    4.0\n",
       "5633    4.0\n",
       "5634    2.0\n",
       "\n",
       "[5635 rows x 1 columns]"
      ]
     },
     "execution_count": 497,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " df.select_dtypes(include=['float64'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 498,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "text        object\n",
       "stars      float64\n",
       "oneword     object\n",
       "dtype: object"
      ]
     },
     "execution_count": 498,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Exclude stopwords with Python's list comprehension and pandas.DataFrame.apply."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 499,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df['text'].dropna\n",
    "# df.dropna(how='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 500,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df['text']=df['text'].notnull()\n",
    "# df.isnull().values.any()\n",
    "# nan_rows = df[df['text'].isnull()]\n",
    "# nan_rows\n",
    "# df=df[bdf['text'].notnull()]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 501,
   "metadata": {},
   "outputs": [],
   "source": [
    "df=df.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 502,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>stars</th>\n",
       "      <th>oneword</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>about 5 minutes walk from kankurgachi bus stop...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>cheap quality food. ordered food from zomato. ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>as the name suggests, this is a compact and wh...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>kankurgachi dhaba is located at a 5-8min walki...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>kakurgachi dhaaba is a hidden gem located at k...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>too much poor service. we have asked for bar m...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>had a great time with family on the saturday l...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>it  was sumana thanks. for. gr8 service   i wa...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>...finally m reviewing the jw marriott ...it i...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>dinner after a good discussion is quite a task...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>its an amazing place for two of us....great fo...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>we went to jw kitchen for a dinner buffet. we ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>im sure that everybody must hear about the mos...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>the perfect place in town for game of thrones ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>average experience. food was alright. the ambi...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>went there yesterday for lunch...this restaura...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>location: this place is located in salt lake n...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>missing game of thrones does the wait till the...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>the food was kinda okay...they provide suffici...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>good news for all the got zombies out there  y...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>few days back i was there w my friends. so, it...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>im sure that everybody must hear about the mos...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>the perfect place in town for game of thrones ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>average experience. food was alright. the ambi...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>went there yesterday for lunch...this restaura...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>location: this place is located in salt lake n...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>missing game of thrones does the wait till the...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>the food was kinda okay...they provide suffici...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>good news for all the got zombies out there  y...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>few days back i was there w my friends. so, it...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5604</th>\n",
       "      <td>had a red valvet cake. it was really wonderful...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5605</th>\n",
       "      <td>pathetic/poor quality of lachcha parantha it w...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5606</th>\n",
       "      <td>ive ordered chicken handi and the food was goo...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5607</th>\n",
       "      <td>great experience, nice food and service and se...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5608</th>\n",
       "      <td>very decent and friendly environment. food is ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5609</th>\n",
       "      <td>good food.. zomato is asusual... timely delive...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5610</th>\n",
       "      <td>best place for hangout with friends. reasonabl...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5611</th>\n",
       "      <td>i ordered yesterday the burger was amaaaazingg...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5613</th>\n",
       "      <td>this cafe serves one of the best burgers in ra...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5614</th>\n",
       "      <td>basically a d brand for kfc though you can get...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5615</th>\n",
       "      <td>the lounge has a pleasant ambience. plonk your...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5616</th>\n",
       "      <td>ambience 5/5 food 4 service was slow.it took t...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5617</th>\n",
       "      <td>the quantity of food was very less n the taste...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5618</th>\n",
       "      <td>the food was very tasty although it was delive...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5619</th>\n",
       "      <td>the food quantity was very less and they did n...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5620</th>\n",
       "      <td>this is a decent veg restaurant serving both n...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5621</th>\n",
       "      <td>they have the best babycorn chili in town over...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5622</th>\n",
       "      <td>its a great place to hangout with friends and ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5623</th>\n",
       "      <td>once upon a time when it was very popular for ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5624</th>\n",
       "      <td>i m a regular visiting customer of chennai kit...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5625</th>\n",
       "      <td>the grilled chicken cheese sandwich with fries...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5626</th>\n",
       "      <td>always delicious.. foods will come in time ......</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5627</th>\n",
       "      <td>one the worst pizza ever eaten ... the conditi...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5628</th>\n",
       "      <td>the pizza was gr8 apart from its crust everyth...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5629</th>\n",
       "      <td>very worst pizza. ill never recommend to anyon...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5630</th>\n",
       "      <td>the worst paneer lababdar i have eaten in my e...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5631</th>\n",
       "      <td>superb food  though costly but superb everythi...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5632</th>\n",
       "      <td>a really nice place to have grills and buffet....</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5633</th>\n",
       "      <td>we went there on january last week.. not just ...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5634</th>\n",
       "      <td>worst paneer chilly gravy ever. it felt as if ...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Blah!</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5601 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   text  stars  oneword\n",
       "0     about 5 minutes walk from kankurgachi bus stop...    5.0  Insane!\n",
       "1     cheap quality food. ordered food from zomato. ...    1.0   Avoid!\n",
       "2     as the name suggests, this is a compact and wh...    5.0  Insane!\n",
       "3     kankurgachi dhaba is located at a 5-8min walki...    5.0  Insane!\n",
       "4     kakurgachi dhaaba is a hidden gem located at k...    5.0  Insane!\n",
       "5     too much poor service. we have asked for bar m...    1.0   Avoid!\n",
       "6     had a great time with family on the saturday l...    5.0  Insane!\n",
       "7     it  was sumana thanks. for. gr8 service   i wa...    5.0  Insane!\n",
       "8     ...finally m reviewing the jw marriott ...it i...    3.0  Average\n",
       "9     dinner after a good discussion is quite a task...    3.0  Average\n",
       "10    its an amazing place for two of us....great fo...    5.0  Insane!\n",
       "11    we went to jw kitchen for a dinner buffet. we ...    5.0  Insane!\n",
       "12    im sure that everybody must hear about the mos...    5.0  Insane!\n",
       "13    the perfect place in town for game of thrones ...    5.0  Insane!\n",
       "14    average experience. food was alright. the ambi...    3.0  Average\n",
       "15    went there yesterday for lunch...this restaura...    5.0  Insane!\n",
       "16    location: this place is located in salt lake n...    5.0  Insane!\n",
       "17    missing game of thrones does the wait till the...    5.0  Insane!\n",
       "18    the food was kinda okay...they provide suffici...    3.0  Average\n",
       "19    good news for all the got zombies out there  y...    5.0  Insane!\n",
       "20    few days back i was there w my friends. so, it...    5.0  Insane!\n",
       "21    im sure that everybody must hear about the mos...    5.0  Insane!\n",
       "22    the perfect place in town for game of thrones ...    5.0  Insane!\n",
       "23    average experience. food was alright. the ambi...    3.0  Average\n",
       "24    went there yesterday for lunch...this restaura...    5.0  Insane!\n",
       "25    location: this place is located in salt lake n...    5.0  Insane!\n",
       "26    missing game of thrones does the wait till the...    5.0  Insane!\n",
       "27    the food was kinda okay...they provide suffici...    3.0  Average\n",
       "28    good news for all the got zombies out there  y...    5.0  Insane!\n",
       "29    few days back i was there w my friends. so, it...    5.0  Insane!\n",
       "...                                                 ...    ...      ...\n",
       "5604  had a red valvet cake. it was really wonderful...    4.0   Great!\n",
       "5605  pathetic/poor quality of lachcha parantha it w...    1.0   Avoid!\n",
       "5606  ive ordered chicken handi and the food was goo...    4.0   Great!\n",
       "5607  great experience, nice food and service and se...    5.0  Insane!\n",
       "5608  very decent and friendly environment. food is ...    5.0  Insane!\n",
       "5609  good food.. zomato is asusual... timely delive...    5.0  Insane!\n",
       "5610  best place for hangout with friends. reasonabl...    5.0  Insane!\n",
       "5611  i ordered yesterday the burger was amaaaazingg...    5.0  Insane!\n",
       "5613  this cafe serves one of the best burgers in ra...    5.0  Insane!\n",
       "5614  basically a d brand for kfc though you can get...    3.0  Average\n",
       "5615  the lounge has a pleasant ambience. plonk your...    4.0   Great!\n",
       "5616  ambience 5/5 food 4 service was slow.it took t...    4.0   Great!\n",
       "5617  the quantity of food was very less n the taste...    1.0   Avoid!\n",
       "5618  the food was very tasty although it was delive...    4.0   Great!\n",
       "5619  the food quantity was very less and they did n...    1.0   Avoid!\n",
       "5620  this is a decent veg restaurant serving both n...    4.0   Great!\n",
       "5621  they have the best babycorn chili in town over...    4.0   Great!\n",
       "5622  its a great place to hangout with friends and ...    5.0  Insane!\n",
       "5623  once upon a time when it was very popular for ...    1.0   Avoid!\n",
       "5624  i m a regular visiting customer of chennai kit...    3.0  Average\n",
       "5625  the grilled chicken cheese sandwich with fries...    5.0  Insane!\n",
       "5626  always delicious.. foods will come in time ......    5.0  Insane!\n",
       "5627  one the worst pizza ever eaten ... the conditi...    1.0   Avoid!\n",
       "5628  the pizza was gr8 apart from its crust everyth...    4.0   Great!\n",
       "5629  very worst pizza. ill never recommend to anyon...    1.0   Avoid!\n",
       "5630  the worst paneer lababdar i have eaten in my e...    1.0   Avoid!\n",
       "5631  superb food  though costly but superb everythi...    5.0  Insane!\n",
       "5632  a really nice place to have grills and buffet....    4.0   Great!\n",
       "5633  we went there on january last week.. not just ...    4.0   Great!\n",
       "5634  worst paneer chilly gravy ever. it felt as if ...    2.0    Blah!\n",
       "\n",
       "[5601 rows x 3 columns]"
      ]
     },
     "execution_count": 502,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 503,
   "metadata": {},
   "outputs": [],
   "source": [
    "nan_rows = df[df['stars'].isnull()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 504,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df=df.drop(df.index[nan_rows])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 505,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>stars</th>\n",
       "      <th>oneword</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [text, stars, oneword]\n",
       "Index: []"
      ]
     },
     "execution_count": 505,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nan_rows"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 506,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['text'] = df['text'].apply(lambda x:' '.join([word for word in x.split() if word not in (stop)]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### INCLUSION OF LENGTH OF REVIEW"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 507,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['length'] = df['text'].apply(len)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Review vs Stars"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 508,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7f14d46817b8>"
      ]
     },
     "execution_count": 508,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1512x216 with 7 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "graph = sns.FacetGrid(data=df,col='stars')\n",
    "graph.map(plt.hist,'length',bins=50,color='blue')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Removal of meaningless thing from text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 509,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fun(x):\n",
    "#     print(x)\n",
    "    blob=TextBlob(x)\n",
    "#     print(\"blob \",blob)\n",
    "    verb=[]\n",
    "    s=\" \"\n",
    "    for word, tag in blob.tags:\n",
    "#         print(\"ffff\")\n",
    "        if tag == 'RB' or tag=='RBR' or tag=='RBS' or tag=='JJ' or tag=='JJR':\n",
    "            verb.append(word)\n",
    "    s=s.join(verb)\n",
    "    return s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 510,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_word_cloud(data):\n",
    "    '''\n",
    "    Function to plot word cloud from a pandas series\n",
    "    Input: data: pandas.core.series.Series object\n",
    "    Output: Wordcloud of most recurrent words in data\n",
    "        for plots to be in ipython notebook include following line before calling this function\n",
    "        %matplotlib inline \n",
    "    '''\n",
    "    # take relative word frequencies into account, lower max_font_size\n",
    "    wordcloud = WordCloud(max_font_size=40, relative_scaling=.5).generate(str(data))\n",
    "    plt.figure()\n",
    "    plt.imshow(wordcloud)\n",
    "    plt.axis(\"off\")\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 511,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>stars</th>\n",
       "      <th>oneword</th>\n",
       "      <th>length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5 minutes walk kankurgachi bus stop, primarily...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>cheap quality food. ordered food zomato. food ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>name suggests, compact wholesome eatery locate...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>kankurgachi dhaba located 5-8min walking dista...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>kakurgachi dhaaba hidden gem located kankurgac...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>much poor service. asked bar menu 3 times got ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>great time family saturday lunch. huge spread ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>sumana thanks. for. gr8 service highly oblige....</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>...finally reviewing jw marriott ...it great h...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>dinner good discussion quite task. jw kitchen ...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>amazing place two us....great food...great peo...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>went jw kitchen dinner buffet. greeted warmly ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>im sure everybody must hear famous world named...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>perfect place town game thrones lovers. cafe p...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>average experience. food alright. ambience isn...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>went yesterday lunch...this restaurant enrolle...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>location: place located salt lake near island ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>missing game thrones wait till final season se...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>food kinda okay...they provide sufficient amou...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>good news got zombies experience dine like kni...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>days w friends. so, literally great themed ins...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>im sure everybody must hear famous world named...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>perfect place town game thrones lovers. cafe p...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>average experience. food alright. ambience isn...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>went yesterday lunch...this restaurant enrolle...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>location: place located salt lake near island ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>missing game thrones wait till final season se...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>food kinda okay...they provide sufficient amou...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>good news got zombies experience dine like kni...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>days w friends. so, literally great themed ins...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5604</th>\n",
       "      <td>red valvet cake. really wonderful. awesome. lo...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5605</th>\n",
       "      <td>pathetic/poor quality lachcha parantha sour ev...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5606</th>\n",
       "      <td>ive ordered chicken handi food good also like ...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5607</th>\n",
       "      <td>great experience, nice food service service .....</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5608</th>\n",
       "      <td>decent friendly environment. food delicious se...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5609</th>\n",
       "      <td>good food.. zomato asusual... timely delivery ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5610</th>\n",
       "      <td>best place hangout friends. reasonable price. ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5611</th>\n",
       "      <td>ordered yesterday burger amaaaazinggggg twiste...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5613</th>\n",
       "      <td>cafe serves one best burgers ranchi.absolutely...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5614</th>\n",
       "      <td>basically brand kfc though get similar product...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5615</th>\n",
       "      <td>lounge pleasant ambience. plonk large sofas re...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5616</th>\n",
       "      <td>ambience 5/5 food 4 service slow.it took much ...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5617</th>\n",
       "      <td>quantity food less n taste poor. small 4 piece...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5618</th>\n",
       "      <td>food tasty although delivered quiet late prepa...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5619</th>\n",
       "      <td>food quantity less send salad mentioned order....</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5620</th>\n",
       "      <td>decent veg restaurant serving north south indi...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5621</th>\n",
       "      <td>best babycorn chili town overall place quite a...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5622</th>\n",
       "      <td>great place hangout friends family. food quali...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5623</th>\n",
       "      <td>upon time popular testy food feel sad test qua...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5624</th>\n",
       "      <td>regular visiting customer chennai kitchen. 1st...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5625</th>\n",
       "      <td>grilled chicken cheese sandwich fries amazing....</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5626</th>\n",
       "      <td>always delicious.. foods come time .... packed...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5627</th>\n",
       "      <td>one worst pizza ever eaten ... condition pizza...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5628</th>\n",
       "      <td>pizza gr8 apart crust everything perfect..just...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5629</th>\n",
       "      <td>worst pizza. ill never recommend anyone. first...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5630</th>\n",
       "      <td>worst paneer lababdar eaten entire life. food ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5631</th>\n",
       "      <td>superb food though costly superb everything st...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5632</th>\n",
       "      <td>really nice place grills buffet..ambience inte...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5633</th>\n",
       "      <td>went january last week.. like barbeque nation....</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5634</th>\n",
       "      <td>worst paneer chilly gravy ever. felt delivered...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Blah!</td>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5601 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   text  stars  oneword  \\\n",
       "0     5 minutes walk kankurgachi bus stop, primarily...    5.0  Insane!   \n",
       "1     cheap quality food. ordered food zomato. food ...    1.0   Avoid!   \n",
       "2     name suggests, compact wholesome eatery locate...    5.0  Insane!   \n",
       "3     kankurgachi dhaba located 5-8min walking dista...    5.0  Insane!   \n",
       "4     kakurgachi dhaaba hidden gem located kankurgac...    5.0  Insane!   \n",
       "5     much poor service. asked bar menu 3 times got ...    1.0   Avoid!   \n",
       "6     great time family saturday lunch. huge spread ...    5.0  Insane!   \n",
       "7     sumana thanks. for. gr8 service highly oblige....    5.0  Insane!   \n",
       "8     ...finally reviewing jw marriott ...it great h...    3.0  Average   \n",
       "9     dinner good discussion quite task. jw kitchen ...    3.0  Average   \n",
       "10    amazing place two us....great food...great peo...    5.0  Insane!   \n",
       "11    went jw kitchen dinner buffet. greeted warmly ...    5.0  Insane!   \n",
       "12    im sure everybody must hear famous world named...    5.0  Insane!   \n",
       "13    perfect place town game thrones lovers. cafe p...    5.0  Insane!   \n",
       "14    average experience. food alright. ambience isn...    3.0  Average   \n",
       "15    went yesterday lunch...this restaurant enrolle...    5.0  Insane!   \n",
       "16    location: place located salt lake near island ...    5.0  Insane!   \n",
       "17    missing game thrones wait till final season se...    5.0  Insane!   \n",
       "18    food kinda okay...they provide sufficient amou...    3.0  Average   \n",
       "19    good news got zombies experience dine like kni...    5.0  Insane!   \n",
       "20    days w friends. so, literally great themed ins...    5.0  Insane!   \n",
       "21    im sure everybody must hear famous world named...    5.0  Insane!   \n",
       "22    perfect place town game thrones lovers. cafe p...    5.0  Insane!   \n",
       "23    average experience. food alright. ambience isn...    3.0  Average   \n",
       "24    went yesterday lunch...this restaurant enrolle...    5.0  Insane!   \n",
       "25    location: place located salt lake near island ...    5.0  Insane!   \n",
       "26    missing game thrones wait till final season se...    5.0  Insane!   \n",
       "27    food kinda okay...they provide sufficient amou...    3.0  Average   \n",
       "28    good news got zombies experience dine like kni...    5.0  Insane!   \n",
       "29    days w friends. so, literally great themed ins...    5.0  Insane!   \n",
       "...                                                 ...    ...      ...   \n",
       "5604  red valvet cake. really wonderful. awesome. lo...    4.0   Great!   \n",
       "5605  pathetic/poor quality lachcha parantha sour ev...    1.0   Avoid!   \n",
       "5606  ive ordered chicken handi food good also like ...    4.0   Great!   \n",
       "5607  great experience, nice food service service .....    5.0  Insane!   \n",
       "5608  decent friendly environment. food delicious se...    5.0  Insane!   \n",
       "5609  good food.. zomato asusual... timely delivery ...    5.0  Insane!   \n",
       "5610  best place hangout friends. reasonable price. ...    5.0  Insane!   \n",
       "5611  ordered yesterday burger amaaaazinggggg twiste...    5.0  Insane!   \n",
       "5613  cafe serves one best burgers ranchi.absolutely...    5.0  Insane!   \n",
       "5614  basically brand kfc though get similar product...    3.0  Average   \n",
       "5615  lounge pleasant ambience. plonk large sofas re...    4.0   Great!   \n",
       "5616  ambience 5/5 food 4 service slow.it took much ...    4.0   Great!   \n",
       "5617  quantity food less n taste poor. small 4 piece...    1.0   Avoid!   \n",
       "5618  food tasty although delivered quiet late prepa...    4.0   Great!   \n",
       "5619  food quantity less send salad mentioned order....    1.0   Avoid!   \n",
       "5620  decent veg restaurant serving north south indi...    4.0   Great!   \n",
       "5621  best babycorn chili town overall place quite a...    4.0   Great!   \n",
       "5622  great place hangout friends family. food quali...    5.0  Insane!   \n",
       "5623  upon time popular testy food feel sad test qua...    1.0   Avoid!   \n",
       "5624  regular visiting customer chennai kitchen. 1st...    3.0  Average   \n",
       "5625  grilled chicken cheese sandwich fries amazing....    5.0  Insane!   \n",
       "5626  always delicious.. foods come time .... packed...    5.0  Insane!   \n",
       "5627  one worst pizza ever eaten ... condition pizza...    1.0   Avoid!   \n",
       "5628  pizza gr8 apart crust everything perfect..just...    4.0   Great!   \n",
       "5629  worst pizza. ill never recommend anyone. first...    1.0   Avoid!   \n",
       "5630  worst paneer lababdar eaten entire life. food ...    1.0   Avoid!   \n",
       "5631  superb food though costly superb everything st...    5.0  Insane!   \n",
       "5632  really nice place grills buffet..ambience inte...    4.0   Great!   \n",
       "5633  went january last week.. like barbeque nation....    4.0   Great!   \n",
       "5634  worst paneer chilly gravy ever. felt delivered...    2.0    Blah!   \n",
       "\n",
       "      length  \n",
       "0        114  \n",
       "1        103  \n",
       "2        108  \n",
       "3        112  \n",
       "4        105  \n",
       "5         80  \n",
       "6         99  \n",
       "7        101  \n",
       "8        118  \n",
       "9        100  \n",
       "10       106  \n",
       "11        95  \n",
       "12        96  \n",
       "13       100  \n",
       "14        98  \n",
       "15       127  \n",
       "16       107  \n",
       "17       105  \n",
       "18       111  \n",
       "19        85  \n",
       "20        90  \n",
       "21        96  \n",
       "22       100  \n",
       "23        98  \n",
       "24       127  \n",
       "25       107  \n",
       "26       105  \n",
       "27       111  \n",
       "28        85  \n",
       "29        90  \n",
       "...      ...  \n",
       "5604     101  \n",
       "5605     109  \n",
       "5606     107  \n",
       "5607     113  \n",
       "5608     102  \n",
       "5609     115  \n",
       "5610     115  \n",
       "5611     106  \n",
       "5613      85  \n",
       "5614     110  \n",
       "5615      93  \n",
       "5616     109  \n",
       "5617      87  \n",
       "5618      90  \n",
       "5619      84  \n",
       "5620     104  \n",
       "5621      95  \n",
       "5622      99  \n",
       "5623      75  \n",
       "5624     109  \n",
       "5625     101  \n",
       "5626     135  \n",
       "5627      96  \n",
       "5628     106  \n",
       "5629      96  \n",
       "5630      97  \n",
       "5631     109  \n",
       "5632     104  \n",
       "5633     112  \n",
       "5634      86  \n",
       "\n",
       "[5601 rows x 4 columns]"
      ]
     },
     "execution_count": 511,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 512,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>text</th>\n",
       "      <th>stars</th>\n",
       "      <th>oneword</th>\n",
       "      <th>length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>5 minutes walk kankurgachi bus stop, primarily...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>cheap quality food. ordered food zomato. food ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>name suggests, compact wholesome eatery locate...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>kankurgachi dhaba located 5-8min walking dista...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>kakurgachi dhaaba hidden gem located kankurgac...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5</td>\n",
       "      <td>much poor service. asked bar menu 3 times got ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6</td>\n",
       "      <td>great time family saturday lunch. huge spread ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7</td>\n",
       "      <td>sumana thanks. for. gr8 service highly oblige....</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>8</td>\n",
       "      <td>...finally reviewing jw marriott ...it great h...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>9</td>\n",
       "      <td>dinner good discussion quite task. jw kitchen ...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10</td>\n",
       "      <td>amazing place two us....great food...great peo...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>11</td>\n",
       "      <td>went jw kitchen dinner buffet. greeted warmly ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>12</td>\n",
       "      <td>im sure everybody must hear famous world named...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>13</td>\n",
       "      <td>perfect place town game thrones lovers. cafe p...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>14</td>\n",
       "      <td>average experience. food alright. ambience isn...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>15</td>\n",
       "      <td>went yesterday lunch...this restaurant enrolle...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>16</td>\n",
       "      <td>location: place located salt lake near island ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>17</td>\n",
       "      <td>missing game thrones wait till final season se...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>18</td>\n",
       "      <td>food kinda okay...they provide sufficient amou...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>19</td>\n",
       "      <td>good news got zombies experience dine like kni...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>20</td>\n",
       "      <td>days w friends. so, literally great themed ins...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>21</td>\n",
       "      <td>im sure everybody must hear famous world named...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>22</td>\n",
       "      <td>perfect place town game thrones lovers. cafe p...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>23</td>\n",
       "      <td>average experience. food alright. ambience isn...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>24</td>\n",
       "      <td>went yesterday lunch...this restaurant enrolle...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>25</td>\n",
       "      <td>location: place located salt lake near island ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>26</td>\n",
       "      <td>missing game thrones wait till final season se...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>27</td>\n",
       "      <td>food kinda okay...they provide sufficient amou...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>28</td>\n",
       "      <td>good news got zombies experience dine like kni...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>29</td>\n",
       "      <td>days w friends. so, literally great themed ins...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5571</th>\n",
       "      <td>5604</td>\n",
       "      <td>red valvet cake. really wonderful. awesome. lo...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5572</th>\n",
       "      <td>5605</td>\n",
       "      <td>pathetic/poor quality lachcha parantha sour ev...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5573</th>\n",
       "      <td>5606</td>\n",
       "      <td>ive ordered chicken handi food good also like ...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5574</th>\n",
       "      <td>5607</td>\n",
       "      <td>great experience, nice food service service .....</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5575</th>\n",
       "      <td>5608</td>\n",
       "      <td>decent friendly environment. food delicious se...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5576</th>\n",
       "      <td>5609</td>\n",
       "      <td>good food.. zomato asusual... timely delivery ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5577</th>\n",
       "      <td>5610</td>\n",
       "      <td>best place hangout friends. reasonable price. ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5578</th>\n",
       "      <td>5611</td>\n",
       "      <td>ordered yesterday burger amaaaazinggggg twiste...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5579</th>\n",
       "      <td>5613</td>\n",
       "      <td>cafe serves one best burgers ranchi.absolutely...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5580</th>\n",
       "      <td>5614</td>\n",
       "      <td>basically brand kfc though get similar product...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5581</th>\n",
       "      <td>5615</td>\n",
       "      <td>lounge pleasant ambience. plonk large sofas re...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5582</th>\n",
       "      <td>5616</td>\n",
       "      <td>ambience 5/5 food 4 service slow.it took much ...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5583</th>\n",
       "      <td>5617</td>\n",
       "      <td>quantity food less n taste poor. small 4 piece...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5584</th>\n",
       "      <td>5618</td>\n",
       "      <td>food tasty although delivered quiet late prepa...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5585</th>\n",
       "      <td>5619</td>\n",
       "      <td>food quantity less send salad mentioned order....</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5586</th>\n",
       "      <td>5620</td>\n",
       "      <td>decent veg restaurant serving north south indi...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5587</th>\n",
       "      <td>5621</td>\n",
       "      <td>best babycorn chili town overall place quite a...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5588</th>\n",
       "      <td>5622</td>\n",
       "      <td>great place hangout friends family. food quali...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5589</th>\n",
       "      <td>5623</td>\n",
       "      <td>upon time popular testy food feel sad test qua...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5590</th>\n",
       "      <td>5624</td>\n",
       "      <td>regular visiting customer chennai kitchen. 1st...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5591</th>\n",
       "      <td>5625</td>\n",
       "      <td>grilled chicken cheese sandwich fries amazing....</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5592</th>\n",
       "      <td>5626</td>\n",
       "      <td>always delicious.. foods come time .... packed...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5593</th>\n",
       "      <td>5627</td>\n",
       "      <td>one worst pizza ever eaten ... condition pizza...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5594</th>\n",
       "      <td>5628</td>\n",
       "      <td>pizza gr8 apart crust everything perfect..just...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5595</th>\n",
       "      <td>5629</td>\n",
       "      <td>worst pizza. ill never recommend anyone. first...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5596</th>\n",
       "      <td>5630</td>\n",
       "      <td>worst paneer lababdar eaten entire life. food ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5597</th>\n",
       "      <td>5631</td>\n",
       "      <td>superb food though costly superb everything st...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5598</th>\n",
       "      <td>5632</td>\n",
       "      <td>really nice place grills buffet..ambience inte...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5599</th>\n",
       "      <td>5633</td>\n",
       "      <td>went january last week.. like barbeque nation....</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5600</th>\n",
       "      <td>5634</td>\n",
       "      <td>worst paneer chilly gravy ever. felt delivered...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Blah!</td>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5601 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      index                                               text  stars  \\\n",
       "0         0  5 minutes walk kankurgachi bus stop, primarily...    5.0   \n",
       "1         1  cheap quality food. ordered food zomato. food ...    1.0   \n",
       "2         2  name suggests, compact wholesome eatery locate...    5.0   \n",
       "3         3  kankurgachi dhaba located 5-8min walking dista...    5.0   \n",
       "4         4  kakurgachi dhaaba hidden gem located kankurgac...    5.0   \n",
       "5         5  much poor service. asked bar menu 3 times got ...    1.0   \n",
       "6         6  great time family saturday lunch. huge spread ...    5.0   \n",
       "7         7  sumana thanks. for. gr8 service highly oblige....    5.0   \n",
       "8         8  ...finally reviewing jw marriott ...it great h...    3.0   \n",
       "9         9  dinner good discussion quite task. jw kitchen ...    3.0   \n",
       "10       10  amazing place two us....great food...great peo...    5.0   \n",
       "11       11  went jw kitchen dinner buffet. greeted warmly ...    5.0   \n",
       "12       12  im sure everybody must hear famous world named...    5.0   \n",
       "13       13  perfect place town game thrones lovers. cafe p...    5.0   \n",
       "14       14  average experience. food alright. ambience isn...    3.0   \n",
       "15       15  went yesterday lunch...this restaurant enrolle...    5.0   \n",
       "16       16  location: place located salt lake near island ...    5.0   \n",
       "17       17  missing game thrones wait till final season se...    5.0   \n",
       "18       18  food kinda okay...they provide sufficient amou...    3.0   \n",
       "19       19  good news got zombies experience dine like kni...    5.0   \n",
       "20       20  days w friends. so, literally great themed ins...    5.0   \n",
       "21       21  im sure everybody must hear famous world named...    5.0   \n",
       "22       22  perfect place town game thrones lovers. cafe p...    5.0   \n",
       "23       23  average experience. food alright. ambience isn...    3.0   \n",
       "24       24  went yesterday lunch...this restaurant enrolle...    5.0   \n",
       "25       25  location: place located salt lake near island ...    5.0   \n",
       "26       26  missing game thrones wait till final season se...    5.0   \n",
       "27       27  food kinda okay...they provide sufficient amou...    3.0   \n",
       "28       28  good news got zombies experience dine like kni...    5.0   \n",
       "29       29  days w friends. so, literally great themed ins...    5.0   \n",
       "...     ...                                                ...    ...   \n",
       "5571   5604  red valvet cake. really wonderful. awesome. lo...    4.0   \n",
       "5572   5605  pathetic/poor quality lachcha parantha sour ev...    1.0   \n",
       "5573   5606  ive ordered chicken handi food good also like ...    4.0   \n",
       "5574   5607  great experience, nice food service service .....    5.0   \n",
       "5575   5608  decent friendly environment. food delicious se...    5.0   \n",
       "5576   5609  good food.. zomato asusual... timely delivery ...    5.0   \n",
       "5577   5610  best place hangout friends. reasonable price. ...    5.0   \n",
       "5578   5611  ordered yesterday burger amaaaazinggggg twiste...    5.0   \n",
       "5579   5613  cafe serves one best burgers ranchi.absolutely...    5.0   \n",
       "5580   5614  basically brand kfc though get similar product...    3.0   \n",
       "5581   5615  lounge pleasant ambience. plonk large sofas re...    4.0   \n",
       "5582   5616  ambience 5/5 food 4 service slow.it took much ...    4.0   \n",
       "5583   5617  quantity food less n taste poor. small 4 piece...    1.0   \n",
       "5584   5618  food tasty although delivered quiet late prepa...    4.0   \n",
       "5585   5619  food quantity less send salad mentioned order....    1.0   \n",
       "5586   5620  decent veg restaurant serving north south indi...    4.0   \n",
       "5587   5621  best babycorn chili town overall place quite a...    4.0   \n",
       "5588   5622  great place hangout friends family. food quali...    5.0   \n",
       "5589   5623  upon time popular testy food feel sad test qua...    1.0   \n",
       "5590   5624  regular visiting customer chennai kitchen. 1st...    3.0   \n",
       "5591   5625  grilled chicken cheese sandwich fries amazing....    5.0   \n",
       "5592   5626  always delicious.. foods come time .... packed...    5.0   \n",
       "5593   5627  one worst pizza ever eaten ... condition pizza...    1.0   \n",
       "5594   5628  pizza gr8 apart crust everything perfect..just...    4.0   \n",
       "5595   5629  worst pizza. ill never recommend anyone. first...    1.0   \n",
       "5596   5630  worst paneer lababdar eaten entire life. food ...    1.0   \n",
       "5597   5631  superb food though costly superb everything st...    5.0   \n",
       "5598   5632  really nice place grills buffet..ambience inte...    4.0   \n",
       "5599   5633  went january last week.. like barbeque nation....    4.0   \n",
       "5600   5634  worst paneer chilly gravy ever. felt delivered...    2.0   \n",
       "\n",
       "      oneword  length  \n",
       "0     Insane!     114  \n",
       "1      Avoid!     103  \n",
       "2     Insane!     108  \n",
       "3     Insane!     112  \n",
       "4     Insane!     105  \n",
       "5      Avoid!      80  \n",
       "6     Insane!      99  \n",
       "7     Insane!     101  \n",
       "8     Average     118  \n",
       "9     Average     100  \n",
       "10    Insane!     106  \n",
       "11    Insane!      95  \n",
       "12    Insane!      96  \n",
       "13    Insane!     100  \n",
       "14    Average      98  \n",
       "15    Insane!     127  \n",
       "16    Insane!     107  \n",
       "17    Insane!     105  \n",
       "18    Average     111  \n",
       "19    Insane!      85  \n",
       "20    Insane!      90  \n",
       "21    Insane!      96  \n",
       "22    Insane!     100  \n",
       "23    Average      98  \n",
       "24    Insane!     127  \n",
       "25    Insane!     107  \n",
       "26    Insane!     105  \n",
       "27    Average     111  \n",
       "28    Insane!      85  \n",
       "29    Insane!      90  \n",
       "...       ...     ...  \n",
       "5571   Great!     101  \n",
       "5572   Avoid!     109  \n",
       "5573   Great!     107  \n",
       "5574  Insane!     113  \n",
       "5575  Insane!     102  \n",
       "5576  Insane!     115  \n",
       "5577  Insane!     115  \n",
       "5578  Insane!     106  \n",
       "5579  Insane!      85  \n",
       "5580  Average     110  \n",
       "5581   Great!      93  \n",
       "5582   Great!     109  \n",
       "5583   Avoid!      87  \n",
       "5584   Great!      90  \n",
       "5585   Avoid!      84  \n",
       "5586   Great!     104  \n",
       "5587   Great!      95  \n",
       "5588  Insane!      99  \n",
       "5589   Avoid!      75  \n",
       "5590  Average     109  \n",
       "5591  Insane!     101  \n",
       "5592  Insane!     135  \n",
       "5593   Avoid!      96  \n",
       "5594   Great!     106  \n",
       "5595   Avoid!      96  \n",
       "5596   Avoid!      97  \n",
       "5597  Insane!     109  \n",
       "5598   Great!     104  \n",
       "5599   Great!     112  \n",
       "5600    Blah!      86  \n",
       "\n",
       "[5601 rows x 5 columns]"
      ]
     },
     "execution_count": 512,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df=df.reset_index()\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 513,
   "metadata": {},
   "outputs": [],
   "source": [
    "df=df.drop(['index'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 514,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>stars</th>\n",
       "      <th>oneword</th>\n",
       "      <th>length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5 minutes walk kankurgachi bus stop, primarily...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>cheap quality food. ordered food zomato. food ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>name suggests, compact wholesome eatery locate...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>kankurgachi dhaba located 5-8min walking dista...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>kakurgachi dhaaba hidden gem located kankurgac...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>much poor service. asked bar menu 3 times got ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>great time family saturday lunch. huge spread ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>sumana thanks. for. gr8 service highly oblige....</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>...finally reviewing jw marriott ...it great h...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>dinner good discussion quite task. jw kitchen ...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>amazing place two us....great food...great peo...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>went jw kitchen dinner buffet. greeted warmly ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>im sure everybody must hear famous world named...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>perfect place town game thrones lovers. cafe p...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>average experience. food alright. ambience isn...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>went yesterday lunch...this restaurant enrolle...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>location: place located salt lake near island ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>missing game thrones wait till final season se...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>food kinda okay...they provide sufficient amou...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>good news got zombies experience dine like kni...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>days w friends. so, literally great themed ins...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>im sure everybody must hear famous world named...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>perfect place town game thrones lovers. cafe p...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>average experience. food alright. ambience isn...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>went yesterday lunch...this restaurant enrolle...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>location: place located salt lake near island ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>missing game thrones wait till final season se...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>food kinda okay...they provide sufficient amou...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>good news got zombies experience dine like kni...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>days w friends. so, literally great themed ins...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5571</th>\n",
       "      <td>red valvet cake. really wonderful. awesome. lo...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5572</th>\n",
       "      <td>pathetic/poor quality lachcha parantha sour ev...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5573</th>\n",
       "      <td>ive ordered chicken handi food good also like ...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5574</th>\n",
       "      <td>great experience, nice food service service .....</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5575</th>\n",
       "      <td>decent friendly environment. food delicious se...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5576</th>\n",
       "      <td>good food.. zomato asusual... timely delivery ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5577</th>\n",
       "      <td>best place hangout friends. reasonable price. ...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5578</th>\n",
       "      <td>ordered yesterday burger amaaaazinggggg twiste...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5579</th>\n",
       "      <td>cafe serves one best burgers ranchi.absolutely...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5580</th>\n",
       "      <td>basically brand kfc though get similar product...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5581</th>\n",
       "      <td>lounge pleasant ambience. plonk large sofas re...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5582</th>\n",
       "      <td>ambience 5/5 food 4 service slow.it took much ...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5583</th>\n",
       "      <td>quantity food less n taste poor. small 4 piece...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5584</th>\n",
       "      <td>food tasty although delivered quiet late prepa...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5585</th>\n",
       "      <td>food quantity less send salad mentioned order....</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5586</th>\n",
       "      <td>decent veg restaurant serving north south indi...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5587</th>\n",
       "      <td>best babycorn chili town overall place quite a...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5588</th>\n",
       "      <td>great place hangout friends family. food quali...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5589</th>\n",
       "      <td>upon time popular testy food feel sad test qua...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5590</th>\n",
       "      <td>regular visiting customer chennai kitchen. 1st...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5591</th>\n",
       "      <td>grilled chicken cheese sandwich fries amazing....</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5592</th>\n",
       "      <td>always delicious.. foods come time .... packed...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5593</th>\n",
       "      <td>one worst pizza ever eaten ... condition pizza...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5594</th>\n",
       "      <td>pizza gr8 apart crust everything perfect..just...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5595</th>\n",
       "      <td>worst pizza. ill never recommend anyone. first...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5596</th>\n",
       "      <td>worst paneer lababdar eaten entire life. food ...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5597</th>\n",
       "      <td>superb food though costly superb everything st...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5598</th>\n",
       "      <td>really nice place grills buffet..ambience inte...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5599</th>\n",
       "      <td>went january last week.. like barbeque nation....</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5600</th>\n",
       "      <td>worst paneer chilly gravy ever. felt delivered...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Blah!</td>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5601 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   text  stars  oneword  \\\n",
       "0     5 minutes walk kankurgachi bus stop, primarily...    5.0  Insane!   \n",
       "1     cheap quality food. ordered food zomato. food ...    1.0   Avoid!   \n",
       "2     name suggests, compact wholesome eatery locate...    5.0  Insane!   \n",
       "3     kankurgachi dhaba located 5-8min walking dista...    5.0  Insane!   \n",
       "4     kakurgachi dhaaba hidden gem located kankurgac...    5.0  Insane!   \n",
       "5     much poor service. asked bar menu 3 times got ...    1.0   Avoid!   \n",
       "6     great time family saturday lunch. huge spread ...    5.0  Insane!   \n",
       "7     sumana thanks. for. gr8 service highly oblige....    5.0  Insane!   \n",
       "8     ...finally reviewing jw marriott ...it great h...    3.0  Average   \n",
       "9     dinner good discussion quite task. jw kitchen ...    3.0  Average   \n",
       "10    amazing place two us....great food...great peo...    5.0  Insane!   \n",
       "11    went jw kitchen dinner buffet. greeted warmly ...    5.0  Insane!   \n",
       "12    im sure everybody must hear famous world named...    5.0  Insane!   \n",
       "13    perfect place town game thrones lovers. cafe p...    5.0  Insane!   \n",
       "14    average experience. food alright. ambience isn...    3.0  Average   \n",
       "15    went yesterday lunch...this restaurant enrolle...    5.0  Insane!   \n",
       "16    location: place located salt lake near island ...    5.0  Insane!   \n",
       "17    missing game thrones wait till final season se...    5.0  Insane!   \n",
       "18    food kinda okay...they provide sufficient amou...    3.0  Average   \n",
       "19    good news got zombies experience dine like kni...    5.0  Insane!   \n",
       "20    days w friends. so, literally great themed ins...    5.0  Insane!   \n",
       "21    im sure everybody must hear famous world named...    5.0  Insane!   \n",
       "22    perfect place town game thrones lovers. cafe p...    5.0  Insane!   \n",
       "23    average experience. food alright. ambience isn...    3.0  Average   \n",
       "24    went yesterday lunch...this restaurant enrolle...    5.0  Insane!   \n",
       "25    location: place located salt lake near island ...    5.0  Insane!   \n",
       "26    missing game thrones wait till final season se...    5.0  Insane!   \n",
       "27    food kinda okay...they provide sufficient amou...    3.0  Average   \n",
       "28    good news got zombies experience dine like kni...    5.0  Insane!   \n",
       "29    days w friends. so, literally great themed ins...    5.0  Insane!   \n",
       "...                                                 ...    ...      ...   \n",
       "5571  red valvet cake. really wonderful. awesome. lo...    4.0   Great!   \n",
       "5572  pathetic/poor quality lachcha parantha sour ev...    1.0   Avoid!   \n",
       "5573  ive ordered chicken handi food good also like ...    4.0   Great!   \n",
       "5574  great experience, nice food service service .....    5.0  Insane!   \n",
       "5575  decent friendly environment. food delicious se...    5.0  Insane!   \n",
       "5576  good food.. zomato asusual... timely delivery ...    5.0  Insane!   \n",
       "5577  best place hangout friends. reasonable price. ...    5.0  Insane!   \n",
       "5578  ordered yesterday burger amaaaazinggggg twiste...    5.0  Insane!   \n",
       "5579  cafe serves one best burgers ranchi.absolutely...    5.0  Insane!   \n",
       "5580  basically brand kfc though get similar product...    3.0  Average   \n",
       "5581  lounge pleasant ambience. plonk large sofas re...    4.0   Great!   \n",
       "5582  ambience 5/5 food 4 service slow.it took much ...    4.0   Great!   \n",
       "5583  quantity food less n taste poor. small 4 piece...    1.0   Avoid!   \n",
       "5584  food tasty although delivered quiet late prepa...    4.0   Great!   \n",
       "5585  food quantity less send salad mentioned order....    1.0   Avoid!   \n",
       "5586  decent veg restaurant serving north south indi...    4.0   Great!   \n",
       "5587  best babycorn chili town overall place quite a...    4.0   Great!   \n",
       "5588  great place hangout friends family. food quali...    5.0  Insane!   \n",
       "5589  upon time popular testy food feel sad test qua...    1.0   Avoid!   \n",
       "5590  regular visiting customer chennai kitchen. 1st...    3.0  Average   \n",
       "5591  grilled chicken cheese sandwich fries amazing....    5.0  Insane!   \n",
       "5592  always delicious.. foods come time .... packed...    5.0  Insane!   \n",
       "5593  one worst pizza ever eaten ... condition pizza...    1.0   Avoid!   \n",
       "5594  pizza gr8 apart crust everything perfect..just...    4.0   Great!   \n",
       "5595  worst pizza. ill never recommend anyone. first...    1.0   Avoid!   \n",
       "5596  worst paneer lababdar eaten entire life. food ...    1.0   Avoid!   \n",
       "5597  superb food though costly superb everything st...    5.0  Insane!   \n",
       "5598  really nice place grills buffet..ambience inte...    4.0   Great!   \n",
       "5599  went january last week.. like barbeque nation....    4.0   Great!   \n",
       "5600  worst paneer chilly gravy ever. felt delivered...    2.0    Blah!   \n",
       "\n",
       "      length  \n",
       "0        114  \n",
       "1        103  \n",
       "2        108  \n",
       "3        112  \n",
       "4        105  \n",
       "5         80  \n",
       "6         99  \n",
       "7        101  \n",
       "8        118  \n",
       "9        100  \n",
       "10       106  \n",
       "11        95  \n",
       "12        96  \n",
       "13       100  \n",
       "14        98  \n",
       "15       127  \n",
       "16       107  \n",
       "17       105  \n",
       "18       111  \n",
       "19        85  \n",
       "20        90  \n",
       "21        96  \n",
       "22       100  \n",
       "23        98  \n",
       "24       127  \n",
       "25       107  \n",
       "26       105  \n",
       "27       111  \n",
       "28        85  \n",
       "29        90  \n",
       "...      ...  \n",
       "5571     101  \n",
       "5572     109  \n",
       "5573     107  \n",
       "5574     113  \n",
       "5575     102  \n",
       "5576     115  \n",
       "5577     115  \n",
       "5578     106  \n",
       "5579      85  \n",
       "5580     110  \n",
       "5581      93  \n",
       "5582     109  \n",
       "5583      87  \n",
       "5584      90  \n",
       "5585      84  \n",
       "5586     104  \n",
       "5587      95  \n",
       "5588      99  \n",
       "5589      75  \n",
       "5590     109  \n",
       "5591     101  \n",
       "5592     135  \n",
       "5593      96  \n",
       "5594     106  \n",
       "5595      96  \n",
       "5596      97  \n",
       "5597     109  \n",
       "5598     104  \n",
       "5599     112  \n",
       "5600      86  \n",
       "\n",
       "[5601 rows x 4 columns]"
      ]
     },
     "execution_count": 514,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 515,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['text'] = df['text'].apply(fun)\n",
    "\n",
    "# strv=df.groupby('stars').mean()\n",
    "# stval.corr()\n",
    "# strv.mean('cool')\n",
    "# strv\n",
    "\n",
    "# strv.corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### STEMMING"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 516,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['text'] = df['text'].apply(lambda x:' '.join([stemmer.stem(word) for word in x.split() if word in stemmer.stem(word)]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 517,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['text'] = df['text'].apply(lambda x: ' '.join([word for word in x.split() if word not in (stop)]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 518,
   "metadata": {},
   "outputs": [],
   "source": [
    "texts = df['text']\n",
    "stars = df['stars']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 519,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5601\n"
     ]
    }
   ],
   "source": [
    "print(len(texts))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 520,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_word_cloud(df[\"text\"].loc[df.stars == 1,])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 521,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_word_cloud(df[\"text\"].loc[df.stars == 2,])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 522,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_word_cloud(df[\"text\"].loc[df.stars == 3,])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 523,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAADKCAYAAABDsfw/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzsnXV4VMfXgN+7vlmJu4cAwd2LlSIV6gaFKnWn7i6/Utev7u5UoFCkuLsG4u7JJmtZme+PhV2WJJCEQEib93nu8+TOPXfm3M3uPTNnzpyRhBB00kknnXTy30TW3gp00kknnXTSfnQagU466aST/zCdRqCTTjrp5D9MpxHopJNOOvkP02kEOumkk07+w3QagU466aST/zCdRqCTTjrp5D9MpxHopJNOOvkP02kEOumkk07+wyjaWwEASZI6ly130kknnbQQIYR0rHV0jgQ66eRfQL9B16I3xLT6/vDIPgwaditjT3um0Xr6Dbr2WNTr5GRGCNHuByBOhkObmCyiL768zeQ6jxN7qOJi212Hk+nQG2JEeETvFt0zfPR9Qm+IaXfd2/Po01sppp6pbXZ5Wx2tqbst3r/SyZBA7mRxB2kTkwmbeBZOUw3KoGDM+/dSsXg+kVMvQhkWjkypwrx/D5aMvQ3kHFUVKINDqVg8H4DQU6dQsXg+AcmphI6fjHC7kalU5L73GiFjJ6Lr0g2Aur27qFq5hNgZs7Bk7kObkIzCYCT/s3eJumAGuF3I9Qac1VUU/fglnAT/r5MNRXAQIdMuRJ2chMtUC0DhM3PA7UYRHkbkLdeB241pyTJql60CIObR+5CpVAink+o/F2Bet9Fbnz6tF6rwKCqXLzpiuwHJqViy9rdK5+a20RxCw3vQq+9lbN34ATXV2QCoNYEMGXEnwu3C4bAAsH71K0TFDCI+aSxymRK3cJKdsYiSok3euoaPvo8dWz6nrrbQW3dSygR0+iiWLXrYKxcdO8RbT+b+vygp2oRCoWXw8NsoyFtFfNJYigrWk7X/r0Z1/nNzIoHBMn75qpY5D5Yf82fQGtYuj0SrlaivF7zwci3f/WDh2y9CGTJYhaMeqqrdAFxwaTkvvxDkVz5qfAkuF9x8gx6DQUZyopyRI9SMPa2Uyio3M6YFcOtNBrRaiaefN/HdD57/gV4v8cgDgZx1hgaTSXD62WXodBIvvxBE/74qb5sH6z8abeEOavdRwMk2Eki69T6BTCaQJJFw/Z0CEJJc7pGRyUTK3Y82KifJ5SLhujsEHoMmEq67XSCTieTZjwiFMdDXRkKyiJ15nUdOkjx/g4idMUsEDR3lp0/y7EeEpFAIQMRfc4tQRUS1+2d0Mh+hMy/1L5MkEX3/bAEIWYBWxD75kFAnJXjOdQECEMqoSBE/56lWtRd/za3t/swHjyEj7hSBQUl+ZWm9LhLxiWP8yrQBoUKp9Dx7gC5CnDL+Mb/rTY0Ehoy4s8l6DtahUGjFuInPi8SUU4VKbRCnjH9CBOjCG9VXkhDGIJm459mwY3ruLmmqVt8bEiwTgOjWVSEydkd7y998NVjcfIO+gXxj5TffoBd7t0eL3r2UfuXJSQpv/YfW/coLQeL9d0IEIOJi5Q3qaukztMX796SYGD6ZqC8vBbfHGteXFiMpFESccR4ylRq304FMo21UTrhc1O3diS61Oy6rFUvWfhQGI26rGaepxlu/KjwSe2EeB3v09sJ87zVrTpafLvaifITTCYDLXIdMrT5+D34Y8gAdKQ88xb5HZjcpE3vVjRR8/M4J06mlKEJDUMVGE/f0I76yiHDs2bmEz7oCJAmEQKbVgkwGbjdBw07B2H8Ilsx9lC/8HU18IiGjT2swIou+4DI0MXHEzpgFQMGXHxI4aDiGPgOQJAlLdgYVi+c3OsIz9h/i1wbQ7NGh224/ps8kQBdBfOJoQAIECoX2mOtRKLRIkm96sSB3FU6nDaulHLU6EIu5rMH9QoDpQK/3WDj/cmOrRxIfvx+CzPMVINAoQ6GAAz+3FrF8pZ0dOx1+ZV1TFdx8gx6Z5F/3pIkaLr6sAoD8gmZ09U8AnUbgMFRhEZ4XghCoIqMJSOmKTBtA0befItcGYOwzoFE5gJr1qwifcjYui4Wq1ctw1dUiD9Ch0Btw1nncFPbSIvS9+npeQIAmNt7X+OGunpPc9XMyG4CDCJeL/Ief8iuT6XVUfvMjjpJS5AY9cc897r1WvXYFbpsVVUS0t0wdFUP2688hnE7ir7kFVXgkxT99RdIdD1HwxQdeOWPfgeR9/DYIQdyVN6KJTfDo4HRS+M0nR2xDm5DsuReInXEt1pzMRu89Vnr1vYwNa17DYi5DpdIzcuzDR7/pMJRKnV894yY+73fd6bS1lbrHjdAQGXffX82+/U7CwmTs2Rp99JuaoK7O/3caGiLj4/dCGDeplH37nZQXxHqvySTJ04c/ieg0AodRX1ZC9EUzUQYGYU7fhS0/l5AxE4mdcS3OOhP24qJG5QBcFjMui8f356ypAqD41++IvvRKhNOJJJeT9+GbWLMziL/qJpAkzPt2t8+D/gsRVv+Xj7OiEmepr5eoio3BUVKKTK3GZTIBoB898qj1NndEpgwJI+6KG7znB+UOH+E12sZho0P1gY5Fc+5tCqfThkYbfFipoN7u6ZDExA1vVb1yhdqvnrYmOFTOzJsCGTE+gMgYBXUmN1dPLaC8xNdzfvbdSPoP1RAYLOPc6QZv+ZjUbFxOz+eYmqbi5odC6DtYw7K/zLz6ZCU1lZ469HqJkhLPSOTqy3V+7ZtMbhLiG74amyo/HL1eQgi89R/K30ts3HaznutvqSI6Sk5FpZv6eo++zan7eNA5MdxGGPsPJmjoKKw5mZT99Vt7q3PMHHQHlf76PTUbVqOKiMRtr/caN4DEW+8l540XvOfx191G5T9/I9NoMfYfRMGn77WH6q1GksmIPG8aqvBIZGo19sI8qtYsJ2TUeG9vPObSK6lcuQRbXg4RZ56PJj4JZ3Ulhd984nHzDBjieZlLEgVffEDMJZdTvuAP7KVF3naiLrjMr42KJQvQ9+yDLjXN2zGoXLaI2BmzGtz7b+OeZ8Oa5c5ZsCORSb1zmn3//c+HMXxcABeNycNRL+g7WMM7P0Rzx4xi1q+wtonuJwNtMTHcORJoI0xbNmDasqG91Whzard5IkfqS0sw9OlP7faqJmUlpSfaRjidIJOfKBXbDOF2U/zjlw3KD3XHHPp36R8/+cmZtqzHtGW9X9mh7qKDNNZG5bJFVC7zjxRq7N7/Kjq9DLlC8vbyj8bAERo2rbbiONDL3r3NjhDQtZfqX2UE2oJOI/Av5baHQph+rdF7/tcvZh67o+EE3ZrspCPLHNLPONqgsXzer0RdPJP6ijLK/vi5NWp30glyhcS5lxk49UwdMfEKgkJa3qEIDJYz+Tw9k8/T+5WHRXa+8g6nw30iSpmaCYm3MD/rpXZoXSLAEIFcqQGgtrLh8PRk4fP/q2Hxn2YiYxU882Z4k3Kzzis6ooyh32Bq1q1EFRGJLe/Iz6sKj6R2++ZOA9DJMXH7oyFMPlfPK49XsGWtjYoyF0vTk1pUR22Nm3XLrHz2drVfeXXlyRGRczLR4YxAe9J98DTU2iDq7XUA7Fn3WTtr1DRVFS6qKlzkZDiOKLdjs71JmfqSIuQBASTd8QDI5WS/9DQAkedPQx0Z7ZkIvfZWyufNxZafA5KEceBQDL3746isoPDrj3HVHZ/Jw07+vYyeGMDCuXXM/6nuqLIaTeMu8Y2rrPTopyZjT/3BSO5OmqBDGgEh3IyOuwqlTEN+7XbSq1aglKlxuH0x1CNjZrC3ajlOt50eoeMBCaMqgnq3laW57xKgDKJn6AR0ymCEcLMs/6OjtivJ5OxY+R5udyuCiTsYLouZnDfnAFC5dKHftZKfvm4gr+/VF32vvmQ+8xCSSkXMzGvRJnWhbseWE6LvoQy/IhVrTT1bf8kFYMYHo4jrF0LOhnJ+fWAjlur6VtU7odsdKOWeuPr5u59rM30PEmnoxoC4C1iy73XsTnOb199SBsSdj91Zx67iBW1a7/X3BHP5zUHe84PRPY/cXMriP8zcNr2Ye54JZdHuJJwOwfKFlibrUqkk5m9NRJJBbqaDa8/xrHR+7r5y4pOVPPFGBENO0VJncrN+hZXXnqzAZu3wcShtSoc0ApIkY3n+x6jlOkbHXU1B3S7qXY3/aHqEjiejei1llkxi9D2I0fdEQqJ/xFR2lC/AZC9BKdMQpI6m2t54FEba0MsBUKn1DDrtXmqrPQu8WjoS0GglPp4bQ1SMHLeAsmIXG1bZePHRCuKSlPywNJYrpxZy79OhJKcquWBsAf97N4KU7koWzDUz5+EK3G445bQAzjhfR1ofNeFRcrastfHh6zVsWdd+8dnm3TsISE0j8fb7EU4Xtds2tosBAOh3TgJb5+Z6z8O7GFn82i4GXpDE+Nt78scTrdNrUfqrKOVaRiRd0VaqnuRIuEXbu0/enVPFu3OaDjDIy3Jw2/Riv7Kn72o4nwXwyC2lR6znkZubvt6Jhw5pBA5id5mxOKvQKPRNGgG5pPR+kd3ChYQMrcKIQRXOwMhzvXLplcubNAIF+/9pE30ffSmMrett/LLfgU4vo2sPFaedFcCLj1Z4ZR5/JZxFf5iptwv+924Ee3bYKSpwct50Ayv+trBysZXLrjUSFCLnnwUWSgqcXDs7iJc+imD6xEJKitpnlCLcbkp//a5d2j6cwJgAind5VmnrwzX8/dIOtv+eR02Rhcn39z2muh0uK4L/hn9hc/6P7a3CcUWm1xH/8mPkXHdvkzKRs6+j5OXWhTrLArTEv/rEEesHiH7oNqp/+Qvrzr2taudY6dBGADiw+k5qsAhPIVMBsKdyKf3Cz6S2vhyZJGd3xRKQJIRw80/ue4hmLN/zTgAfSDPQWoacomVi31y/Mtlhybw3rLLx/ivVLJ6n4tYHgnnpsUoCdDImna2jR181KxdbmX11KVaz70VUUebiqTfCGTZGw9xvj+5H/bcj3AKn3WP4+56dwJpPPUneakts6ENPXOqNTjo+LTEAqoRY6nMLjqM2x4eObwQO4HTbCdbEUWXLJ0QTT4DSs1JSpwylyLyX3RWLvbISEhZHFclBQ8msXguATJIfdejbc9hV7FrjmTtIG3p5i91BO7fYmXaNkd++q6Ou1vMSP3zSKmufx19dlOdkzw7P3xazG6vZjc7gsRiHGgCAXds8cyFRscfv3xkwsDuWTb6eStjVU3GZzFg27SH0yrNQhAXhrrNS+fUCbOk5JH34MJmX+lISaPt0wZ5ZQNzzN1Mzfw3GCYMxr9tF5Tee+QZ1ahyhl5+BOiUWl8lM9a/LMP21plW6lu4z0ev0OOotToZelsKqD9MB0IWqsdUdv5HSgLjz2ZzvWzuglGsI16dSWLPDWyaXKRmVMguNwoDLXU+dvZythXOxOUwN6os29qRvzFR2Fs8nv3orAEMTL6O8LpOE4EEo5Rrq7GXsKV1ElcWXgypUl0zX8DEEqIIoNu0hvXQJTrfnuxRp6EZJbToKmZoJ3e+ktHYfm/N/ZEDcBVjqK9lbugSAKT0eAKDItJutBb+0/Yf1LySgX88OaQQ63KYyDrfdLzx0VeEXVFg9PfUqm+eHUGnLY37WS1RYc4jR92Bvpc+d0zP0NASCFQWfeg0AcEQDEBjWhb6jbyIoPJUhkx5k2OmP4bA1/NEejTsuL2HVUitX3BzI7+viWZOdxPPvRvjJmA/kIXE6BbU1vpe9yw0yCXoPULN0dyK/rYln5o2BjBinJbmrZ9Qjkx17VtmmsGzaS/j15wGgH9EHRUQwVT8sJuKWiyh89D1yb3qB/HvfQD/qyO4WRXgwtUs2kjf7NbT9uqLt3QWA8GvPper7RZS+8R2OwrJWGwCAT69YzsoP0lHrlbwxxTepKZNLfHPT6lbXe6xEGroxsfvdLNv/Dgv2vMCi9FdZm/NFAwPgFi7CdMmoFAH8ted/XgNwkPzqLSzd/yYL977IzuL5DEuc6b02LGkmqWGjWJP9KYvTXyOvahOndb+LKGMaANHGXgAkhgwmt3Ij4fouyGUqIg3d/NqZv/s5qqz5/BfQjxkGgDI6EkVIkN+1mMfv8js3ThpL/MuPAR530qHIw0KabEMeaCT+9SdRRoWDTEbA4H4AqFMSUSV4cgspIkKR6QIO6BJB/KtPgCQR8+Q9qBLjAJAUigY6Hiv/mpFAU7iFy+vyUcjUGFRhLa6jpjyDbcvfpvvg6aRv/MZTXyvdQjkZDt56vop3X6zm5vuDmTbLePSbDmHaNUY0Wolrzy9h325P727AME2rdGkp2r6pyHRaDOMGYvr74MrYRgzPwY+mCfeZ22LzyR24veLzeUTcdjGOonIqPvnjmHWtK7dRV+4/Ub53cfumXwgJSMDpPnoWUL0qjAHxF7Jwz5xGr9e7fCterQ5fhlqZJCdIG8u+Ul+np9Zehs1RS2hAEsWmPejVnvUgQdoYiky7CbLHYlCH43TXY66vbO2jtTlRgT1IDh+BQRPJ6v0fUWsr8bs+JPky9hT93aC8NZjXbgbAUVSCbkg/nJXVR5S37fG4F911zY/gctWYcJZWIA8KRLjcWDZ4DK49MwfHgfxWyqgIjBPHeHp7wjOngBDULlqOYdwIKj79noABvTCv33qkplpMhxsJtBSzo5LRsVdxStxVDI+ZRrm19Qm58tIXI4S7VQZArpBQqX0vTKdT8PfvLQ8DlCs9dZQW+9waZ12sb0q8TalbtpnAKcNRxkdi2bQHAEmjIqBfVwCUUaFo+3TBbbXjttrRdE8EQNsz+ah1K2PDqVu1ncLH3seeVXj8HqIdUcq12J1HnrMRCPrHnY9MavkqWYVMhYSEw+WfFqHeZfGGtupUwUiSjEBtDCZbMTXWIgzqcExt8DJtS4prdrN6/0fYnI2vM1mf9WWbGAAASfL9Lpvz03bbWpnO+0BOqUbLgfDrZ1L55c+UzPk/yt/9wnu5btVGND26IgvQohs5pHVtH4F//UhgZ/nCows1E01AMBaTJ3RNqw/DEJJEReF2XM6jfylCw+V8+3cMX75voqbaTXiknCnn6rDbWmZQVi2xMm5yAM+8FcG8H+vQ6iQSU5S4D/Nmpaap0BtlhEV4XiYRMXJ6D1BjqnaTm+VZHKbWSPToq/aTGT5GS2Ge0ytzKDV/rSH+xdsx/b0O4fK4qkpe+JzQq84i7PrzcFtslH/wKwhB+Ue/EXHrRQiXC+v2DMxrdjSo71AkCQzjBqIf3htHaSUlL32Fq6Z1k9xBsQHUFFkR7vaNB1fJ/d0FTre9QdnhSEjsKp6PSqEjTJ9CeV1ms9tzuO0I4UapCDhMjwBqXB7D6hYugrXxyCQ5ZnsFNbYiDOoITLbixqr8T6AbPpDapatRRkdSn9n6LADC2rycRM6KKgIG98OyYSuqxDiUEQe9EwJXzYHstuNG+Op1ODCv2Yjh1FGoYiNbrV9T/OuNQFuS0vc8EntMYfOSV+g28FJqyjMI6nc+6RsbLp46HFO1iwVzzVx+YyAyOZSXuti8zs7n79Qc9d5DmftNLcZAGedNN3D/86HUVLq5+NQC3v0hyk/ui/n+m4UPGKrhg5+jqSx3ccbgPAB69Vfz9jdRfjIDPtP4yRyKsDuQVApql/i2YrRnF1H42PsNZOuWb6FuecN4/EMniwse9OTP1/RIQn9Kf3JmPYNwugi75mz0o/tT8/uK5nwkDZj+7ii2/pLDyg/SW3V/a9Eo/V17IQHxuIRvxFZpySMheNBR66m2FmB3mpnQ7U5WZ3+Cpb7pmPpDEcJNlTWPcH0XMss922ga1OFolAYqLZ6XW629hChjd2qsRQgENdZCogzdKTJ1jJTm4YaupEaORq+JYEPWl1SZ84g0dicpfARrMz4BoGfs6dgddWSULidAFUzP2ClolUHkVKwjt2KjX32OgmJkugBin74X5HIKHvAsAgy96hJUsVEoIkKJuvcmqr7/HXtW7uHq+FG7dA2xzz+I22yh6KlXmxZ0uwk66zSCzz8d2+59WLfvOXD/amIevwu3zU7dqg04y3yh47WLVxHzxF3ULlvbVK2tp723ljxR20sqQ8JE/A13HlMd/cfdLhRKz2bQvUd5toXsNeKa4677yXIEXzRBRN59WZvXqxvWS8Q8NktIcpmQBWhEzBPXCt2IPq2u74GNZ4uEgaFtrqdMkguVPECM6XKDUMq1QpJkften9HhAdAkbJYK0cSI+qL8Y0+UGERPo2+hdkmRiRNKVIiF4oAjSxokwfYroFjFOqBWebQUjDd3ElB4PCLVCJwAxLGmmOCXlWiGX+bZQHJro//kr5RoxpccD3vMgbZyYnHaf6B19hogypokxXW4QI5Ku9OraM2qSGN/1VtE13LPtpIQkxne9VejV/ts8quRaMSL5SjEw7sJGn/VEHWPSbhEGTWSD8lFdrxPBuvgDn6skxqbdJvQaz1aWp/acLTRKg5CQxMius4RRGy2Uco0Yk3aLCAyIbZfnOObvnkYjEt56RijC/b/XbfL+bW8DcDLtMXy0I7n3WWLQhHtE2tDLRY9hV4jo5JGizyk3tEndsTFykbc3VgQFen5sMln7P29HPfqdmyCmvztSyORSu+vSeRzb0RwjAAiN0ijG97hTyCS5CDOkCEAEqILFxN73iwP7lbToGJFytQjTp5yQZzRqIkWkMa3Ra7qh/UXcS4+KqAduFfLgwAbX/7N7DCtDwoi98gZwu6hatYyadSuRabQEDhqKcdAI5FotNRvXUvH3nxgHDCFw8AhkGg2qiCjvnrnqmDgizjwfRVAQLquV3DfnoOvek5BxE8l79zVvW6GnTqZi8V8AZO34nexd8xBuF3KFmqjEYezf0jarKuNi5RSXuqiuaXz9QCfNx1RkxWl3c8MvE8hYUYK5yj9X0Ir32mdlZifHD5vDRLUln4TQwWSXe1wm1vpqLPZKksNGkFm2CoMmErO9vHmpME7gZlsRhm7edRyHY163BfO645t+peMZAUki+tIryH75aWRaLYk33Y29MJ/68lIkpYqc159HYTCSePv91G7ZgGnzekyb1yPTaOny0DMAyFRqYmbMouSnr7Hs34syJAxVWDjm9N2ETTkbdUwc9sJ8JIWCmo3rDm0crS7Mm0q6IGNZmzzSbz+EEx8rJzpazrplHh/9OReXcdp4DbfdZCBAK2GvF/zvZRMRYXKSkxSMG6Nm/gIbOp3EDz9bWL7KTv++Kv73VBBxsXLuvK+KBYtsBBplZO+OITjWE/O9ZF4ETzxbw9Lldgb1V/HJe6EolfDr71bue+TIoXEdhQtfGer9u+85CQ2udxqBk59+CecToApGozDQP+F8tuT+SK2tlD5xU9FrIghQh9A7birb8n6lxuJZoJVXuYn+CRd4jYBAsCnnO9KiJzKux+2Y7RVsyv4WmjACQdpY+sadS4U5m0MzCQxOmEaA2pNoMqdiPblVG1HIVHSNGE+4IRWZJMfpsrM2+1OUci09o6b4yReadjIy5RpyKzcQF9QfpVzDkvTX0CiN9Io+g6CAWNzCRVxwf1bub14WgzalvV1BLXUHKYNDRdcnXxLJ9zzmPQz9BwuZRitkGo1XLuHG2SIgpav3XKbRiq5PvSwAoU1OFYm33e9Xb9CI0QIQgUNGisjzLhGAMPTp7yfTffB00Xf0TSJt6OUibejlbTokHD5ULdYti/IrS0lSiJBgj3uoW1elyN4dI269wSA+fS9UaDSSyE+PFZMmaMTrLwYLnU4SezZHi1PHej6DvZujRWqKQgQaZaKqIM5b55J5EWLcaLUAxMfvevyLapUkkhIVJ2To23l0HifjISGJMV1vFoAI0saKST0fEGH6FCEhCaPG87tUyjViTNebRaA2RvSKPkP0iztXyGVKAR53FCBGplzTQF4h14jJPR8UXcJGCUCoFXqhU/l8+31ipoqk0GGt0vu/6Q6SQLhcZM15wq9YptHith2WRbOxmFwAIRpd4wSeLQJDxp6GTKvFOHAotdt9QzG1Nphty986Fu1bRNdUBbfcYPCsHREQaPQs69i7z4HNJigtd7F1u4Ppl+gY0FeFySRY/I/nM/hnhZ0J4zV8833TaXi/+NrM6y8G8/X3FlavbWXscyed/AvQqoJQyT2htdXWAm9ElkYVyICEi/xkdaoQwg2pbMz9BpfbE0p9cNW3XhPRQP4gOZWe7Wftzjr0ujDM9RWNyp1oOpwRcFRWkPP6/0i++1EkuZz6shIKW7gXqzU7g+LvPifu2ltRBgXjttm8G6YLh4Pc/3uFpDsfoma9f4oBp8PMkEkPtjqVdEv55N1QuvcvwlTr5r7ZRkaN8CQ/qz/Efeh0CiRg09Z69HqJCeM1LFpiY+xoNXNeM2GqdVNXJxgxTI1cBv36qLz3Vla5ue3uKtK6K8nbG0t8946X9+RIJA8PJyotiB1/5lFbakMbpMJpc+GwtS49csLlN6OOiKK+spyyxX8SPm4KqrAIMt54BrfdhjI4lORr78KSk0HVumVIKjUhw8eR++mbnvtn3oQlZz/WvGyiz70MhGD/q4+34RP7OPv/JjD3hkVHlFEbVFy99CLeGdRwz+P/GlZHjXeRnVETRYDKk3vMVl9DXuVGX8itJgKzvYIQXRLJoSPYUfgHbuFEozBQ77JgtpdTVLPTT/4g/qvFfb1Qh9uGVtm2qSBaQoczAgCOqgqyXnzyiDK577x8xOv2ogLy33+j0WvC4UCmUGLa6J+/Ji99ScsUPUY++LSO1Usiqatz89V3FrJymk5+ZrEILr2inBeeDub1OcHMvr+affs98nc9WMX7b4awcLGN+Qt9C1pmTtPxxYeh1NW5uf3e5sWidwQCglRc+OowEgaGgoDMNaXUltqYdE8fnPWuVu8nACBTqsj/5gNcFjMl8y0kXXsXmuh4LNn7CB01AXe9nYIfPkW4PJ993MVXo0vuhjkrHVtxPurwaIQQ1O7aQtCgESgMgbjtNtz1bTsSO5oBOBph3YMxxhnIXHTk2Ph/C0K42VrwC2O63kRFXTbltZ7UEAJBoDaWsV1vQZLkmO3lbMr7jj3FC+keNYHRqTcgSTLqXRbWZ3/JprzvSYuc6Cd/tJTceVWb6Bd7HiNTrmFV5odmoU5ZAAAgAElEQVQn4nH9ae/5gOMdIirJ5QIQuu49RdLsh49+jySJ0FOniJjpVzd6XR8cL5AkIZMr292P2Xk0fpz73CBx49zTRPyAUCFXyURkmie0rueUWHHLvEmtrjfh8ptF4pW3es/lGq1Ie+RlYew9UACiyy0PiZhz/eP40x55SYSMGC8AYew9UKTceJ+Iu3SWCOw7WCTNulMY0voIbVxSu3xOaoNK3Lix8XUfQ67vK/rP7NHu/8vO48jHf3NOoIUk3fEgyGS4rBZKfv7miLKGvgMJP+NcHFWVFH39SYPriT0mYwhJZOfqD+k14hq2r/i/46T1vxNJIaPbmzdg3pVL/uu/H7d2uo6N4teHNpG32d/naiqyYgg/tmR7TnMjqSwOzD3JtAEY+wzC2Md/VbDCEAiArTgfZUg48gAdZYt+x1aYhyYmAUd1Bdb87Bbp0X9mD1R6JcY4AzEDI/ju0j+x1dhJHB3LkOv6EJIazHsjfCvZI3qFcsrdgwnvEYK1ysamT3ax788s3E43036aitqgYvcv+1n71lb0kQH0nZ6Gy+Gmx3mpAHxz4e/tnoajk+PDv94IZL30VLNla7dtonbbpiavG8NS2L78HQCE6Azkbzmel6V0+E46bYxMIcNhaeg6UxuUrZ4P8HKE+HG3zYolcy8VK/1dMQcNR31FGcLlRJIrsJeXYC3MxdCjH5JC2SpVel3Yjd9uWsTfD630luUsLyBneQGXfHeWn+wpdw9m40c7UOmVdD8zhR3f7kVtUCFTyPj6/N8ICNMy7aep7P09i+ocE1lL86jcX82WzztGOolOWs+/PotoW1KWt5k+p9xIvzG3UJrXtLFoDZIkMXTwrRgMB3KLK7QYjfFIkoxRI+5Hp/Mkjho96hHUaqOfvEKhZeTwezEa41EoNJw67lkAVCoDY055jICAlqfPPh4Ip4u9N7xN3qtzj7mu7v93U5PXPp7xDxe+Mozb/57COc8OYtQ13bjux1M597lBfDS9bbYKbYzcz98mILkbYWMmI9cbkRRKgoeNRTgPGCQhqNmyjqp1yz1/b11PQEIKxl79W9Ve1tJ8yvc2by5n3TtbGf/ocHpf1I21bzacE7GUW6nJrUUfFdDI3f9OLn28e3urcFLQoUYC+j6J1G1vXZa/Y7n3IMXZa6gu2weAzdy24V0aTTA6XRR9e/s2CMnI/AuTKY/Cog3ERA9h3/7fqa7Jwm43odWGNJAP0IZhsfg25K6vr8VqrUCtDsRiKW9TfdsTVUQQ6timN/AoTTfxwcVLGDmrG0lDwpBkEtlry1jx3l6qC5sOmT1WHNWV5Hz0GuHjzyDmvBnIlEpqtm5AOH0ZWW3F+bgOupSEwFbkcQkdb4KSA9m/IIcVczY0S76+zoEh5sSkKO+kfelQRiB65lj23du6sMxjufdQDr781QHB2C1tG1EjhItVa15o4GoqLFrP0MG3kJE5n4LCtUeUVygO93kLpKYWRRwFeYCaqCvGEziyB/IANZmPfYW+XzJhU4ey81LPhid9f3uY9FvexZbjMz5dXrgCTVwYO6f7doDr+5sve6hp/X6yn2x8fqbra7PQxIXhdriw7Mmn5Kt/sKR70iCHXzCC4HF9UMeHNaiz9LsVFH++1HtelW/mj8c3t+q5myL3M/81Ii6blT1PzfYrq68so+DHT5uso2bLOr/z3M/ebjsFgVOfGEFIahCB8QbO+2gSq17eRMmOciQJ0s7uQpeJCZjy65h/9zLcjqZdmju+T2fSc6cw88/zsFXb+X76n22in0ojZ8rNSQyYHE5wlJrsbSY+vXsXVcW+6Ki7vh1EbHc9drOLzX+V8utLGdjNHjfe7Z8NoK7KwYe3e1KTaw0KXlg3ms/u28X6uZ79BV5YN5oPb9tBcIyGydcnEhyt5qkz1lKR71lDM+uN3qQODkIXpGTUJb5su7f3Xorb5XH3xXbXc+49XUgZFMi2heX8+Nw+6qp8xjxtZAhn3p58THp2Gx7MGbckE9NNh0wu8eebWSz+uGH23uNNhzACmsRwoi4eRUDXaLo8fgkAGY9/S9iUAQSP7QUyibrtuVT8tZnkhy8k42HPhFiXZ6aT89LcRu89VnoMvZwtS187umAzsdmqsForSIgfQ07uUgBkMgVutxO7vYaamlziYoeTm7eyUXm9PtpvFHDMSBJJj1yCrncCpjV7Me/KI+nhi3GUN77Jx9HYf/fHyA1akh+9pEmZ2Bum4Kiopfqfnci0KkIm9KXL/64g85EvMe/IxbIrn/riatRxoUTNGEfO877QO3vev2ek0xya8tUvfqzh9pkxAyPoenoSH536PW6HmzEPDKX7mcls+Xy33xqBH2bM8/5dnW3iu2lt8+I/iEwuceP7fYntrmfeW9nk7jDRa1wY1aW+hS/dhgdTmG5m7ksZ6ENUTL0zhZhuOl6d0TKDPnZmHEqNjB+f24e11kVloc/I/PjMPiS5xNQ7UvjtVd9+DQcNQFi8lju+GMD+9dW8d+N2Ln60G7Pe6O2nw43v9WXNz8Wt1tMYpuKG/+vLqu8L+fXFDJQaGSExJ2aHwMPpEEbAllNGzstz6dnzJu8LXB0dTPC43ux74HMQkPrMZRR98Q+FHy8mYfZUJEmi8MNFjd57MiKEYNuOz+na5UxGjrgPmSRj9VpfT7qgcC29e00nN29FA/lTRj6A2VLGtu1tt3jNOLQrut4JVP29lbzXfgPAmlVKylPTcZlbHtNu2VtwQO/Gr+t6xhN65mC2TX3aW1bx+wbS3r+ZmGsns+/29zHvzvPKAtSsbPxFKJNL3h90J6AJ1uCqd4NboNIrCe0aRMGGE7+TWO9xoaQODuL/rt/GzmWeEXXGRv/9NM68NZlXLvPNt5Vkmnng16G0lMiUAJ45a12j34ODow672UVlga3B9YnXJWIzu/jwjp046918+fAe7vxyIN1HBLN3tWf0n7Wlhq8f2dNqPcMStCjVMlZ9X0hhest3GGxLOoQRABBugUyl8Nu3VlLKkSQJIQQZj3l6/4ogHa5aGyBQGLVN3tsSBk+8v82e40hYrZVs2/F5o9cqKtP5Z/njzZJfvPRB79/rN/pcGE9sn8pjfX5rli4hkwcCUPSJL9Klbksmlj35qOPDm1VHS4icNhrwd/EcRJvSst2UHtx8Dm9OWdDA/x/dK4gZ749izshj38P4RJB69o0UrJqLtfzYVnJnLsr1Lvqqr3Pw89UL2kK9FtN/UgQlmRavATgcfbCSlIGBfmWF6Wby9zS9w1xTmWHWzS1pdUdgwORwtAYFr2wd61fea1woe1dXoQ9Wel06rdUzc1MN9wxZzqlXxXPLRzHogpQ8NHqln8vpRNFhjABCULV8N91fvYr6khqynv2R8j83kfrsZQiXQJJJ5L+7gPCzh7D/Ac8QN/XZy6havrvRe1vChoXPH48nOq5E9wgkOC6AXQtbt7n6wS9ta2LDJXnL98c92GDB2/OOIth6jFFaOlJk7/6577S3Cm2Ktc6JIUzV9PVaZ6Mvbn2QL4RWCP8Xqj606fpai6XGye4VlSx8zz+Q5OAL2lrrRBfcMKy3pXra6pz8+UYWf72TzcAzIrjq5V68cdXxTRvdGB3HCAD5//eX33nlom1ULtrmV5Y++xPf33f5/j783n87aeOjqG8kVr651Jd5humqqCCstb5UE8rwwAayksr/a6SKbChzNOyFlej7JVMxb+PRhZuwSxFdjQy+NAWAcbf1pN7se36lRk7XcVFkriptsW6dtA0ZG6oZMz2W7sOD2bumYVCFyynI2OCfzjymm46gKLX33FReT3iC1nueOrj1OXeU2sYj5NPXVpHYx0hBurnRTpDLKeg1JpQF7/qMxLHo6XIK1s8tYdoTaa15jGOmQxmB402oOp7ugaPQKYIRuMmq3URG7foT1v6oK7sQHK+jy4gw9i4tofeUGH68fzNZ68qJ6RnIGQ/0RqVT8Pere0hfVoLGoOSBVVO8Lp7rvx3Dwld2U5Fdx/CZybgcgoEXeMIP3zpnKS6nm1FXdmHgBQloDEo2/ZTLotf3NKpL3eZMQk8fROjpg8jf51ndq++bhDLU0GBOQNczAes+34hDEajDVdu8TbcPUr10B6GnD/KsJzvsdyfTqHDbfJOHTlPjYZ61ZTaMUZ4fXkyvIFxOX0UOq5Ptv+Wx7J3Gn7elhPYYRkT/U5EplLjdLkrW/4UiwIDKGIohvjum7J3IlCqq9m2irmA/6sAw4kZfgDownNKtSyjf4Zngj+g/HplKTWjaMJAk9n47B6fNjDGxJ8mTr2T/b+9gLsryfRZKNWmX3IMkV+Cqt7Hvp9dw2a30mPaAny6V6RvoOeNhyrYvJ7TnCBRqLRW711K0tm0ne1vC1oVl5Gw3cdUrvfjj9Szy99TRe1wo/3yRj6nM8//97bUspj+dxvq5xeiClZw9uws5203eOnYvr2DQcz2YcmMSe9dUMXFWQqtXMvcZH0a/08KpKrahNSi8/v7572Rz34+DmfV6b1Z+V4hGJ6fXuFB+em4/5mrPaCCxr7HZetZW1HPaNf569p0QRsqAQPZvrKG2op6gCDX71rfPfh6dRuAACklF/5DTqbTns7v6H+QyFTZn6yJhjoRBGUqto+k1BrpgFW+f9w/3LpvEvOd30vfMWAp2VDP9zaH8/PAWqvItXPPpKD66ciXmysZ3I6optrJnUTGl+2tZ+UmGt1yukKFQy3hj6hIM4RpunTu+SSNQsyYd8648Qib2R2EMwLwzl4iLRmEvrEQRqPPKOWvMRF9xKqrIIByl1agTwrHnlaMI8slICjkB3WKQ69RIEiiDdRgGdqG+rMYb1WPelUfZT6tJfelqajfsx22xowg1ou+dgDWzmPw3fH78+qJK7AWVRF99GvWlNcgDVFj2FFC3LZtvbl7NXcvO4KvrVx3XNQF1hRnUZO3AaTOjCY6k63m3UbLpbxRaPXu/fYHeVz5J9sLPCe46iLrCDJImXUHe0m+xlOXTc8YjWMrysZR4epJhvUax4+NHUGh0OG2eSUJTzi7sNQ0jnmJHnsOeb17A7axHpQ/GZfcY230/v+6nS2X6BlSGEGQKFXu+fh5lgJG0afe3qxFwOQVvXr2Vs25LZuJ1iQSGq9i7ugqXw/dyzNpcw3n3duGm9/thM7vYtqicX+fs915f92sxIbEaRl4cw+QbEnl15maufLFnq/TZvaKS6c+kIUlQmmVh72rPKLSywMaLl2xk6h0pXDGnJ1aTk72rK3HYfL7EN6/ewtmzU5qlp6nUzif37PLTs6asnsS+RoadH43WoKC2vJ7/nX/iOpyH0mkEDqBXhqCUqcmoXY/J0YahloeRoOvLzuolTV4vy6zFYXdRV2HHVGZDo1cS0ysQe52TjFUevTLXltP1lAi2zM1vcftrv8wGPL3myrwjRCUIQfaT3xB1xakEjkzDMLALWY99RfgFI/2MQObDXxJ95QSCxvZCplFhzSjCtG4fIZN8q2DVsSF0+d8V3nNtajTJT0xDOFxsP/85b3nRx4tIeXYmYecMQ6ZU4KiqxZJeSNXi7f6quQXZT31L6pwrkWlVOKvqqC/29aJyN1bgPs55btRBEUT0H8fBVBhytWcEYqsqxu104LDWYS3LJ6T7EFSGELShMSSffo33fk1guNcI1BV4FiAeNABHwpjUk7x/vgOgvs7nUkmafIWfLgdTc5RvXw6Aw2KivhGjcqKx1Tn54dl9/PDsviZlXp7W9Gp8IWDeW9nMeyvbW/bEJP9sv/cOXd4sXT6evbPJa2U5Vj66s+nr+9dXH5OeOdtMvHZ5265jaS0nvREYFn4hFbZc4nV9UMhUVNrz2Vm9BJvLNxNvUIbRPXAUQcooSm1Z7KlZTr3b5444Lfp6Nlf+iUaup4thCADry3/B6vIM3/qFTCZa2w2AkRGXeu9bXvIFZmeVXxvBqhhKrBkN2pBLCroYhhKlTUUj1+MSDlaWfu3Vc0DIGYSoY1HKNMTrenvv+6vgLQS+Hobr8AU8EhTuqEalU5B6SgSVuWZShofxz7vp2Osc2A/4vZOGhBLdw+eLt9U5CYppmALAVueLPjhaoJTLbKfg7Xl+k7WR08f415ddStbjXx9+q19UkS2nzC/080hkPth4dNTh2Asq/BajHcr3d65ttLwtSZp0BTs/fQxXvY2owZPRx3QBQLh8uYmE2wWSRH1tJbaqYqr2baJk0yK0oTHYqn1zE64WpJGuzd2LTK7A7XKi1AXitJlR6oxkzfuogS6eun0hkB0laDbt9nHseW3pMdcj1yqZMO8mdjy/kML5u/yu9XvyTLY+2jGixI43J70RAM8LeFvVAhQyFWmBo+kfcjpryr4HIEARyLDwC6i0F7CtaiHdA0cyIPQM1pb5RwAl6vshlxTsrlmGQRmGzeVz9eypWUGFLY/ewRNYW/aj98Vtc9c1aGNTxe/0DBrboI1BoWdjUIaRUbuOmvpSFDIVdpevZ7e7ZhkSMroZR5Bu8i3oOdQANEW91cVXt6zjjAd6ozEo+e3JbZRneXT74+nt3DFvAhlrykhfWuy9Z/232Vw0ZxCzF56GtdrBOxcdv5w5JxtBsQHUFFmPa9bL8p0rSbvkXlwOO5V712M3HSGNiBBkzvuI2FHn0mvmo9hN5WT++UFTW90CkHDqNNSBYSScOh2nxUTBqrlYSnIoWPkzadMfQJLkuGxm9v36FvWmSnrNfLR5unQAhOMYk/w1A6mp2NL/Iu29l8DR9hMYFn6hkEly73mYJlFMifXldO8ddKoYF3WVVyZYFS2mxN4qQtXxXpnToq8XoyNnCglZk+2Ea5LElNhbhVZubHCtOW1Mib1VhGuSjpr/u1fQ+HbPQX4sR5fnZope39zT7noc6bjp94li1Kxu7a5H59G+h1yrFJOW3i5ipvRsd12O19EW798OkUXUfUiXqdLu7wcPUcdRaS/wytQ4PMNso9J/QVO5LYfm9LobozltON31lNmyW1V/J21LYLSWvE0duzfcSRsg2luBjkGHcAcdilu4cIlD4r9lGmICuhMT4J8WVi3X+Z0f6r9vKc1pw+5u36Xfh3LGWRpeeSOI886qYNfOtl2BmPFA83z27Ymp2Ira0Loc/Z146HPLSLpO6+c9z1uwj/VP/H1EuaZkmsspX11JQEwgxYvT2fak/6LBIa9dSMWGXIRbEH9uX5QGDXVZ5ay90T8VTOSYVPo8MoW6rArS317ewBAcbANgwbiGub+63zyGsOFJaCMMOOrsVG0t8NPloB5xZ/fx6rD37eVUby9s8rkmXBHL8HMj2bygnD/fafl2nd2GBpG+7viFj3YIIyCT5N5eeLgmEbnkU7vElkGgMpJVpV8jjpPpb04bOkUwoep4KuxHzgJ4qO7Hiz9/t3H/Q8ffr9pcbp+t57WXm15S39b8+eQWhlzWhYl39yZjRQnmKv9Q2hXv7T1hunRUtr+5iu1vrkKpVzP1r6uPKLfnk41HlGkuK6Z/AkDfR09v9HrKjCH8c/FHZH3pCaU0pIYTMjCeyk2e31yv+yYSOiievye+CUDUqd04PIHuwTaGvnlRo23sfWsZe99a5j1PvWZEo3r8Pfktrw4jPpjOhtk/efU4nEWfFuCwu9Efssr4sie6EpGkRa2Vs3NFFb+9ns3p1yeQNjIISSZRnmvls4fSufrFNAacFsaeNR4j8PZNO9p81XuHMAIDQs4gu24LCpmKHoGjqar3LUzKMK1nZMQlDAg9g8zajWjlBsI1SeyuWY7D3TA5VGs4tI088w4UkqpBGzX1JfQPmcI+0xpMjnIUMhW1jnK/yWGAcE0ykdou2Fy1KCT1UY1GUwSHyPjx11C+/drC9JkBvDKnjl9+OvJo55Mvg0lKVuBywscfmPniMwsXX6rluht1aLUS9fXw+it1/PyjlX79lTzxjJHQUBkKpcT8P2w88aipgfz4U8owGmX89lcon31k4ZLpWv6aZ+elF2qJjpHz7P+MDBik4qyzPRkSp0wox3Wc7dOFr/gSefU9p2Gu/k4j0DExpZfiqPF9x20lJvRJvn0ljKnhVG315Vmq3tG6lCmHYi0yIcllCJfvzWtK90V22Uo8EYbqkJZtxvPt0/txOgQyucQzi4by+xvZDD83kg/u2k3erjqkA476T+7bS+pCI2/dsOOYn6UpOoQRMDnK6BcyGbmkpMKe5xdnb3WZWF32Pd2MIxgadj52t5lyWx5u0fqUCYdzaBt9gyfjFPYGbawv/4WuxuGkGAajlgfgcNtZXvJFg7oq7Ln0DpqAhITZWcXqstbnD09MklNZ4eaay6v4aW4o69bWU1jQ+BtWJoMXn69jx3YHgYEyfpsfyo7tDtauqWfBX3aqq9ykpir4/tcQfv7RyqzrdfzwrZUvPrOgUklERnm+lYfLHyQuTo5GKzFpfDlbd0fy849WMjOcXDWzijkvB3LP7JpG9ToetFWCuLCQ7pRX+gxG99SpOBxmMnMWM6D3lWi1oQjhJq9gFflFazEa4uieOhWVUockySkt30l6xvHbS/m/Rn21fydHCPwS9KiCtTi2+Tp+zrqWZ7tNOK8fkeO6oo0JRBWoRVLIKPzLP1vtoXp4w6xbGG10ySOpqAPkOO1uAowKJJnEO7fsZNI1cYTFaVj4UT7bl1a2WP/W0CGMwD7TGvaZ1jR53eKsZktl04nH/i5696htlNmymV/wRqvbcIp6dtcsY3fNsiZlALZUzj+qLi3ht19t2GyCXbscDBiobNIIxMbJee8j//wlySkKQmrczLpOhyTzfKGNRs/L/vtvrDz5nJFefZT89L2V9es8LpUuqQo/+ZnTfF/Uzz72jHpyspxERcvIzKBDU16ZjkEfQ21dITKZgsiwPqzb/DYSEvuzF1BbV4hCoWXYwJsx1RWQGHcKRcWbyC9ai0ymQK0ynjBdz5p3Nbnz9rDt9VUARI1MpMuFfVg522OEdNFGJv9wGT+N8iSlSz6nJ8nn9sKQFEz5pgJ2fbieql0nd14lcZSFLfXVVhQGX05+pbHl+flTrxnBnteXUrW1AHulhZgpPVqsR3PQBSp57/Zd6AIVDD4zAsDjAnowHV2ggsf+HMy9o9YghECplnt+b8cp+WGHMAKHk3Dv+eS+8BPQNttGdmTkB/6DR+uISBKMHl7WwBWza38kU6dUkLHfSWiYjHWbPV/IZf/YmTSunNMmqbn7fj3FRW4ef8TEm/8X5Cd/KLW1nh/HYR20DowgLmYYu9N/Jjy0B6a6fGz2arSaEPr1muEnGaANo7B4I91Tz8agj6GodDPVNdknTNOqPaUEpvr2kg7tG01Ir0hvLqbAbqHea/1njyblgt4Ur8whf+E+Ui7sw9i3z2PFHb9RvqXpCc6Tndr95QQPiPOeB/aManEdRYv2UrjAl0rFkHrsadOvnpNGTFcdCqVEVJcAfnk5m8AIFbe825uasnoK9nrmy+78rB/OejeSTOKfLz3/B+GGjfPLeOD7gVQU2Hj3tl1HaqpVdBgjEDltDIZBXXCUm1AE6dAkhhNx0SiMI9Ko25oNeDYZUUUFUfKlZ2FU5GVjqducSfh5w3E7XKgiAqndnEnJl/8QMmUgQaN7Islk1O3I8d5zsjCo9zVs3PHhEWVsNsGVV+u4eJqWV+bU8cdvnqHwG+8EERkl5413gigvd/HYQyb27Hby7kfB9OqlQKGUyNzvZNaVVXz+qYXPvgqmziz48TsruTkeK/HUs0Ymna7B6RDUmQWff2KhqtLdQP7D948eFfV/b5tZsTacqio3U6cc/9BNQ7iGMx7tT8KgMNT6hl/xp/v+0uy66utrSU4YT2z0EFauexHw7OqWX7iW7DzPd0avi8JiLUcfEMnqDa8ihAtdQATjRj7C0lVPtc1DHYXMn3Yy4vkpaMN12KutpF7SF4fJRtLUnmTP3UXimWkUr871GoAFl35FXZ7HRZf+1RZOeXUqY946hz/O/AR7desj6VpL4kUD6H6zbzV61KmeFfy7XlxE/u/N84fveH4BkeO6ctrCW6jLqmDP60ux5PlSaxzexqSltzdoQxcfzIT5NyOcLkpXZpL9dTOy2h6Fj+5pmJ9rzrSGKaMbKwP45qn9jZa3FVJbDG2OWQlJOqIS6thQEu46l313fYQkQbe3b2DvDe8gyWV0f+9m9lzjceNICjld/ncF++/+GIQg9cWrKProb2JvOoP0W98DoOtLV5Mz5yfibzuLjAc9u5KlPH0ZmQ9/eSQVjgtKRQBqlYE6S8Ndno5mBIJDZGzYGkGX+OImZf6rnP/CEFJGRbDhm0yq8y3I5BKRaYGkTYjhi1krKMtofmJAlUrP8EG3U1C0jozshd7yfj1noNdHI5PkmK1lbN35OV2TpxAe2hMh3DhddrJyl1BStu0ItbcdmjAdZ/x6Oavu/gOHuZ6xb59H4fIsHHV2Nj6zhNN/nkn273sI7RNFxJC4Juv5+7JvMGX7XpwHo4OOFP7ZlMzF5+h48clQug/Lw2oTnH+Wjp9+P3lCqf8NCCGOeczdIUYC6phgbDmlIARC4Lep+aEIpwvTunQMA1Nw1dm8biJ7QYV3BseWW4omPgxVdAgpT112wp6hMaIj+qHTRrA749d21ePfRsKgUBY8v41tv3km3eUqGa4f3JiKrYy4uhtzH2p+787tciCXKSgs9r9n666Gk/4htkACyupZXfAFJnvr/OtDYi4mt2YTJeaW9f5s5WZsFRYMySEgBLU5VVTuKCHxzDSUBjXaCD1Vu0oI7etxkWx9eXmjG7hYK9o282pltZtZMw288b7p6MKdtAsdwgjYi6vRJEZ4Hc3qOI9/U7gFMrX/tpGV8zYRffVpOGutlM9dhyrciCY+DEkmIQRoEiMo+XYFjrIash79CuEWSIr2WTgdEtQVu/3ERc38V9AYlX4ZUvVhGmoKLexbVsyIq7q2qK6EuFOoqNqP1Xb0SI1V+Z8zNvG6FuvbVlTtLsUQH4hco6RiWxGVO4rpfeNwQvt4XvxVe8ow53VHj20AACAASURBVNfA4DjKNhViyjr+0Sc//mbm0vP1vPepb/T15v/C6JqiJCBAYvSZhTz/aAgBWonhQzTMnWdmxVo7S1dYuWaGgYvP0SOXw2nnH3u457ESOfs6qr7/nfq8I8+bBAzqS+CUcSjjoil+9o2jyrc3HcMI5JVTuzmT1DlXUl9STX3RgeGqENQs30XXl6+mvrSGnOd+wGmy4Ky1IkngKKtBFW7EWWcl4Z7zUYYbMW3YT31RFRXzN5H89Axwu0GSyHzI17MLCUwhJWECel0kNnsNRaWbySlY4bselEqXhNMw6KJwuuyUVuxkT8Zc7/VBva/B4bSwbc+BfY8VGsYNe5id6T9QVLaFvmnTCQ5MQqnwxBbHRnkymy5a9SjikBCA5LhxxEUPRaHQYraUkJ41j2qTZ3RTVenudAU1wf+zd5bhUV1bA37PuGXi7gIkwZ3iUAoUalBqVKj3Vm711m799quX3uqtU6ctddpSirtbcAJx90kybuf7MWXCECFAQpKS93nOw5x99ll7nwlz1t5rr71WbYmFoDgdhbs8L7m4wcHsKTYjV8mQyluv8MND+xEc2IPdB+a3V1fblJoD5YQNiUERqCbzy53UHKzA7XITPT4Jc2k9thoL+X8eJvGS3qReP5gtTy/12VErU8txWtp2h7koinzzo5EbrvajvMKz3nT/41XYHSJSaYMDwYq1Vg5nO/DXS5k0Tk1+oZMrZ+iYPKsEUYQhA5Rs23XyLp9tSdlrH7Sqnnn7bszbdxP9wqPt3KO2oUsoAYDSL1ZCExELit5vSBsZOLEfwdMGY9pfQPHHDbZJR2W915voKDXLd1OzvLG9NkAfz6A+N7Jp51tN2uoBkmLHs3X3e95znSacsODelFc1H3/8WHYf9LxUzhl4N4a6vGbNQYVlW8gpXAWAnzaS4QPuZNn6xonYu/Hl8+vX4rA2uEFF9w1kwj/Tqc438uGslS3c6UtZxe5GNn2FVM3EhDtZnPUqOkUwI6KvZl3Bp1idTZs7JiTczp7yP6g056KRBzA8+irW5c/j3MR/crh6HVk1m1BKtcgkSkwOj9Jyi25CNcnU28tPKrFR9k/7SL9lGNZKE/l/ePY35Py0j+TL+rL3fx4X66rdJfw46l2ixycx5q2LCegZgqPORvWBcrK/30NlhmfEPXP97T6yYyf3IHZyD2zVFn6/8FMAQgdGMebtixvVAbyuqCDw1od13H+HJ1TD1HM1XHeFjtm3luPvJ6HO6NFC5ZUuAgMkWG1OeqV48vEqFAISCbhcsGN3ywpAmRhL0OwZSPx0WHbto/qbX5DotEQ+chf16zbjN+4cDL/8iWmTJweALCyY8HtvAbebumVrqV/liewbdNUlqPulIcikuC1WSl98B2VKAgEXnoc8OoKy1z7AdiQXAN3oYeinjEOiUCA6nRh+XeqV35XoMkqgNdSs2E3NCt8frWl/Aab9rd+QVWcswmypZFDvG8gv2UBuoa/fv0SQUlnjmxDDaC4jJnJ4q5VAa3E4GuyzVlvHpJ7ripiqfV8Yi5/fzeLn236B1mivot5eTqAqihJj00rA6bZRac4FwOwwUGVpiB2TV+tJKmJzmdApgr1KIFgdR5Rfb1bkvnNS/bHXWo95+XrIeH0dGa+va1S3aFU2Rauym5V1vJymqNhZ3Kp6AO9/WseUiRq27rTxyD3+/PxFOKXlLu56uGlvsexcBx9+Uccf30bicotcOqcMs6V5/xH95HEY12+lftVGZKENu4hlYcG4602UvzmPiEfvwpaZjbOmltDbrqXo3y8i0aiJfOJe7HlF2HLykei0FD/1KqLdgSwoALfZgmX3ASy7DxD19AM+bVozszDv3IvbZEYeGUbEw3d2K4G/A263ky0Z7xIXPZq4qJFo1WFk5i7yvpClMiUOZ+PFM4Vc26jsKMLxAUxOgfaKi9TN6SKc8t/G6T5WWTX8H1HKdBisRY1v6GIs+MUEeNZmTGbR6xk0/iJf+/4j//Eovw1bfMO8fLnAyJcLWhdzyrhuC0FXz0QRF03V59/7XDNt2YXocOAoKEaZHA+5BShiIol5+TFvHVl4CLacfOr+WIFo95jEnNUtD7zkEWHozxsLEgFEkGjULdbvrHQrgSZwumxk5y8np2AVY4Y+SF/llezYO89zzWlDIWscJ8TuON71reFHLW9BQXQ0c17vR2y6Hv9wJXaLi4J99bx3k68njEwp4dybE5l0WyJSma9Cuz/d4zYpV0mYeFMig6ZHoNbLyNxQzR9vHaGq4Mz7nEukQpOeL22NThGMnyIUg7X5RUupoCBEk0ilOQeNPIAQdfwJ5ZYaD1JhziUpcDjZNe2fJe3vgGVfJsVPvoq6fzoht8ym8sOGdRxBKkF0AILwl/+IgOhyUfjQc6fcnkSnJfS2ayl59nUcpeVI/XTEzH3ytJ+jI+gS+QQ6ClF0kVu4hgC/WJ+y4MCePvV0mnBqanO853ZHPSplQ6rHQH1Ck/JdbjsSSceGPI5I1rFneTm/zj3Mlh+LieurJyDCd7v9rCfTmHRrAh/csoPfXz+CyylSlmXilUs8dlSpTOC2Dwcx5c4kSjKNbF9YQuqYYO7/bjjhyWdeAf5758VNptZMHBHKjfPHnbZ8l+hkbNxNDI6cyf7KZd71gJEx16GS6hgQfiHDo6/CTxHKjtIfSQ4cwfj42xgUcQn7Kpa0qg236CRSl0a0X+/T7u/ZgCI+BkQR8/bdqPv6hnrQjhiEPDIceUwk9uw8nJXVOMsa8i0rYiIRZJ7xsH7qeAS557M0wN9bfjwSlRIQcdV6/va68Y2jjXYVumcCxxEanEaAXxyGujxsdiPR4UOoOW77v79fDOkpMyip2IVcpiElYTIbdzTEJq+sOUzvHv3w94tDpw0nIXpMkyaD2vpCIkMHoNdFI5OpqDac+WA7L124wec8f28dvUYFs/kHjzlCrZczbEYUm74v4vDmag5vriYsXsOQS6Kor/TEExp5ZSxJgwP5/fUjLP/AowzXflXAw7+O5Ir/pPPm1VvP7EM1g7XeQXCC7rTlSAUZa/Ibb+TbUPh5k/U3FzXOwbw461Wf8yqLx+tra/ECb9n6gk9Po5dnF7rRw9AM7AMuF9VfNKR9FR0OJFoNYXffSPX8n7wmnvL/fUr0i/9GkElxlJRT/vYn4ATR7iDq2YcQJBLcJjOlr75H0LWXooiOQBYWTMgNV+Cqraf05f9Rv2ojUU8/gNtqw7hhG84Kz/pG6G3XIAsNRhqgJ/S2a6h47wvshR3v4tocXWLHcDftxyWP9CJhUAAB4UpUOhkyhYQVH+Wy6A3PZqWgaDWPLx3Nms/z+flFj8fJzMdTGT07lsfPWYW51sEza8Zht7h4borvAuSkWxOZdm8K79+yg0Pr2z9cRFgPPUOuTGLQZQnsXVSI3XRM8iGVlB7jI8jeUM6PD566UjrWO6ibzo1EpyX2tafIu/Whju5Ku3HW7Bjupn2I7+9P/ynhrP0qn5JDRsy1DnTBCuL7NZiyqossHFhbyYjLotn6SzGxvfUMmxnFtoUlmGsbfMqbDBj3V9mZGmfUV1jRR3gW56J6B+ByNjTssDjZ82sBa95tHMflZLC7LN0KoJu/Fd1K4Cxm3Jx4PvzHTooPNfiiJw8NbFTvy3/t4a4vhnLP18OoKbGy6pM8ln/YsAZSnmMiYWAACpUU+zH++WEJnvWAyvy2DUXQHBaDnW/u3MgDa6Yx/7YNGIrPTLvddE7cRtPfehbQVnQrgbMYqUygtszXLW/YjChqy3z97PtPDSc4Ts1DA5Y3KWf7byUkDw1k7Jw4lr3vUQ5B0Wr6Twknf08t1YVn1kMof3sVbne3hbEz0HvC7ThsJjI3NL1e0k3H060EzmIOrKnkuv/2Y9svJSg1UvpMDEWplTVSAj1HBGE2NB9OYMuPxfSdFMa0e1KISdNTVWRh2IwoXA43C5480Ox97cV393W7VZ5JZAoNCrU/5trOu/jZTfN0yYXh868JYdYdYdw0cj8JaWpufiKKx2d37TRWgVOn4LbZqF25qqO74kPamBBmv9SHQ+uryN5WgyAR0AbIGXpJFNoAOf8e3vowDN14iEnVUXiw6U1QD30/hN/eyGb/2jOTWrAtSB1zE3azgeztPzS61j0TaF/O2oXhC29oyKCUe8BCdNLJp5FrCqlWS+S9d1O/aTN+w4YiyGQUPOvZUOI/YTwSlQpZSDDqlBQKX34Vt8mEMjaGoBkzkKiU1Pz2O+b9B/CfMB55SDCiw4mgVKJJT6P8i69wlJSgGzHcK7vmj8UYt20n4rZbUCYkIDqd+I0YDkDhS694gtt1MFf8Xzq5Ow18+eAen/LyXDPXze3bQb1qmRFzUrDU2sn42ROi4ZqPRhHTP4i8bZX88uh2zAZ7h/avOQXQ1viHpRDbZwoa/0gQBGymGjL+9F3U9g/vSeroG3E5rVQX7iYv43dcTs9M8PgXuEyuZuiMZzmy+Wsq8jwbCnuNup7ASI9ffnhyg6/8xgX/8n4W3W6i0yYRk34u5toScnf9Sn1lw5qSJiCS+H4Xog9NRHS72LHoBZw2382X/uE9UGoCiE6bhFITgM1cw/7VH2AzVdN7wu0YSg8DbiJSRmG31DZqo60Z+P391O3MJeu5H09cuZPTJZWAVOqr/I71Ajld5CEhuI1GCl94CXlEOLLAQJw1nqilfqNGUvru+1T/+DNukwmJUkn4zTdRMf8bnFVVRN59FyVveeK9SLQ61Km9qPj8C2y5uegGD6Lmt999ZEfdew/WrGxK3/+Q0NlXYi8p7VQzAYlUwC9Eic3UOG9x4kD/Ju7oHPS/OI6MhQ0xekKT9ax4Yz+DLk1gwj3p/P5M0xmcOgNtOTFPHXMz5dmbyMv4DYlUjjbQN5mMf1gKaWNvZv/q95ErdcT1PZ/UMZHsX/WuTzTblsjZ+TMafTjGmkLy9zSdgzswMhWVNoiDaz8mts9UUkffwI7fn8flsKLSBtNnwp3UVWRzcN08pDIFvUZez76VvnGTInuMRiKVk7vrF1wOC34hidjMDclvonqNxVpfSfb2H4hOnejTRnsg1SgJGH5yYck7K11yx3BJrmekIpEKzLo9jOx9bbvwaNzhCezlKC1DmdCwzd96+DD2oiJcJs8oRREbg9tqxXLoEI7KSiyZh1GnpXruLSvDVW/EVliEs7YOiVrdSLa9qMhHfmfD7RLZt7KCgdMjuOzpNAZfGMn4G+L5x0eDGHNNHHuWdc7E5P5RGkr3e/I06EJVLJu7ly1fZrHy7f0kjw5v17afWzsKuVKCLkjOm/snAPDwj0MJCFfSZ3wwD343hOTBvgo0cYA/zyw7h9n/l4rYhgvaEqmMsuzN1FflUVt+hOJDq3yux/aZSn1lHnXlWVQVZHBo/WfoQxMJiu7T6jbsZgNutwuXw4rNVO09jkWQSDmw7mNqy4+Qs+NHZAoNur8UUnTaRFxOG5kbPqe27DDVRfvQhybiH+77glX7hXFgzUfUFO+nriKHogMrfDSmRCrnwLqPqSne36iN9sBltlG9qm0DRnYUXXIm8P5TRfxvWSrf7u1LSa6N525t22mfIJE07O895jfpth4XzraF36t4NKP7cZndfWQjtCijM/Dlg3sYf308/aeGM/zSaMy1DkoPG/n+mQNs+r5zBjkT3SJOm+d773dRHJs+82x8qy+zogtWtmvbBfvrieqlIzBCycH11UjlElQ6KYYyG4YyG3tXVfkoAYlU4Pq56Tx17kYSB/hz31cD26wv1voK0sfdSknmWipyt2G3NkQ6FSRSdMFxFOxd7C0z15Zgt9TiH96DqsK2i7pqqin0mneOjt7lKj8A9GEp1JYfwe12HnOHiDYgmtqyhmi9NaWHWpydtNRGe7Bz1mvtJvtM0yWVQEmejRk9MtpFtuhwoB83Ft2QIQhSiXdNoCmsWVmUf/4lkXffhUStpmbRH5j37MV/wvhm7zlWds0fizHt8pgmDMtWEDbnWvRjx+A2mSh6tXP8J7NbXCx5N5sl7zYfdriz8d29W7jmo1HIlFKq84xs+DgTgLTJURTsat8F11WfFTLs4ghi03S8MWcXI2dFsvbr5pVlULQKXZAnflTOrlrKc9tuVrvzj5cAzws/JHYA8f0vYNvCZ46r1cK6ougbAVeuOrWQG3brMTkR/hr0CILkr38FJBIZ2oAob5XdS/6L47g1Aaet5XWUltropmW6pBK45aloPnymiAfeiGfIBD3/+3cBa39rm3j7glyOYclSDEuW+pQ3Z6u3FxVR8ubbTdY9KsNy8CCWgweRarVNygZwlJdT9Mrc03+AbsjdUsHb5y8lMFZL8b4Gu3HhrmoOLGnfVH+5u+uYekc8dqsbl8PNoGlh/Prf5hVodZEVk8EzCo7t7UdofNuHIxbdLirytiNTanzK6iqyCYxMo+iAZ/+Hxj8Shdqf2nKPp53dWo9KF+y9Rx+a1KR8t8uORKY4pb7Vlh9BFxSLqbbkjG0tl6gVDPrhAbZNewG/vnFEzBqBtlcUUo2S6jX7Kfl6PdaixoOF5MdmEjiql/fcklfJvts/PGF7QeN7EzQuDW1KJDK9GrfDSd3OXAybD1O1bE+j+uqEUCJmDsevfzxSjQJzTgVVy/ZQuXR3u3xHXVIJjJzqz/pFBkIi5Dx86WEeeju+zZRAZyQkeShhKSOQylXs+e0Vb7k2KIb4oTOQyJUU7vwdQ9EBAqLTiep7HvsXewLaJQybicNqomj3n6RP+ScKbSBOu9lbf9g1c9nyZUOyjN7T7mPfov+iDY4jYdhM5CodgkRKdf5u8rb+dMafPX1yFKNvSmHhUxmUHmx9PmZjpRVjpe+i4KEV7e/HbjU60ejlZCyrACCxv56CfZ5R6jUvpBHVU4taJ6Ou0s5PLx0hd3cdnz6wj6eXnsOhjdXsX9N2MZbi+19AXUU2DqsRhVpPeNIISjLXeq8X7F1M7/G3ow9LRq7QEtdvGsbqAqr/MgUZSg+SMuxKYtLPo7b8MFGpE5t8CRmr8glJGExQdB/sllqkcpWPKaclCvcvo99599Jr5PWUZW/EZbcQnjKS3J2/4LS3747v4El9SbzvAkDEXlkPAgRP7EPAOT05eN9nWPIrferXrD+Es9aMTK8mcHRqq9qQB2hJeugi77nb7kSiUhA4qheiw9lICYSeP4C4O6d6cqK73IhOF359YvHrE0vQ2DQOP/MdoqOxo8bp0CWVgNMpMnFmEL99Vkn+YStqnbSju9SuVGZtpTJrK1JFwyhRKlPSc8JNZG/4BpuxirTJd3Fg6TsYig4QN/hCAASpjKD4Aexd9F+kMiVFu5dQW3IIlV+It35zRKaPpyJrK+WZ6xGkMhTqjvEG2r+kmP1LiolI7bzeSMfz/EVbvJ/v7bfa+/nLR5veOJe9o5anz9vY5v3QBcUSljAUqVyFw2akpsQ3blJ9ZS77V39A2rhbcDtsVBftJW/3b17be0XudpTaIMKThhOdfi77VvyPHudc3aidgn1LkKv8SB56BYIgYKkvZ08rlYDNVM2eZW8S1/d8egy/GolMTkXudtyuts113BSJ911A1ap9FH64HIfBY37qP/9u5AFaoq4d28j9s3rVPu9i8JBFrcgfLAikPHkpLoudos9XU7P2II5qI4JUQugFgzDuK/Sp7tc/nvi7piI63eS9s4TKpbsRnW70AxNJfOAC9IMSib1xIvnvN7YknBaiKHb4gceK1+rj1Z97iO+vShPlCkGUygTx0829T+r+rnpIFWrvZ314stjvooe958mjrhbDU8eIgBjWc6QIiEHxA8ReE2/x1j9W1tH6w66Z61Pee9p9IiD6R/US+1/ymJg44nLRLyypw589ItW/w/vQffw9DolaIQ5Z9KiY9sb1Ip6Nqt4j7MLB4pBFj4oDv7u/RRlDFj0q9n73lhbrBI1JE4cselTUpcW0ql9pr18vDln0qBg+c3hjWePTxSGLHhUHL3zIp7wt3r9dcuXkX5cc5rbxB3DYRVxOkeuHn76r1tSUB9ugZzA0+nLCte3vP2ysKkAiU+IflYrKLwR9ZA9qiz2hnsszNxDddzJxgy8ic+XH3vr+UZ4p7LH1XQ4bfmEeW68+PBltUDQATpuZ3QtfJGfTApx2M4OveL7Jfoy8PpkH10z1jtQ1gQoUGhkPrJhM8shQAILitATH60CAu36dSGR6AAAPrpmKf6TaW0cdoCAkyY+HN5zfHl9ZN934UPD+Mo43b9Xv9uwvkWqVSNSnts5xlMjZowEwHig8QU0P2p6RAMTePJEhix71OZIeuhgAQdb2Vo8uaQ4CuGCOZ9fwrnX1FGbZTlD774fbaSdz1cckDJ2BVK4md/P3WOsa/PbD08ZQnrnRO7V3O+1E951E4ojLcdkt3vq5W34gefTViG4XtSWHqSncD0Bo8jAC4/oiut24HFZyN3/XbF9ytlR67fXmGjsJQ0OwGp1kbfDYxavzTQy/Oomqr4xs+iKboVcmsPDJXRTvM1Bb4vGGCUnUcc6cZARBQOUnP2MpIrs5e3GZ2/e9IdWegjuyKDa5KN2edEklMGSCnh79PZ4Os+4I593HC9m8tPWLhumhkwjVJCERZDjdNjYVfoUoukkMGEqMvh9yqZLCuj1kVq0lVJtMStBI/BShbC1aQI3Vo9VlEgU9g8d65azLn4fD7bsQOSbuJjLKfqXOVo5GHkh66CQ0cn9yDdvJr92JXKJkZOwc3KLLp83WYq4uYv+fbzd5TSKVU5HlG0itqbpVOdupytneqDx3yw/kbmkcC6Ypjk3eAiCKYtP5BYBdvxQw5pYeqPRytn/vyaalCVRw2dyhvH/ZKiy1Dv61ekqr2u2mm9OinccYbsvJhycRRdh724eNZijtSZdUAtc+GMk90zymj8h4JY++n3BSSkAuVbMu/xNcogOVTI/DbUUQJEgkMtbmf4xSpmVM3M0U1e+jwpRFhSmL0XE3+MjoFTIBmUThlXMsblyEapPZWrwAq7MeAYEBERext3wxFkctI+PmUGcrxWSvRi33Z/GRV3zaNNkbRgKCIPEkyHa78AtNxFC0vxVPKFBbfAhbfftn82qK4n0GFBoZKaPDOLKunKA4LVkbPLMUp81FxsIChs9OYu2HHv99hUYGiNRX2BhxbdNuiGcTPabfTuGmhViqOudmvG5ah6WgClVs8IkrHoMgEdAkh2M+UtpOvWpMl1wT0Pk32MVK8mxo/U7OTpZds9n74j6aJBwgz7ADAJvThNlRg0ra/I7DME2yj5xjCVbH0zfsfKxOj2ugWu6PnzKUQZEzGRV3AwISNPIgn3uaa1OlD6XfxY8yYOaTxAycdsJnC04YyMBZT5O3teMCWzksLubftZmxt/bkvmXnceUbw6jMadjss+XrHIZfk+g19xiKzGz9Jpc7fh6P0+aipqBho9Blc4dw24JxXPbqEG74bDThPfVn/HnONId/f7dLKQCJSk2vpzvH5sbOhGGTZ5CjSYloVf2jL/6Iy0a0W5+aokvOBCqKG6ZZg8b6UV50ktOuZqZaTvdxNsIWNlMKQvNylDIdBqvvj1gU3azOfY9jE87LJU3YDI9r01JbRsZP/9d8R46jKncnVbk7W13/dNnwadMhvEsP1DLvunVNXnNYXchVvop7yav7WPKqZ4F//bwj3vLvHtjWRj3tppszS/WqfUTMGEbP566kcN5KDBsycdZbECQCgWPTkOk1lC9s+P9d8MEyer4wm6AxaYhON6ULNmDJq0SmV6OMCCBgeA9El5vi+U3/rk6VLqkEHrnsCH2G6xCB/dtMPHr5kRPecyzxAYMI0SQgIMHutrCl6Jtm6/YNPx8/RSgaeSB9w8/H5jKxuXA+a/I+JjVkvFfOuoJPcLg8awKl9QcpMx1mVNz15NZso6h+L+vy5zEw8hL8VZGY7NVsL2mdvf3vRJ9p0Ux9uA+GIjNvTV/R0d0BQSBtxv3krfkWc2Uhfa56gpzln2Mqz0MdFEnylJtx2cyU7FyGIScDmUpLv2ufJW/Nt5jKc+l18b0c+P5l7CYDaTPu58CPc5Eq1aTN/Bc5yz/HaigntPdoSncuRa7Rk37ZIxz6+XUih5yPseQIFfv/2oOh8fcx3aXPeoj8dd9hLM1BECT0ueoJDv/xPtaaUiQyBVKlGoe5rtm+K4JCCJt+KfKAIGo2rcGwdT0SlZqEfzyA6HQiUamp3bmZyuWLAIi76W7k/gG4LBYqlv+OKXM/vZ5+jUNP3+/tU/yt95P3wWuoouMIm3oJgiCgjIzBZTKS87+XEV0ugkZOwH/QcCQqNVmvPtXmf67w1AACYrQcWtYwwHo441Je6t/+v6X4u6YSOs03rpM6PqTRfoFt017wfhadbvbd+TGhUwcQNDaNmOvHI9WqcNvsGDYepnK570ax+r0FbL/wJRShesIuHEzSQxejjA7CUWPCXmbAsCWLmrVtn6SpSyoBAInUE3ek10BPHtuM9fUnuKOBveV/NipbfOQVn/MNBZ4Y6lXmvCZlON22JuVsLVrg/bw+/1PvZ7PDwI4S3x23ruPaPdrm35W9i4rYu6jzmDmUuiDUQVEkT76pocw/FFN5HpbqEoq3/o46OBpDjm+cqpqsHbidDixVRWjD4hEqJKiDoug7+ykfOVZDORX7PAv9DnMdtrpK5Dp/qg5tJm70LDQhMVRlbsVY2nxYCVF0U3loMyGpIyjc+DOBSQOoytyC0i+46b5X5BN1+RxKF36Lo6aK+Nv+hbWkEHtlOfKAIA49fT8yPz2Jdz1CXcY2nHW1VK36E1PWIRRBIcTe+E8KPmna2QAgbOolVK9dhjFzP/p+g9H3GwyAIJUiyGTkvP0iMj89ipAw7JVtG2W2x4Qo7GbniSueALfF7vOyPhZLXkWT1/LeXkze24ubuOPEVCzeRcXi1ocvt1fUUThvJYXzVp5SeydLl1QC1zwQycBxfj6r+w9c3Hol0E03gCeIq9vF3q+fbTJCpUIXiEIb0Pg2b2Ayu1IqngAAIABJREFUwUfOnvm+wdmkCjUu+/Hx7AXqCg+y77sXCYjvTdTQ6ThMBnJWfNFsNysPbiJt5v0UbfmNkNQRVGVuabbv8sBglOFRRF/ZoBwUwaE+L2RnfR326kpken9kOj2mLI+Thb26EnN2JtqU5kMiSOQK3E7Pi1h0OhEkDWa9mi1rvfI1oeHYK8tR+sm5YcEkts8/wvAberH7pxzWvOUx+0WkBzLz9XOw1tlZ/cZestZ6bOJj7kwnIEZH7JAQBInAvEuXcsHzw4juH4TL4ab/zAQAPpqxFLfTzbA5Pek/MwGln8JHfmclXNuDMlPDjureoedhc5ooNu4nPWQSWnkgIm5ya7eTX7sLf2UkaSETAAG9MoyDVavIr207k2+XVAKTLg9qkw1i3Zzd2OqrsdZVEN5/AqW7lqMOjsJqKEd0OfGL7klwz2HY6qsJ6zuO8j0N4R+CegyhvuQI6uAoTOV52E0GrHUV3utH5TSHJjQWS1UxNdkZWGrK6HXxPS3202EyYCrLI6z3aGRqvxb7jgCiy0XW68/6ZKaTqI4LTCeCR4m1kDxGELzrXhKlZ/2qfMkvRF16LbayEgSZjPI/Gma3buuxCq9hccs/SoNcLeXTy5dx88LJ7P01H2O5hUvfGsk75/5OYJyOqz8Zx/wbV1OdZ2Tg5Ul8c+s6fn10C+oAJRaDne/uWMf0Z4dQcaSOLZ9lNjyXTIJMKeHDi5egC1Fx88LJnV4JlJuOoFOEYLRXIhXkROh6sb7gUwZFzGRvxZ/U2cqQS1SMjL2O/NpdpIVMIKtmMxXmLKJ0aRQb29Yk1CWVgKmubQModXOWIopk/zmPmHMuoc/sJ7HVVZK1+CPkukASxl1F1p8fYTcZSL3kPsyVhVhrSnE7HUiVGlKm3ELBuu+xGz1RSrP/nEef2U8iCFKstWVkLf6o2WZDeg3HP6EvuF24HDby13pMiPHjrkIdFIlSH0L8uNnkrvwSU7nHHFlxYD1Jk66nZMeSFvvuqKnGXl1B8KgJVK1djjIiqkWzjLW4AG1KKqYjB1EEhaBJ6knVmqW4bTbUcYlY8rLRJKSgCPJszlSGhFO3dyfli39uUBDHK5gm2D4/C5vRgSHfhF+4Gl2IClu9x7OuJt9I3uZyEkdFUJ13hLwtFZQf8gSEtBhOvKFr+3yPc4Kx0ooh33SC2g30u+c1dr9xf7PXk2beTsnahVgq2taEKSISq+/HgcoVROh6UW0pxOo0ejwII2Y0qi+VyHGLntmXm3Z493V03KDWxg76Oau/9/jhUD+f85+z+p/wfpW/UgxOCejwuCXdR9c9ZCqtOOiW1zq8Hx11qOOTxfib7xEFqdRbFj59Vov3KP3k4sMZl3rP53w9UUwYESbK1TLxjqXTREAMjNOJd624QAxO9BMB8fynBzcpa9LD/cXzHh3gU3as7KPyW/s8/e45+b9l+IgpYuigCaf9XQ6KmCEmBAwRx8bd7C0bHXu9mBTYEDfITxEqAmKc/0AxLaTp52qL92+XmQlcPXDvad2fOj2eoEQ9K57rdjk8XWLnPoNE69mxnfeP5mMuhd52Ha7aOqq/+bnN+xB+/+2UvfZuq+oGXXlJu/ThbEOm1SK6nDTMAFQowyNPSZbD4uSHf27gjiXTsBkd/Pl/O6jKaXldb+eCbC56aTi3/zkNa62NTy5f3qq20m95hswvXyH91mfI+eUj6nMP0GP2A4huF6GDJhDUZzhSpZrqfZsp3bAIfWI64cOnoAqJJPvHdzEV5yD3CyDm3MvRRCQgup0E9RlO5hcvI4puFAEhRE+4FGVAKBXbV1K1e/0J+5Rft4uB4ReRVdMQPXZH6U+kBk9kfPxtCIIEk72aLcXfIiAQ49eXcG1PBEFgfcFn2F1tGGa7o2cBpxJFVOffMBKRKwRRFyA94T0XvTlWnPjYkA4fTf0tDkEQJVqNGP3sIy3WC/3HHDFw1oXNXpdHR55yH8Lvv73VdVvqQ/fR+kOQSMSICy8XE+/+t5h458Ni4p0Pi8FjJ3V4v050JF58i+if0k9MvORWMWLkdFGQSsXU6x8T+93zmhg27DwREOVavdj79ud97ut5zUOiNirRpyx28lW+MwFBEHvMfkBUh3kihabe+ISoiYhvs74HqWPEEdFXixKh4R2XGDDU+/msmgkcy4vf9eCuyZ7Y6PpAGU9+kuQNI3E80+eOJnpQKCp/z8JWn0tTAHh76Lc+Acpih4XjF6Fh6M290YVrqC818/MdK6kr8tgYQ3oEMPreAUQNDCVrRSFrXt3hY6889rrb6SZnTTF/Pt72MeI7BaKI22RGdLewqAhUvPdZi9f9xo+k+qv29/Gu+f7Xdm/jbEB0uyn9dcGJKzZBTOq5BIT1xGqq4sj2BehDEonqMRa5yo/assPk7/e4WycPvBRtQDSGskPeMv/QZGLSJiGRKti/9gNcThvhiSMIiR1AfWWOt15zmEvy0MWkULlrLWFDJqIKicJSXojCP5iqjL9ceE112A2VLcppCoU+CHVIFAkXHuOqGxCKubRp1/KTRS7R4BZdHN1kKpMosRwT5aAt6JJKQOvXEO2iqsyBX0DzYSPWvLwDQSpwyTvjKdtfzcZ3PFmTjo9Q2f+qnsiUUta8sgOb0UHUwFDqSzxTLv9oHbPmnUvR9nJ+vXcN4x8ZwvS5o/n+pobp6LHX5RoZo+4e0JaP3CwqPznW+vZPwNFN5yckqh8xPcaj1Ueya81bmGrbN5XmyRAaN5idS17m2MiCKl0Iu5a9Rr8Jd1NRsBNLfTnZGZ4F58Hn/5v8/UsQBIGUIVew/Y+GUOYqbTChsQPZu+Y9eo+5BV1gLMaagmbbNpXmEjFiKsVrfyFixFQ04XGYS/Pw79Efl+14F96TR3S7ODivaTfj06XcfIQQTQJjYm/ELXoWhdcVfNKmbXTJ2EEb/6wlIU1NQqqK+/8bx/ZVzdsS68vM1BWbcDndOMwO6opN1BU39iCoLzbx8x2ryFlbTPHOCrbN24/o9iiKS94dT/bqIn69dy0FW8qYf+UfRA0IZdCcBn/qY69nrypi/pV/+Fw/qxAEJBo1kY/eQ8jN1zS6LAsMQBYciESpRBYc6D28SCSEP3A7sa/9B/25Y1GmJBI4YxrHhyaNePifqPumoeqVQsSDdxI797gk6n/1o6k+tOoxJALy4MbxoxKevNLnXNMrmrhHZgEQ98gsNL08ORl6vnuHtx9emdL2+8lVFu9m1+o3sVvbdqTYFuxaNheZQsPQ6U94y3wTwYsERqTSc9hspHIVMoUGQRD++u6O/bt7PgsSKYIgsH/9xxgNLcfrN+ZnUl+QiehyUbFzNVFjL6Eyo/nQC7GTr6LHVfejDAghdvJsNBHx3mvlW5cTmDaYHrM9KVnttVUc+uIl4i+4gbQbnyTp0juQyE8vD8GxiKKbfRVLWJP/EesKPmlzBQBddCbwxSsl/G9ZKoJEYMfqOj554fRHPOaa5l3RVP4KUqclkDotwadcF9rgGtfU9V3zmzZRnSqR6QFc+eZQrHUOlr1+gMNrygBPYpdBl8aj8pOz48d8VrzZ9lvLTwpRxG22IDqb3t3prPG4/7ltNpxVNY2ua/qlo+qRRPnbH2PZ6zH72Y7kNKpX/s483EaPQq/+xkzkY/c12Y9TRRbkh6Py5F6ogqThxZb3/AKSnruOip82Ur+tdekW2wNdQAxJfS9GqQ4gK+NHqssOMGzqE2xb8gJSuRJdQCw1ZQcZMP4e9m/6FIlURp+RtyCKboqz1lKS4zFrxqdORqUNxj8kCRDYsfK11uUBFgT6jP0HottNSdYGb7HTbqbv+LswlB7EUl+B024mJvVc4vtO985iRLeLrB3f0XvsbQiClIMbP8FqqqIsZxPpY24F4MD6j0+YjrJso2e3r+HQDgyHPIEij3cPPfy1JwhewZKvm5Vjqykn86tXfcrstVXk/vrxib+HTkqXVAIWk5sbzmlNSOW2wVbnIH9TKdvm+b5cLTUNU8nDS/NbvH66KDQyZr89jLkTlxAUp+WGz0bz6fXrMVXbkCmlvH3hCnShKu5aOJGMhQVU5RpPLLSTouqZhNti9SqA5jiqAIAmlcnp4qwxIgvU4az567sUALFxPbfFjsxfiyCVoEoI85bbS2rIe/E7er5zOweu87xgRLfoswmrvZHKFKQPv57MnQswlGcybOoT7Fn3HiZDEVr/SJTqAHQBMRgqDiOTqbHb6hg47h62LX0RmVzNwPH3YjQUUV+TT0TiOezd8CGHtn+NXKFtfSJ4UWTPqsb5rG2WWjI3f+k9d9hMTdYzlGViKMv0KSvP20Z5XrenX1vQJZXAqeC0OJGpTu1xC7aWEdEniMrDBq+J6HiCEvUtXj9donoHYDV6RtbV+SZyNleQMjqMjIUFbP7KM0o2VlipLjChD1d1aSUg0Wpx1Xa8SUN0uYl/ZBZuhwtBKiHvuW9xGRsrdmtBBY6qOmL/NRNr/l87hwWBpOevQ3S6qPp96zFCRVJeuxl7uYH8F79v92fQBcTicloxlHteorUVRwgM60VdTT5a/2i0+gg0+gi0+kiMtUWoNIFo9ZEMm/KYV4ZaF0J9TT61FUe8I3SHvfWbsrrp3Jw1SqB0bxWp0xNInhCD3eygYHNZq+/d+tE+rvxqCtNfHc2eH46g1MpJGBPFmld3Yq31mJG0oRrvdXu9A78IDfmby7zXT5eWsnXZjI5jK3psqV0Yt8WKVN98LoczSdbDnzYqy/3PcVFnRSh49adG9bIeatp+e+T+5ncTtz0iTcVEr6/OIziyN1r/aCQSKX6BsdTX5AECbtHFlj+fa3SPy9l26RjrKnOoq2xs4uvmzHPWKIHVL+9g9cs7mr2+9aPm443UFZv4YIJvkpbMJfk+58dfL9l9Cp1sgbxtVXz/0HbuW3YetnonK946wMEVpaj85G3bUCsQZFIkahWCRIJEq8FtsfrEqRGkUgS1CkEmQ6KQeeqYGpsOtIP7Y913CGeNAYlGjfWAx25e891ClAmxxM59BsPCP7EXFqPum0b9yvUnNUM42o+jfRCtNkRX1wo5kiz0JpE0NolLMNLys6cOvQaVNhiFSk/q0Gs4uOULaiuzObRtPv3G3IFSHUj2np+pKvH8X0/qcyFVJXsxGyvoOfByNv7+BG63kx0rXmXo5H8jSKRY6svZv+kTXK6TT5X4dyasV4A3tEVXRxDPYC7LZjshCGesEyPj5rAhv2X/9W6aJ+Diqfiff26j8ooPv8C83aP54t97pdH1qi+/x7jON+dx6C3XoErrCYKAo6yC0hff9F6TqFQEXDQF9cA+yPz1WA4epvLj+V5lcvyOYYlGTexr//HuYNafO5bAyy5sVT86MyejBLo5+xBF8bSn/d1KoJtuOjHdSqCblmgLJdAl9wmcDh2v8rrp5uwgXp5GuCz+xBU7GcPm9GTy4wO57bepnPtgf+5cPp24oaEAJI+NZM78icQMDPHWV/rJufXXKSSNjiAwTsedy6YTktR1cmF36jUBf1Uk6WGTUcl0lBoPcaB8GXKJknNT7mVx5kuAZ2R/qHI1ddYSov37EevfH7lURWHtbjIr1xCgiqZ3+BRkEsVfWcI8aiDGvx+JgcP/CtPqYk3O+4Rqk0kOHsmm/IYEHynBozlS1bY5Pbvp5mwgVtaLQmfH7Y84HTSBSj6+dCn/XHUBy1/KoPf0OPK3VpC1poSsNSU+SsBW7+Cn+zcx662RWOvsLHtpF5XZXWfW1qmVQGLgMIrqdlNYuxuV7MTeIlJBztrcj1DKdIxJuJmiur30j7yIQ5UrKa0/SIAqmuFxVwNQbS6gzHgYh8uCThEMQIUpi9TQiehVEdRZS5EIUgpr23iFtxPy8OfpfP1CHvkHut3+ToQGHSOFqdRRgxY/8sikUMymv3AORmo5KO5ERPSacZaJvm6gQ4QJbBNX+pRp8WOgMAYFKlw4sWKmRMwjn4YXqAQZqcIgwojCjp1C8QiFNJ+WsqNRCzo0kq4zGj6equw6nDYX5iobxgoLSl3LDhgVh2tZ/eZewnoFcGhp50mh2ho6tRLYVfILEkHK0JgrsTrryShZ2MicI5MovZ+zqj27EW1OI2Z7DSqZHoVMTWm9Z9ORwVqE2V6NQqpmdMKNrM/7FIfLgt1lQRAkiKKbDXmfMCbxFtbnfUL/iIvYVnRqAbPOBJf8Mxab2cUfH5/ejumXrjtzG+9OhXu/GMDr17Y+R+uZYIu4HB3+jBDOw18IYqu4kknCLGxYyaZ136eAwEhhKkpUrBAbu5geS6owkG3iSg6yAwVKRgnTiCCObeKqNngaSFEMJEnelw2WhRjdDV4vw1RT0Ur8WWn+lh6KQSTK+1DnrsIuWql2laIQVMTJU9li+YM6dzUAAdIwgiQRAOgkAV6TkMld65UtIGGIajJ20UKtuxKloCZKlgwIrDQf54LbAbgcx8UBOoHlPWFEGKNvT8dQZGLotT3Y+kXXmQF1aiXgr4qgzlbOvrI/GRF3LeBJ8O50e9zVgjRxaBSBLYnA4bIQ4ZdKaf1B9KoINIogpBIFIh5lARAXMJCsas/WeJfopKhuH/EBg9EpQ1qQ3DKDzwti+9Jq7/mcZ5KorbSzcWEl1z6VSHiCGrdTZMlnJSz/qpSkfjpmP5aAIBGIT9dSX+3gvjHbCY9X+dR/ePJOgiIV3PBsMikD/XA63Iy7PIx/T8vA7RIZe1kY026JRiKBn98uZMPPns1L598URXy6ll7D9AgSeHx6BkaDk6d+7EtsTw0vz9lP5nZPDKajMhQqCU672yvn1ZWDWPZ5CeOuCEerl7H6+3J+eC2/yec/G7DgmTnV4XmxuXAiE+StXnjSE4QabatG9AXiEdx/pYK0Y8NMPWq0p9bx08Tsrme3bY33PEHem2TFAHZaVwBgcDVkMjO6DZQ5G0fUjJX3IlAaxhLT596yfMdBRmkubseenz7Tnx1CaA9/lDo5xiorK1/djbHSyrRnh/DDPzdQV2phzvwJlB4wULCt4sQCOwGdWgnE+PcjXNcTu8vKvrLF3vL95UsYm3gbVeZcyo1HWpSRUfIrvcOn0CtkPJXmXCpMWVgctRQYdjA64SacbjtFdXt87sk3bGd0/M0U1J766HPnihqie2goOmxGoZYwbFowj1+QwX0fpDLvsSxy95rQ+sv4z8/9WP5VKbMfS+C394rYtbKGcy4KYeTFoQgSuPPNnj71k/vryMowMvfmA9z8UgpFmWafmcChLXXsWFqNPljO49/28SoBgEUfFfPeA4fxC5RhNHh2Hz8zcw/P/d7fp+9HZRgNTqKS1V45IdFKFGopj07dRUCoghf+HHBWK4Gj4X2dosPnvLVo0AFgFGtPWNeMb5BEFy4kNB89tz0pcfoqrTp3NcHSSASEVn8HEdJ4r1I7ikU0UurMRS3osIgdt+P92BzG71/gee9kr/dsLv39iaZDVfzvvEXez+9NW9xknc5Kp1YC+8qWsK9sSaPy4rp9FNe1nEz6WDfQdbmNgzsdrFjJwYqVjcoBXG4nEomMwtqMk+xxA26XyLjLw5j/XC7Dp4VwaGsdNWV2YntpuefdxtFFlWoJDpvnR+G0i0hlAiHRqkb1v5ubT1ZG8z+QyCQ1U2+MAgE0ehlSqYDrr7DZR23+9TVNB3Y7XoYg8YS4OSoHYOkXJQAYKuyU57VdbKQW6fQuXa3voKQJu0JrXpwuWv6bnUksou/akcVdj14ShEJQYRNbF7BPLfHDJjZegzK569BI9FhcXTfsSVejUyuBjiIhcChVphzMjtPbEXjOhSEseCWfcZeF8cs7nnC3LqebB8bvaJTP4OsX8vjHaz0ozDQjV0r48tkcBKH5+k3hFyjjzjd78uTFuzHVOnlz45BT6vdRGSXZFvTBch85lvqGHbcns8XkpY2jeHryZi5+IIkB54Wg1Er58cUs1n1b7CPniid7MHBKKCqdjKJDRhb+Nwe7te3jtLc3btHdpB1ZdYwJ56g5SSfoO62i8w333Fwdz4Oe7EyoqS/oqKRuzhzdSqAJ2sol9IOHjvDO1qEsfKeQPWs9CuXR8zP459u9iE/XIpUJlGRbePHafUSlqFGoJUQmqZFIBK59IpH/u3Jvo/oPTdqBzeJ5Kf7+fhF3vN6Dcy4K4cmLd1Nf42T5V6U8+Gk6y74ooTz/xCP1Z37qR3i8iptfSqG20sE3L+R6ZVhNLtb9WN4qOSdCFyjnn/P689rVO/nmac90+7+7xjD0wnBeu3onAP3ODUGukvLIqIZww6kjA+k9Nui02z/TFHCEWFLoJQzAIprRCXrCiMZGw0jZQCU7xbX0F0YhESSYRCMyZOgEf1w42SdubaGFtsXk9pikgqVR3sXbMGks/pIQHKJvzKBkeT8ybKu952HSOCpchdjFhv8njr8+q4Sm1y2y7LtIV57jU6YWdCQp+rHM9NXpP1A3reZvoQTCx/cgcfYQ9r28jPojTS/GjPjgKo58tIHKLW2T9q017Flj4Lb+viEKKgqsvHG7b4jkXkP1nHNRKHcN3YrT4RkFXfd0UrP1j1KSbeGJi3xdWL95MY9vXvQ846IPG9YKmvMgempGYxfYrAyjV8axcq7v6Zsu8+mZJ+c+u+rLQpz2hlF9WbaZkFiV9zx9dBDrvvHt58ENNdSWt13gsjOFEwc7xDUMFsYiFWTUYWCXuJ5QwTcxexVlbBWXM1AYQ5SgxIkDC0YKxKwz2t+jL/EeioGoBR1SQUa4NB6TuxaFoPKp6ycJZIhqMlWuYuSCEjcusuy+62dmdz1mdx0x8h64cWMVjRjdBqpdpQAUOY8QJoujv3Icte5KFIKKaFmKN3tWN2eOv4USKFt1mLJVh/FLCe3orpwSfkFynHY3R7PTqf2kxKZqOrZT7UBFrq+92GF1I1M0mBsCI5VUFja2KVcXdx4lYMbo9f134/LZB7BK/MWnrpFaVou++Y0NYuM8tvXUskb8rcn2ssR9ZNF4/ev4vQani1O0s826lJ6KQUTKEjG5a9lhW06YNJZoWYpP3d22NcTJ00iQ90EqSNlmXeJ1Dz2KiMhO2wp6KoYQLUtGKsjZa1vve926klHqiwiTxeIUHVS5SjjiaNoZI/bNRxDdbmoW/Ilpw6mv1XXTmL+FEmgVnSBGUnPsXFZN3zH+PL94AC6nCCJs/LVruJedDDZzy6M8QYB2SNPaTSsxumvYYV3uU2ZwlZNp3+5T5sLp80JvDpO7zus22hQibtZZfj6hHFmwP0WPvoHbZGmUYrS1KOKjkIUFYd6695TuP56Rz5/Hhn8vbVXdQfePYsdrJ/6+jmfCOxex8s6FJ33fydIplYB/WgS97hqLIBHwSwkl839ryf8pA33PMFLvHo8q3I8Dc1dQsanleOQBvSPp+8RUqrcX+CR7GfzKDDQxAYguN3nf76Tg592M+eYG8n/MIOaCPsj9lBT9vo/DH21oQXrb4XKJfPJ4y77iibNup2T1QqwVXWs34slQXWIjOEaFyeCbKjAwUtnMHd10DGc2X4U0OABn5l/myVMczGkGpyFaOyYctvv4jWedjE6pBNL/dS6Z/1tL1fZ8wsf1oGz1YYIGxJB2/0TWX+fZXNL3sSmoo/3J/6Hp6aMmyp/Bc2eyfKonXd3oL64DQB3ljy7xr4VGuYSka4biMttRR+gp+n0feQs8OQdGfHAVVTsKqN5R0M5P2zpyvn+3UVnYOVNw2+1Ubm9b00BH8e1/DvP2/nHsXl7JkW21BEYoGXZxOHWV3bHsz2aEJpSOPDwYl8mCVK8j8ql/kH/bfwAI/edVVLz1NYJMhjRIj7PcY6aShQTiKGx9Iqm2ZNdbG09cqQPplErg0Dtr6PfEVIw5VUgUUspWH0YUxdMagBwdQAjAmsvnNUoD2effU3CafG3PnSFDl19SOmEjpqAKiSTnh3cxF+Ug9wsgetLlqCMTEN1OAvsO58hnLyOKbhQBIURNvBSFfxCVO9ZQnbGekMETUAQEIzqdSBQK/BLTKfjjS0wFR0iefR8yjQ5BIqX2cAYlK1sOX9CeiG6Rtd8UM+C8ENLHBFF4wMiHd++j5/CADutTN50TeVQoweePBkFAolEhSCWILjfGVdsIuXkmxrU7sB7Kbbf23U43adcPImVGOgo/JbXZ1ex6ayOVGaXeOtO+uwpdtJ6CZVlsfHJZIxmR58TR59ah6BMDkSo8G//KdxSz6q6GdaQTtdEmiKLY4Qcex2DvETejv5h693gRoaFMqpKL476/SQwZFi8C4vifbhG1cYE+9/mlhHo/C1KJOHbBjSIg6nuGieetuFsMGRYvChJBTLx6aMM9ySGiRC4VJ6+6x0fWiA+uEoMHx4nH962jjh7XPSRqohN9ymKmXCWGDJ7QUCYIYso1D4jq8BhRqlKLvW56QlRHxIshgyeIcRdcL6bf9YLol5guBvYeLkafd4UIiEH9R/31fclEhX9whz9n99E5D//AhGav9R18oyiTq9utbVVP37alfloxft5/RHlUqCjV68SEL54XBanEe107rK8Y8cStYuidV3rLQm6dJfpPG9NmfZqx5Abx3A9niFGj4sWwwdHixPcuFi/5Y45vPQFR4acUz/nPpEb3+8X5i7NW3yKmzRkoKvxVYviwGDHtuoE+dSa8c9EJ22iL92+nnAkgQPT56YSPTQaJhI03fYW9xsyORxeSds940h+cxP65KzDl1wDQ/+lpaKIDkKrl2GvMHHh9JfVZlez+z2LGfH0DVdvzqfxr/UB0iwSkRzB2wY0IMimm/Gp2PtL+iy9nAoU+CFVIFHEX3eQtUwZ6PKas1aWoQqOxlBeiCo1CqvS4/YUMnoA6LAbD/q2YijpvVMr2RqLT4jY2H0U17F+3Yvj2N+wFpxes7+/Inu3zzmh7gkoBoojLUI9+yiifa8rEaExb9mAvKify6X94y90WK7LQluOMnQwSuYR1D/6BzeDZD7Gjzsrkz2b5VhLBXt+0Z1vEiDjcDhcHPt8JIpRtKaTnFf2AnT4o/KufAAAgAElEQVT1TthGG9DplEBg/2giz0tl5UXv43Z4vEkSrhhE7rc7qD9cwZa7vmt0T8bTixqVARj2FrP2qk8ale987NdGZUvGv+FzvunWr0+l+x2LAKLbReZHzyIe42YTMniCN7eu6HY1VAYOf/Yi+uTehI+ajsNooGDRF8dLPWl6PjKd0HPTAXDWWdh86dunLbOjKX/1g0ZlirgoZKHBmLfvaeKOrs2AYbcBUFV+gIJcT7C49P6zUaoDqanMJPeIxzMmLnE8MQljyNj6ASajx+YeGTOM8KiBGKqzvfUCgpKITz4XiVTB7q0folIHEpc0AanMMxjZu/MzWrvo66yooX7ZZqJeuIf6JRu9dn8A3fghhN57DaLVRtXHDabN+uWbCb3zSmJefxi30Uzx42+d1vdTc7DS+3IGMJeeXJgLmUaO2+n2jOe9NH7+02mj1X1pF6mngWFvCfVZlYz89BpEpxtRhA3Xn/6LqasSM+UqlCGRKAJCiJk8m4I/vsRS6vGUqNi6nNhp1xGQNogjX87Fbqji8GcvEXfhDagj4rDVlJH380ctyu918xOIbhduu42qjLM7eY5++kR0IweDTErtT39i2rQTdb80/C+ahDwmgvK5H2I7nAtA2L03okiOB6cL/xlTACh5/NUO7P3Jcfct/hQWO/nx98azn8iYYeza8j4AiT0mExkzDLOpnP0Z8wHo1ecygkNTqao4SH7OKvyDkrz3BoemEZ88CbOpHH1APAFByVjMlST3ms72jQ0vXpOxjIN7FvgMVprDmpnbqKz660VUf+0Z/NX+3hDRtOqTX4BfGtV3FFdQ/Njpvfh9+lRj9jk/2TS9Bz7dgdNk5/INt+F2uLFUmPj9svkt3tNeqYA7nRIQXW72v7r8xBXPEgr/bH5GYqv+//bOOzyqYm3gv+27qZveQ0ISCEVAQFBBUBRFvXZFUARFVNQr9oLXD2yIBfsFrFwRVERQEREVFaQISJFOCOm9bzbZ3s73x8KGJbsJgQCBnN/z5CE7Z/Y9c8LuvDPztipyFnpPPDZ9LYU/eCfMO+w9VL3ZvSozFGRhKHBHIWd9OL09h3tG42o0UvafN1DExxDz7INYs/Mx79qPedd+4l58zKtv1TvziJg4GntpBQ2/rPUjseMy+c5QPlnou/qV0VDB4Z1icGgStdVZSCRSd80NBAKDYynOX+PzvSZjFVZLPbu2foJEIkPAhQQpCuWR6SMkgICAcCg45ORMbh2dlCu7sfuDv91HQqeRDqcEREROF8bN7i+jvawSe3EZyrQuOOpOLIlgR6RLkpyMrv4rZTXUF3mOg+pqsmmoLyI0LIUefcai0mipq87CZPQdzGg21VJWvJk+AychkUjYvf0zXE47B/Yspe959yCRSNnzz+c47GYQBPpf8BAWs459OxaelGc9nUjlUhRBSmQqGcpQNQ6jzX0EBMiUMrTdIsn+evdhnXjaEJWAiMghJBLJEd/F9vlmTnsinFuvDyI1WY5C0eRy/PSLtbw1161gXp8ewVMv1HLfhBAemqQlJVlOvd5FYp8CL1mTxoVw34QQUpMVbNxq4cVZdWz5p8nw+K/LAxl3SxAD+qiIj5OzbqOFl9+qY/3mpnPlf/5IIjPDrQBemxbBa9PcpVVnvqtj2qtNZ+uHj4MOo9cVoNd5j+cwCoUGp7MplqOybBuVZd5Rxrrag+hqm1fb2r6x/Y5oOhLdxvSh35SmBHnXr0wBYOura8n7YT8uh4vyv4oYPH0Eg6ePAKBoVQ7bXl+H3Xhq42JEJSAicoiAC/pjWL0RRXwMiqQ4bLktF8xxmS3IIlvOcPr0FC0vztLx93YLg/qrmf5EGIOuKCE33zsq+vmnwrntxiCWLDdSVeOkb2+l1/X3Xonk/rtCWbHKxIpVJu4YHcyaZQmMGl3Ouk3ufEuPTg4lKkLGsp+NFJU4mP5kOMsWxNHv4mKKy9z1CF5+S0f3dAUvPB3Ogm8aWfGr+2x7/8G2Tzza8DS6pI2gQV+Mxaxr8/vPZI5OGWE32Fh8YZPizF60i+xF/hMsDnvrKpw2J6vv/wG7yYZEKmHo66PoMeFcds1xJ508OmXE0fdoL84qJRB7378I7NuVxq3ZSNVKggZkkHP3myjjwom552qUceEILhd5D76PLFBNypuT0f20Ge3IATRu2k/1F78jC1STseAZsm58HoCUN+6lesFvGHd1XvfJ9mLI+zdQvaUYQRBIuaE3lhoje/+7gbpd5Z4+UeclkTlpMKHpkThMNsrW5LJv7l84TPZmMrqNH0hDfm0zGceLLDCA+BlPgkyGbuF3OOrqiZg4GkVCLPLoCCImjsapb0S3eAW2vCIMqzcScd/tJLw+FafRRMUL3h5mYaFSFi5u5LX33BPk6vVmunVVUFXtxGzx3mWMGhFA34uLm7UDDBmk5v67Qnn7g3qeeqEWgDnz9GRtTObtlyMYeJm7VsV1d1RgMB6RpbXaycK5MVx2cQD/+9J9/r/0RwNDBh3yyNlvY+mPx+9xUl+XS33dqc12erYQMyiRXycsof5graetobCegOigUz6Ws0oJOBqM5D7wHmlzplD4n//RuHEfipgwuv73IRw6d3k+CTJCh/fFsPUAimgt9b/9Q90PG0mZdR+m3flYctvuBx6YFk36Y1egiglFHqRCImu5EMeGkW+0eF0RGkDkiEwSbhqIPDQAiQRsOhOm/GqyZ67AaW7bqq295QX3SiD6sp6E9k1GGRWMrdZA454SqlbtRb+z2OPa64uM8QNYddN8Di7YRmhGJMPn3cpfDy+jZrt7Ius2YSDr7mvKzBnSNYKrfrmXHy6afcwyjgeXwYj+x9/R/+jtlFA7b7Hf99grqql44R2/10NCpBhM3pO60eTC4Wg+0U98uMqnAgB47jH3buPRyVoenewdPd23l4oe3ZTsz7YRGyPj/jvDufA8NQlxcqIj3VGoqcln1dccgLBRVxB8/mDqfvgRw/btp3s4beb3e7+n778vIKxbBIpgFXaDjS0z/6T0z/xTPpaz6tMhOI6oenXILx6JBMHhJOfed8DVtEqSBbpXQy7TofPSQx4KHkeFQ14L0gD/yctkGiVdp1xG9GW92ucBJBKSbjufhNGDkAV4HweoY0NRx4Yy4PN7KPx0LZU/H4NvejvLk6oUpD96uScG4DCahDA0CWFEX3EONauzsNY0+pEA+gPV2PTuv7mpwt1PHeFOmy1VyKja7H0E05BXi7nacMwyOhKFxQ7uvj2YvQdsbN9ppX9fFRPGhDDl2ebppK02//YH6aE1xcP/qfGpQCoqHQweoObXb+LR6Z3M/lTP7n025HIJ382Pbbfn6SjItVpCLh5O8Ysv4TI32TuUCQkoIiMw7mxbnYvTQe2eSv582Hf68FPNWaUEfGGv1GErryPihiHULl0HgETh/7FdZisus5WAHsmY9hWijPV95iuRS+nx4g2E9kt236fehH5XMfZ6E/IgNSG9ElDFhHj663cVY8yuRLfVt6aXquR0f+5aws9P82oXXII7e+6hPEYKbQDpj48isGs0eXN/92u7bG95AD1fuYnQPkne8hwuJPKmnU/kJZkUzfefNtdaf0S9gCMVLu4AmsOT+5HY6r3bWpLR0cgvdDD71UicLigucfD2B233NsrJtzPiIg1rNpjZd8D3ru3he0MJ0Ei46Jpydu119xl2gcZn3zPdI1MeFoazoQGXybv2ROA5vXFZO07tiTOFs14JIAiUzPyKmImjSPvoUSRyGXkPvNdi/4qPVhD3yI3gdGHYmu2zW9z1/T0KAGDrbR94H4NIJAz59QnPS0uJjvwPV/u9bdrDl3tN2JayenY//hW2WgMgQRUdTNigrqRNGem+/w39sTeYKF7oO0Nhe8sDvBSAXW+i6LMNVPy4A4lcSkByJFGX9STu+v4kjhnsV0ZLM5DDaEMZqm7WrtQe1XYGzWJdUxT0u6SYAzn21jv74culjdw7PoRnHwnjjgcqvR4/KFCKwejyeB6VljV9BieMCfYpr6bO3Sc54dR+/cOuHIU8MgJNWhpIJJS8PgtFeBjhN9yAPCwMl9mE7scVmPbtR5WU6NVe+vospCoVMZPuRh4WhkwbSuLUpymf8wESiYTIW29BlZKC4HAQfP5gSl57A1wujxypWuWRLQsMJO6RKeByIZHL0a38GcPWba0/wFnKWaUEar9xB+3kHprkjf/kAO7dQMnM5kFXh42/AAVPNqUFaPhzFw1/tryljL9hgOd3/S4f5+CCQNWve4i+vDcAUZf1JP/D1ThNzVdyERd1I3pk05GSfkcR+577Fpf18MQhYK1soGL5DqIv60Vwz3gAkscPoX5rAY1ZzY2i7S1PHd+Ud8VWZ2TXlC+wVrrr0goOF8a8KowfVaH7O4/eb9zq56/WMi6Hi5gLunBwQdMXMqRrBJqoU28say9qdU4MhhNTWhv+tvDmnHoef0BLWoqCr783kBAnY9gFGnbstnLfE9Ws/N3E9VcG8tVHMSz8ppHAQCnd0xQ4fZhncgvsHMyzM+mOEGx2gbp6F5u3WVi9vnlVt/YkeMiFVMz9kOoFXyALDMRlNBLz1BNUf7kI84EDKCIjiZvyb0pmzCRm0t1e7YqoKOzV1ZTPnoO6ayqRt46mZOZrHtkVH35M1G1jsJVXoF+9BsCjNKq/XISjtpa4Kf+m/P3ZuEwmFJGR5D/6OIrYGOIfeRhLbh4OXefycDrMWaUEThXa/l1QRTcd9RR+/KfPfgWfrPUoAalSTuTwTCpXNlcumdOu8/xe+s0WCj5a4/feux7+grSHRxL7r34gkdDn/XHNDM2BXZvKbLaHPICeL98IuHMBbR07t1kq7sPod7TsVtka+z/axEVzbyK0exQOk53ydXmsvKrl1BcdlVEjAtiw2cKN1wRitQpIpRIiwqQ8OllL9/OLqNW5Z+jDHj8t8cxLtTzzkv9+875oYN4X3hHAc/+n99nX6YSeQ07s/+l4MO3eg63UXRTJaXSnq3BZLJgPHADAXlODOfsgysTEZu0hwy7CXt22antHyzFnH0TTIxPjtiZDsr2iEltpGaouXTqUEpAqVaQ/N5PsaY+13vkEEZXAcSBRyLxeuxy+vWGEo9pb8xoCqNuY02qf2g057knbDyFHHNu0hzxZgBJNots2ot9R5FcBHMauM6IIC2zWvuEh71oFdoPVy+sHoHZHGevuX+pX9rHI6Ch89FY0yf0KmrU//1Q4fXsr+WPdyV15nwlIVSo0mZmYs7JQREai6ZZBzeJvmrWbsw60KstltiAPb7LhWYuLPXIcNTVoumVQ/6u3f78iNgZlQjzWwsJ2f7YzhdZnJZFmWCu9V1yqqBCf/Y40DAPY6lr3yTbmtb7aMeVVtXj9yJ1Ae8jTJIZ7CvqYilpftZpLOs6K6mgiEvvS59JHCNTG++3T59JH0MZmntB95HIJsdEyv9era1uut9xZqPj4U7SXX0bS9GlE330XNd8sQbDZmrXbq1r+jAI0bPgLdddUEp5wr56PlHNY9mE5gt1O4rNTib1nErVLlnaoXYCbU2f3EncCx4G5RIddb0IR6nZLjBia4XPFHTG0W9MLQaBhT8v1gQWHE6exde8G21EZDI9GHqJpX3lBTW6y9vrWV6+OxuYePh2F2pKd1JbsbFEJtAcOh8DyX4zMfSOKjVstOB0QHSXj8osD+H6lkd37Ol/JzJqvm8dc2EpLKX+veapxf+0Alrx8L3vAYexVVZS+8eYxyZEoFJS8MvNYh94q6c++Qtmi/xFz7WjkoVoceh0l8z/ArnOn4lDFxBN1xTVouqQiOF3kvzMDp8k7g2vKw1NRhIZhq66g+ucfTpkDhKgEjgPB4aRi+U6Sxrlzg0SP7M3BN1Y265cwepDn96rf9+FoaHkCdVocxzgAAZfNgVTp+79PplGeFHnAEcZl/zgtx+8J0zFony/f+Aer2LAigdHXBRGgkaDTu9ibZWPsvaen1q3IyUV7/jCqVn6Hy2JB0yUVe717d6EIiyBp0r8xF+RS+sWnSJUq4sfeRfGnTcpJFROPtbSYqh++QR6qJeb6Mads3GecEpCpNGQ+NIO9s06+waQliuavp2L5P2Q8dRXaASkMWfWkz376XcVkv7oCW7X/AKrDyINUyINUOAwtr96VEUF+J2xwu4PSv0u7ybPWNh1jKSNb99RRhPr2Tz8ZpA24BZlcRc7Wr3E53cpHJldx7qinydmyiPrKbM7713T2/DkHc6PvozGJREr/K6ey7acZBEd0IUib2C5jM5pc9LukuE3viYmT89Vv8Vx3YQnGRteR8Y0i7YTTaCT/0cfbXa6jvg7jgX0AmAub0swkTpiMIWsvFUu/8LTFj72LsKGXoFvvdhuPvXEshXObdjGm/By6Pj6t3cfoizNOCXQkbHVGct9dxTlvjUWh1eCyOZFpFDjNdmw1jeTN+YP6bQVtkhmYFo1+Z8sTR2BaVIvXDQcrgL7tJs9yxBm/Jjmixb4A6vjWC8NrLxyGVKWibvWqVvtqUtMx5/s2cIfF9WD/uk88CgAgKCwRp91CfaU7xqO+6iDamO5+lYAqMAyFyq3cGmsLMRuaR/SeKmISZNRUOWnUn77Zv++4F5GrAqnO2kjRhiWtv0EEAIfRt81PFhBASN8BhPQd4NUuDw5t6hPovbhyWU6d04CoBE6AwK7R9H5zDPIgFdvvnof5GIymrRF+QXqrk3b4BRktXj/STbM95DkMFixlOtTxYWj7JbdaCEQd27oSqP9rLeGXjGy1H0DEyCsp+ch3ymEJEndxkiMQDl05ftrvLPaJFyMYMkKDQinB2Ohi0o3lXDIqgHGTQ1GrJdht8Mm79fz5i4lZn0YTmyAnKlbG138k8O/bKqiucJKYouDJl8JJSlHgdAjccknLtqUTZefC6fS+ZepJvUdnwmk2Y8w5QN1a77xUTpPB5+8AMk1z77qTRYdWAlKlipiLriakWx+cVgv5X74LgrtObvrEZ5CpNeh2b6ZqnbvMXJeb70MZFgUuJ7Xb1lK3YwNh5wwmctAIJHIFgtNJ9V+/0Ji/n663P4xu1ya0vQchlcnJ/vjltg1OAhlPX+kxmp6IAqj/pxDtuV0AiL7iHEoX/42tznfRc2VUMFGX9vC89uWXbylrSk3QHvIA6v7KJf7mgSjCAom4qBu1a1tw2TvO+Tf8kpEEpHfDmLUP3brVKGPiCB9+Ker4ROLHT6JswafNlI+uIovE7iPI2fY1LqcDqVSOUVeMTK5EG5tJfUUW2ugMSvb/5ve+VqMOu9X99wkKS0QdFHl8D+ADbbiUsSNLsZgFYuPlNOpdbN9s5c9fy9HrXKSkK/j42zhWfmvggTEV9D1PzbOvRXDrCPdEL5XCK3OieOXpGrJ22wgOldKrn4q9O05megQBQRC9l9oLU/5B1AnJWCvL/C6erBVlXosrdVKXUza+Dq0EYi++DqlSzYG5z6MICcNpMSNTaZBIZeTMexV5YAgZdz+Dfu9WrLpqqtatwFxZgkytIW38E5grSzAW59KQswen2YgqIobU26bQ+Ml+lNpInGYjOfNeRRURgyIkDHvDsbuJqaJDCOwaDbhLYp4IxQv+QtuvC0jcdoHuz13LvmeX+DSwZv7nGi9DbcmXm3wLFWhXeRUrdxF/80AAut4/AsP+Mqw+7BwBKcc3gWq6pKJJTqXkkzkkjL8Hc0EeluJCKpd8iebx/1D2ue+Asfydy0jpcw3nXvEMEomUHatm4bCZ2L9hHl3PvYH0AbeQu30p5ka3a2D38+9AHRSJTK7CZmkk759vMenLyd68kP5XPou+6iC6iqzjegZfzJ+tx2J2f7ErDuX0T0lTcNs9EUikgADBIf49teMS5aRnKnnj4xhP25zXdSdZCYi0J3VrVpF8/2PEj70L/ZaNOC1mtIMupHrlMo+HkO6vtcTeeBv6bZuQh4QSMXwkp8og1KGVQHB6Lwq/cRdR8DVBO4wNWHU1yIO1CC4nlip3GminxUz2Ry8B0POR18n9fBZOsxGn2YhM1WS01B/YAYC1tpLQ7v3Qt0EJWCsbaNhdQsg5iUhkUp+GYafZhl1nwpBTSc2aLGrXH/S5EmjYXcJfo2bRdcpIYq/uS8g5iZy//BEAT2CWROq9vK5evZ+Dr/+E4PD9QWlveeaiWhr3lxHcIx5lZBADv5zsludwuYPgjhB38I2VZDx5pU85/hBcriYZEoknIZwgCEgU/kshOu0Wcrd906zdWF/K7tXNXQMPbFrgU05jbQHbV77SpjEfC+MmhzLjqRpsVoGoWBn1dS5mzI7i6vOKMTS6mPSIlnMHN8+XdJjyEgf5B238uszI/DnuCGCF8vQlywuOTydj1H3UZm+hLncbcnUQ0T2Hkv3TXE/R+G5XP4BUpqB8xypcDhvx/Ueh1kazc6Hb0KkOjaLnjU9Stv0XTDXFpI28m/J/fqVi55lbWzznlWf9XrPX15E78zmvtoriAq/X1opSL8Nx465Tlx67QweLGfKyiBzkLr2mCApFIvOvs2z6OiIGXeJ5rY6KRyKTIyDgMLqDu8L6DfF6j7ane2WriojBVN72iMGsl37wmWfnMDKNEnW8lshh3cmcdh193rnNb1/BJZD7zq/kvrsKu77JKCSRSrwmbIfBQv4Hq8l+5Ue/E/bJkAew95lvqF3nnVBPIvdWAJU/78aU33KAmlSpInb07QT37U/s6HHItWFYigsx5+eSOOlBzPk5WIoKDj2IQOPuHcTddmeLMjsqVrOLxasT+HFzEu98FoMmQMLSBY18+Ws8i35LwGoRKC3078rrcsFT91TR+1wVyzYmsmJLEvLTuHSL7z8KY2UBhesX01ieiy5/J0GxXdGm9Pbql/PrJ+iL9tFYlkPxxu+Qq5rOuEMSM3E5HVTs/IOG0mwaSrIIiu16QuMKCJHzQdYwv9cf/awPST1a92xrScbZSofeCVSs/o6Yi6+l++TncViMFCya7d8gKQgExKfQ7b5pSKQyrHWVFH37CbodG0i78ylcNiv1e7dgq3ef3bscdmTqADLunopEKmuzTSBiaDcSxwwmqPux52sP7hnfqlG14scdVP+xj6hLe5I4ZjCK0AAEQcChN2HMr+Hgaytadfk8WfKcJhtZLy4jtE8SUZf1JOScJJSRQdhqDTTsLqHqlz007ClBqlJ4jqN84bJZqVj8RbP2uj9/p+7P5qvB6uXfHvPzdjReeaa5rei9GXW8N6Opnu+CD5py/OzcYvHYAw5TWuTgyXtaj5g92UikMgKju1C27WevdptRT3B8Brr8prxYDkuTDcpqqPPqL1OoEFxODhvgBZcTidR/dHV78PadHb/GwOmiQysBp9VC2S/NowyPjBHIW/CW5/ei7z5t1rdizQ9UrGmq1Vnz9x/INIFI5QqqN62ielPrLopH40m4BjiNVkqXbqV61V5cR9kGlBFBBKZEEntNP4K6uZVFxIXp1G5oXnD7SJwmGxXLd1CxfEebx3Yq5Ol3FaPf5d/jyGW1s+HylquniZx5yJRqJBIpDqu3k4HDYkCuamGVfdSaR1+cRXz/UURlXoiptpiQxExK/15+EkYscix0aCXQUfEoALONnQ8txFxc57OfrboRQ1Y51auzGLjgHhRhgWgHprSqBEREOiJOmwXB5fI62gGQq4MwVh/7caqptoTiTd+TPOQmrI11VOz8neosPw4ObRmfQ2DkxERG3p3IhiUVLHu7AIBzLg7nvvd68s5du8nZ1rTrempRPyQSCUk9gzDU2fj5o2KPjKG3xKEJkXnkRCVrGDs9neguGlxOgWlXbAFg5MRE1EFyopLVdDsvlJeu246x/syKmO9wNgGpRkPK+7NO9zCOiapf9vhVAEfistpp2Oc2WivDWz6XjLpuEN3eu4eYMUPbZYztLU+k8yK4nBgq8whN6uHVrgwMpbG8bQXnIzLOo3TrT+xZPOOQQfjEYzNkcgkKlZQZ12/nktsTiEl1O4HsXlNHVVHzfFYrPyjitVv/4fNnD1B20MSaL8o8MqZfucUjRyKFe9/twfdv5fPcZX/z6i3/kNq3qWDP8LFxrPq0hGeGbz7jFACcRTsB7VWXU//Tr8fU12k2tkvaiSP98VtCFqD0xAEYclrOG1O97G9cNifyI1IvxIwZSnA/t+Es55nPAQg6pwuxtw93J6N6aTEui43EB69ClRCOVK3k4GPz/MoTOXb+82IoM6b5zst/NpIw8Cpi+14KgDo0mqhMd36svD8+R5e/k+yf5hIcm0bmNQ8REJGI025lx8LncFrbEOEqkWA36UkYeBUJA6/CYTXSUJJN0YYlOO0nlnzwp7nuOJeqQjPhcWoq8/2Pa/yM7pRmG1GopHz9clNE+mEZ+mobVYVmIhM1xKUFUJLlPgYzNTjI39nkHr3z91qK97eeIbij0iGVgOB0EnrpxQRdOBipRk3xsy8AoExKJOKW65GFhVG7aAnmvfuRh2mJGHMzqtQUAgecC0DpjDdOiY9t9OW9KV/+T4teNTKNku7PXuMp9H60d01rBPZMIrBHEjlT3ZN/QPcEzAfLSHr0WnKe/Ax7bdOHsfTDXxAcTrf3j4RTmY32rMVu9/9H1IZJiY6RkZ115q3+/FG69SdKt/7k1Tbnqxje2ajjsAN1Y0UuWct9R3ADZK+Y4/XaaTOz7dPDuXokZFxxL4LTzoEVs3HZbSgCguky9BZi+11K6ZYV7fg0LbN1ZRWLZ+S2mqyzpsRMVaGZy+9O5OeP3LYwuVKKw3Z2JHbqkEpAIpMhkcspfek1ZCEhKKKjcOobiJk8kZoFizBnZZM0YzoV787BXlVN5dxPiBw3hpqFi5rJksvVDBs6jT/WNPnxhoelU6fLITnpIuRyFRp1BGFhXdm85V3s9pbTKh9JYHo0vWeNoeiz9c0ibSVSCQmjBxF77bmoopq2jqaCtuWlUSdFYjpY7pnQNSnR2GsbcTaavRSARCkn4d7LkaqVCHYHJf/96biC2KQaJeFDM6lZJXpTAKxvAxcAAA8XSURBVLz+UoPfa9fcoKFrhpwXpp7dO4UHxrZf1lN1aBQhCd3Y//1bmGoPeUHVgqW+CmVg6+lGjocJr3YnOlnNhJndaKixs+S1XPJ3NnLhjbGcOzISqUxCdbGFD/+9z+f7BRfMeWAvo6emMXP1YGQKCf83cgvHmKO3w9MhlQBAw9r1ADgbGlDEZSALDcFltmDOcq+kLdkH0fTojr2qbSXnjiYhfjD/7JxHds6Px6wA9LuKPUXXQ3ol0PuNWzGX6nCZ7UjkUmQaJcrIoGaVxHR/57d5fObCKkKHZHrcLS2F1TjqDchCNMjDgnDoDCCB4D4pyIM1FMxciixYPPo5FVw4TEVF+clJrxCS2ZeEa8eRP/8dLJUnN1fQqcRmrMdptxJzzsVU7lmLVCZHm3IOwfHp5P3++XHLNTU4mJy51vP6lZuagq3mP3OA+c94pznJOC+Uxwf/hePQTm/s9HQGXxftJeNIOTXFFuY8sLfZfVfNKznuMXcUOpxh+DAuc9PZoAQO+Z23f6SkTpeHwVCO3e47t44v9k1dQsXyHV5lFjUJYQSmRxOQEokqJsRLAdjqjOS9/xv7/89/2UQAqVpJ8hPXE3FVf7TDe6OMDsWUVYphdxHpM8eT/up4jFklCA4XJe+tIGXqTaTNGIcsUI0puwx5eDCpz48h/q5L/crrDGzaG8v5Q1TcMDqAn9dFs3JtNIlJTX7o4RFSflwdzbbsONZsjWHW7DCiY7z91H9eF83eonhmzQ7zeY8NO2MZNkLN6NsD2VsUz96ieI6MZQyPkPL0tJAW79ESDVk7cTT634WcKOqYBEK692mxz9BLNcxfHsf6g8n0O8+dI+viKwL49Dvv2Jh7HnF/rpJS5Ly/IJrv1ydw8/jgZvIAXA4bB3/+CLkmiIxR95A2ciIB4Qnk/jYfXcGp230GhSk8J8aaYDmJ3YOoK++cqTgkwimqXtPiICQSzyCkGg3Jr79EwUNPeK5rumdgPnAQZWICEbfcgCw8jLpvvsO0a4+njyImmpgH78VlNFH22ltHSmf4RdPYuWs+9foCwrRdkUiknuOggIAosg6cucFIJ0rYkO4k3nUxquhQzIXVFH38O6oYrddxUEBqNImTRhDUPZ76v3Mo+mCVV4EcqUpB/O1DCb8oE2VkMIasMvJe/wFbddMk1vOdCWhSo3GZbdStz6Lk09U4zU3VtfoveYycl78lZcoolFEh2KoaOPCfRVgrjs34fiSb9sZSW+3i2suqcB7Dnn3Kk8HcflcQg3s2j/6eNTuMJx70nU5k2W/RbN9iPabjIH/30PYdTOTgESCVUr3+F/R7tnquZdz/fxR/O8+zE4gYfAlKbQSCw4FUqSQorSelyxdiLMwh5Y4pKEK0uCxmKteswJC7D01cMrFX3IQ8IAiJVEbDgZ3Ubl5N3KjRBCSkIDgdOC0mcj5+3X3m4YdFq+J59dladmxxT5L3PqYlKFjKWy/Ucc3oIJYvNpCQLGfpnwnUVjXtjGa/puOnb499cSXSdgRBOOGVcYc7DnKZzV4KAMB8wO1Xbysppfxt3yXn7JVVlEzzFfUrcCB7GT173orgcqLT5VBVvcdHv5OLJi0Nc27b3OhOBen/uYHaNfsoeG8lqqhQUh+9mrIvN3iuq2K1ZM4aR+PuIvLeXE7SxEtI/7+byHpyIeC2fXR7+VYCUqMo/WIDxuwytIMzsNU02StC+nbBlF9N8bzVKEIDSLzzYjQvRnlkHCbmuoEUffgbTqOVoF6JWCuP/6x9/Z/WY1IAACVFToKCTm4+nsP3kMnxGpepKJf87D3IA4JIuWOKlxLwhSwgiKDU7pQsW4CptJDQngMwlxdTs/4XDPkHUIZFkjLuIQq++C8Rgy+hftff6LZvQCKTowgOxd5QT9Hij4i/eizWmgpqN69u87N8/1UjC1bE8d+ZOq4b41YCEom7pOY1F5bg6mAJSM/treKfPR17lR90QW9iptxMydQPsRb4T0VzMuhwSuBkUFG5g4rK5tGyRcXrTtkYwq64EvMc3wrsdGKraSTvjeXuHD2Ay+H0qjIWd+uFOM02cmZ8h2B34tCb6fHmHYScm0LDPwVoB2cQ3DuJ7GmL0W9xKznDXu9z0oTxw9j/eFPiNnNxLb3nTmo2FnVSBPWb3a56jXvaVpHraHS1LXhsyWHeV5EkJcvQhktRKNpfAcjkcOu4QCbdH9TiPVQR0YQPuhgJEmRqDRKp1J1Mzw/WmgrUMQlYKktQR8cjVanRxCZhyHefedt0NRgLsgnqmkn9zs3EXXEzmphE6vdswVSc51duW6gqd7J7u41bJgQTHuk+4iordlCc72D85FA+m60no4eSwlw7NtvpP2m4dlRQuygBVWociuhwDJub2wZOtL9h4x4ix11xokM8LjqsTeBINL1Sm7UFnpdJ2A0XebUlz3qA8JuHt9rWGpE33Uzc5PtJmPIwAOrUVGIm3EX0uDtI+PcUv/0AYidOInTYcGLG34lUpUIZG0v0bbejSkwkduKkk2LXOBEadxd55TIy7POewEP6JtO4sxDB7l7eGQ95KgWkuc+Fg/sk4zRaPQrgaCRyGYGZCV5t5oJqr53CYfRb22eSao2p00NZ+pWRcTfWcH6vCl58tu1HTsdyjylPBLd4D5kmkMTr76TilyWUfD//mOQKTqfXv0iaF9U5jCE/i5yPX8WQn0X08KtJvO6ONj3D9Dcj+HxFHIkpcqa9GUnvc1Wea0sXNnLvY1q+/8r9/+hywROT3Inulm9M4LHnw2gh32OLaEOk5P6dwqP3hbF3XRfKdjUll0tPVfDzogRyNqeQtSHF0z7xthD2b+hC4fZUxt3stkckxcv56csEHr5Hy771Xdi3vguyE0hRFDiwB/Jo3zai9uh/ujgjdgIRY0dQ8px3XiDjliyMW7zzvut/3YIsJKDVttao/f4795dMKvVM2orISEreckcyK6KjsVdVNe93aDIVHA4qP/8MAFtFBdWLvkL9TCoV83znxD+dOBq9g3OcBu/X8hANESN6EzHCO0ukMiLIc91e5z9QRhagbJa2GsDR0NwTy6E/dvfcE+GSkWpe/r+mo6bMXv5TVbeE2eRCrfG9jrpkpJoVy8yUlTr93kOqUgECdkMDEee1baFyJJbyYoK6ZmLIy0IZFklgSjdq/lqFJjYJS1UZDVk7sdZUkjq+abHislpQhLY8Qb3wuP9CSRvXmBnew9sturTIwROT2ifRXUqSggCNhF4XFRIXI6d7mpKD+Ta+/jiOex+rZNsuK2GhUgb3V7N5u4W1G80sW2mkVuek9kAaC5c0Ulzm4KrbSvnfuzHc9fDxu7nKI0KJvu861N2SEBxOQi4dQNFj74PLhaprPFF3XY08UovLaKboif+22D9kxADCrhsKUil136ymcW375PM6IQRBOO0/uH1/BEBQpca5f3cbiwVASJ37mHBkH0AIHtZXiBw30qst9PLzhPCbh7fa1tJPQGYPIeaOCQIgSAMCBIlUKqhTU4WkJ58SkEoFJBJBERXtsx8gxE6cJChjY73lSiRCl+kveD1TR/npO/8Br3GFD+8pRI7s43md8shVQu+5k/yOPXxYD+G8lVOFkH4pfu/R/dXbvF5rUqKE81ZO9Wrrv+QxIX7skHZ5pk17Y4XJU4L9Xu+SKhe2HYgTNu2JFV55Wyukd5MLe4viPdfHTwoS9hbFN/u5+bYA78+WViq8OSdM2Lg7Vti0N7bZPeYtimh2D5n89P+fd/QfbYhUcJZneLVdOixASEtRCKbCdEEm8+4fGS4TjAXpQma6UgAEZ3mGIJc3fV7/925Mu4wr5sEbBe01Qz2vNb1SheS3p3iP/coL/PYHhMAB3YWE6ROFhOfvFtIXvyRIZFLPtZTZjwuqlLg2jak95t8OtxPQXnU+imgtuh82YNzWtuja9sBaXIT20suIvXsSzoYGapa6C207TSZibnfnvi99/11cJqPPfj4RBIw7d5Aw5RFK3337FD3JsaGMCqHrE/+i+uedKCODiRs7hIpvmpJ5lX+1gZ7v30XG/91E+TcbUUaFoB2UTtFHv+FoMKP76wDG7HLSpl5P6YK1mPKq0A5Kp/KHrZ4dQunna0l55Cpq/9iDPCSApLsuxph98oxf5/eqaPF6Yb6DAd29798ruczz++efGPj8k9bTAOjrXTz+gG/PocJ8BxPHeK+kj7yHSNvJL7JzMM/Okw+G8+p77pxdKqWE4CApggDlVb49AfQNJymyVxDadLwrCw4g9rExFD81B6fBROrHz5yccbWRDqcEKmd/hyxIQ5d3HiJv0usASJSKVvPwtxdOo5Gy2c1D4h16PVULF7Taz9+RT83337XfINuR3JnfkzBhOOHDemAuqKbgnZ9QJ4R7rlsr9ex/9HMSJgwn87XbsdU20rCjAJfV/YUTHC4OTP2KhPHDiBt9AYrwIBp2FCA4mlxEDPtKSJo0gu4vj8FptqH7K5uST/845c8qcmbjcsENd5bx1gtRFGxLRSGX0O2CAvKL7Mz9rJ5da7rQaHCRW+CdxuOD+XoKtqVSq3My4DLfNbSPBafJgiK6KarZmluGVK0koF8Gph0HUcSGY9qV47e/VKMCQcCha0R79QXHPY72pmPFCUgkJE6/E6QSTLvyqFuyBoCou69G0z0Je1U95bMWIZFJSZp5L1KNCktuGXWLV+OoayDx5UlI5DKs+eXUfPkbrkZTszZHdduNgOrUVEKGDPVSAiIiIp0LZUIUMY+MRhakoeB+t31QlRpH1MR/uW0CJgtFj7/frD+CQPFT7nxKkXeMIujCc9D/vImQywZS9Mi7CE4XsY/eSuCgHjiq6nHqDVR9+iO2wpZ3tEC7xAl0LCUgIiIiInLMtIcSOCNcREVERERETg6iEhARERHpxIhKQERERKQT0yFsAiIiIiIipwdxJyAiIiLSiRGVgIiIiEgnRlQCIiIiIp0YUQmIiIiIdGJEJSAiIiLSiRGVgIiIiEgnRlQCIiIiIp0YUQmIiIiIdGJEJSAiIiLSiRGVgIiIiEgnRlQCIiIiIp0YUQmIiIiIdGJEJSAiIiLSiRGVgIiIiEgnRlQCIiIiIp0YUQmIiIiIdGJEJSAiIiLSiRGVgIiIiEgnRlQCIiIiIp0YUQmIiIiIdGJEJSAiIiLSiRGVgIiIiEgnRlQCIiIiIp2Y/we1CSH0BjEXYwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_word_cloud(df[\"text\"].loc[df.stars == 4,])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 524,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_word_cloud(df[\"text\"].loc[df.stars == 5,])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 525,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "yelp_review_tokens = df.text.sum()\n",
    "yelp_review_vocabulary = nltk.FreqDist(yelp_review_tokens.split())\n",
    "yelp_review_vocabulary_counts = np.array(list(yelp_review_vocabulary.values()))\n",
    "yelp_review_vocabulary_tokens = list(yelp_review_vocabulary.keys())\n",
    "\n",
    "# A Zipf plot\n",
    "ranks = np.arange(1, len(yelp_review_vocabulary_counts)+1)\n",
    "indices = np.argsort(-yelp_review_vocabulary_counts)\n",
    "frequencies = yelp_review_vocabulary_counts[indices]\n",
    "fig3 = plt.figure(figsize=(12, 8))\n",
    "plt.loglog(ranks, frequencies, marker=\".\")\n",
    "plt.title(\"Zipf plot for Yelp corpus tokens\")\n",
    "plt.xlabel(\"Frequency rank of token\")\n",
    "plt.ylabel(\"Absolute frequency of token\")\n",
    "plt.grid(True)\n",
    "for n in list(np.logspace(-0.5, np.log10(len(yelp_review_vocabulary_counts)), 20).astype(int)):\n",
    "    dummy = plt.text(ranks[n-1], frequencies[n-1], \" \" + yelp_review_vocabulary_tokens[indices[n-1]], \n",
    "                 verticalalignment=\"bottom\",\n",
    "                 horizontalalignment=\"left\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 526,
   "metadata": {},
   "outputs": [],
   "source": [
    "dict1={}\n",
    "dict1[1]=[]\n",
    "dict1[2]=[]\n",
    "dict1[3]=[]\n",
    "dict1[4]=[]\n",
    "dict1[5]=[]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 527,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>stars</th>\n",
       "      <th>oneword</th>\n",
       "      <th>length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [text, stars, oneword, length]\n",
       "Index: []"
      ]
     },
     "execution_count": 527,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nan_rows = df[df['stars'].isnull()]\n",
    "nan_rows"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 528,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>stars</th>\n",
       "      <th>oneword</th>\n",
       "      <th>length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>small north indian</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>cheap low low also</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>compact north indian</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5-8min bus</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>hidden much</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>much poor menu next</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>great saturday huge</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>sumana gr8 seat</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>great hyp</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>dinner good u</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>great neha cant</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td></td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>im sure famous</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>perfect also</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>due</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>.we sprite fresh</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>salt</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>final long well</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>.the</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>good red</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>great</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>im sure famous</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>perfect also</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>due</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>.we sprite fresh</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>salt</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>final long well</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>.the</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>good red</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>great</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5571</th>\n",
       "      <td>red good</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5572</th>\n",
       "      <td>pathetic/poor sour even pure</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5573</th>\n",
       "      <td>ive chicken also good</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5574</th>\n",
       "      <td>great nice agreat</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5575</th>\n",
       "      <td>good</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5576</th>\n",
       "      <td>good higher</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5577</th>\n",
       "      <td></td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5578</th>\n",
       "      <td>next</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5579</th>\n",
       "      <td>cafe</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5580</th>\n",
       "      <td>similar chicken</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5581</th>\n",
       "      <td>pleasant plonk</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5582</th>\n",
       "      <td>much nice</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5583</th>\n",
       "      <td>less n poor small never</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5584</th>\n",
       "      <td>quiet late</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5585</th>\n",
       "      <td>less send</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5586</th>\n",
       "      <td>decent north south indian small</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5587</th>\n",
       "      <td>babycorn local</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5588</th>\n",
       "      <td>great less prior</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5589</th>\n",
       "      <td>popular sad</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5590</th>\n",
       "      <td>regular</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5591</th>\n",
       "      <td>chicken sandwich</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5592</th>\n",
       "      <td>come garma wohi</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5593</th>\n",
       "      <td>ever eaten bad poor</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5594</th>\n",
       "      <td>apart crust still good</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5595</th>\n",
       "      <td>worst ill never first bad cheap</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5596</th>\n",
       "      <td>moti</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5597</th>\n",
       "      <td>decent perfect</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5598</th>\n",
       "      <td>nice interior</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5599</th>\n",
       "      <td>last lot space..</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "      <td>112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5600</th>\n",
       "      <td>ever</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Blah!</td>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5601 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 text  stars  oneword  length\n",
       "0                  small north indian    5.0  Insane!     114\n",
       "1                  cheap low low also    1.0   Avoid!     103\n",
       "2                compact north indian    5.0  Insane!     108\n",
       "3                          5-8min bus    5.0  Insane!     112\n",
       "4                         hidden much    5.0  Insane!     105\n",
       "5                 much poor menu next    1.0   Avoid!      80\n",
       "6                 great saturday huge    5.0  Insane!      99\n",
       "7                     sumana gr8 seat    5.0  Insane!     101\n",
       "8                           great hyp    3.0  Average     118\n",
       "9                       dinner good u    3.0  Average     100\n",
       "10                    great neha cant    5.0  Insane!     106\n",
       "11                                       5.0  Insane!      95\n",
       "12                     im sure famous    5.0  Insane!      96\n",
       "13                       perfect also    5.0  Insane!     100\n",
       "14                                due    3.0  Average      98\n",
       "15                   .we sprite fresh    5.0  Insane!     127\n",
       "16                               salt    5.0  Insane!     107\n",
       "17                    final long well    5.0  Insane!     105\n",
       "18                               .the    3.0  Average     111\n",
       "19                           good red    5.0  Insane!      85\n",
       "20                              great    5.0  Insane!      90\n",
       "21                     im sure famous    5.0  Insane!      96\n",
       "22                       perfect also    5.0  Insane!     100\n",
       "23                                due    3.0  Average      98\n",
       "24                   .we sprite fresh    5.0  Insane!     127\n",
       "25                               salt    5.0  Insane!     107\n",
       "26                    final long well    5.0  Insane!     105\n",
       "27                               .the    3.0  Average     111\n",
       "28                           good red    5.0  Insane!      85\n",
       "29                              great    5.0  Insane!      90\n",
       "...                               ...    ...      ...     ...\n",
       "5571                         red good    4.0   Great!     101\n",
       "5572     pathetic/poor sour even pure    1.0   Avoid!     109\n",
       "5573            ive chicken also good    4.0   Great!     107\n",
       "5574                great nice agreat    5.0  Insane!     113\n",
       "5575                             good    5.0  Insane!     102\n",
       "5576                      good higher    5.0  Insane!     115\n",
       "5577                                     5.0  Insane!     115\n",
       "5578                             next    5.0  Insane!     106\n",
       "5579                             cafe    5.0  Insane!      85\n",
       "5580                  similar chicken    3.0  Average     110\n",
       "5581                   pleasant plonk    4.0   Great!      93\n",
       "5582                        much nice    4.0   Great!     109\n",
       "5583          less n poor small never    1.0   Avoid!      87\n",
       "5584                       quiet late    4.0   Great!      90\n",
       "5585                        less send    1.0   Avoid!      84\n",
       "5586  decent north south indian small    4.0   Great!     104\n",
       "5587                   babycorn local    4.0   Great!      95\n",
       "5588                 great less prior    5.0  Insane!      99\n",
       "5589                      popular sad    1.0   Avoid!      75\n",
       "5590                          regular    3.0  Average     109\n",
       "5591                 chicken sandwich    5.0  Insane!     101\n",
       "5592                  come garma wohi    5.0  Insane!     135\n",
       "5593              ever eaten bad poor    1.0   Avoid!      96\n",
       "5594           apart crust still good    4.0   Great!     106\n",
       "5595  worst ill never first bad cheap    1.0   Avoid!      96\n",
       "5596                             moti    1.0   Avoid!      97\n",
       "5597                   decent perfect    5.0  Insane!     109\n",
       "5598                    nice interior    4.0   Great!     104\n",
       "5599                 last lot space..    4.0   Great!     112\n",
       "5600                             ever    2.0    Blah!      86\n",
       "\n",
       "[5601 rows x 4 columns]"
      ]
     },
     "execution_count": 528,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 529,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>stars</th>\n",
       "      <th>oneword</th>\n",
       "      <th>length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [text, stars, oneword, length]\n",
       "Index: []"
      ]
     },
     "execution_count": 529,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# df['stars']\n",
    "nan_rows = df[df['stars'].isnull()]\n",
    "nan_rows"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 530,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       5.0\n",
       "1       1.0\n",
       "2       5.0\n",
       "3       5.0\n",
       "4       5.0\n",
       "5       1.0\n",
       "6       5.0\n",
       "7       5.0\n",
       "8       3.0\n",
       "9       3.0\n",
       "10      5.0\n",
       "11      5.0\n",
       "12      5.0\n",
       "13      5.0\n",
       "14      3.0\n",
       "15      5.0\n",
       "16      5.0\n",
       "17      5.0\n",
       "18      3.0\n",
       "19      5.0\n",
       "20      5.0\n",
       "21      5.0\n",
       "22      5.0\n",
       "23      3.0\n",
       "24      5.0\n",
       "25      5.0\n",
       "26      5.0\n",
       "27      3.0\n",
       "28      5.0\n",
       "29      5.0\n",
       "       ... \n",
       "5571    4.0\n",
       "5572    1.0\n",
       "5573    4.0\n",
       "5574    5.0\n",
       "5575    5.0\n",
       "5576    5.0\n",
       "5577    5.0\n",
       "5578    5.0\n",
       "5579    5.0\n",
       "5580    3.0\n",
       "5581    4.0\n",
       "5582    4.0\n",
       "5583    1.0\n",
       "5584    4.0\n",
       "5585    1.0\n",
       "5586    4.0\n",
       "5587    4.0\n",
       "5588    5.0\n",
       "5589    1.0\n",
       "5590    3.0\n",
       "5591    5.0\n",
       "5592    5.0\n",
       "5593    1.0\n",
       "5594    4.0\n",
       "5595    1.0\n",
       "5596    1.0\n",
       "5597    5.0\n",
       "5598    4.0\n",
       "5599    4.0\n",
       "5600    2.0\n",
       "Name: stars, Length: 5601, dtype: float64"
      ]
     },
     "execution_count": 530,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['stars']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 531,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5.0 0\n",
      "1.0 1\n",
      "5.0 2\n",
      "5.0 3\n",
      "5.0 4\n",
      "1.0 5\n",
      "5.0 6\n",
      "5.0 7\n",
      "3.0 8\n",
      "3.0 9\n",
      "5.0 10\n",
      "5.0 11\n",
      "5.0 12\n",
      "5.0 13\n",
      "3.0 14\n",
      "5.0 15\n",
      "5.0 16\n",
      "5.0 17\n",
      "3.0 18\n",
      "5.0 19\n",
      "5.0 20\n",
      "5.0 21\n",
      "5.0 22\n",
      "3.0 23\n",
      "5.0 24\n",
      "5.0 25\n",
      "5.0 26\n",
      "3.0 27\n",
      "5.0 28\n",
      "5.0 29\n",
      "5.0 30\n",
      "3.0 31\n",
      "5.0 32\n",
      "5.0 33\n",
      "1.0 34\n",
      "4.0 35\n",
      "4.0 36\n",
      "4.0 37\n",
      "5.0 38\n",
      "4.0 39\n",
      "3.0 40\n",
      "4.0 41\n",
      "5.0 42\n",
      "4.0 43\n",
      "5.0 44\n",
      "5.0 45\n",
      "2.0 46\n",
      "3.0 47\n",
      "5.0 48\n",
      "5.0 49\n",
      "3.0 50\n",
      "4.0 51\n",
      "4.0 52\n",
      "5.0 53\n",
      "5.0 54\n",
      "4.0 55\n",
      "5.0 56\n",
      "4.0 57\n",
      "5.0 58\n",
      "5.0 59\n",
      "5.0 60\n",
      "5.0 61\n",
      "4.0 62\n",
      "5.0 63\n",
      "5.0 64\n",
      "4.0 65\n",
      "5.0 66\n",
      "4.0 67\n",
      "4.0 68\n",
      "4.0 69\n",
      "4.0 70\n",
      "1.0 71\n",
      "5.0 72\n",
      "5.0 73\n",
      "5.0 74\n",
      "5.0 75\n",
      "5.0 76\n",
      "5.0 77\n",
      "5.0 78\n",
      "4.0 79\n",
      "5.0 80\n",
      "5.0 81\n",
      "4.0 82\n",
      "5.0 83\n",
      "5.0 84\n",
      "5.0 85\n",
      "5.0 86\n",
      "1.0 87\n",
      "4.0 88\n",
      "5.0 89\n",
      "5.0 90\n",
      "5.0 91\n",
      "5.0 92\n",
      "1.0 93\n",
      "5.0 94\n",
      "4.5 95\n",
      "5.0 96\n",
      "5.0 97\n",
      "5.0 98\n",
      "5.0 99\n",
      "5.0 100\n",
      "5.0 101\n",
      "4.0 102\n",
      "2.0 103\n",
      "5.0 104\n",
      "4.0 105\n",
      "5.0 106\n",
      "5.0 107\n",
      "4.0 108\n",
      "4.0 109\n",
      "4.0 110\n",
      "5.0 111\n",
      "4.0 112\n",
      "2.0 113\n",
      "5.0 114\n",
      "4.0 115\n",
      "5.0 116\n",
      "5.0 117\n",
      "4.0 118\n",
      "4.0 119\n",
      "4.0 120\n",
      "5.0 121\n",
      "5.0 122\n",
      "5.0 123\n",
      "5.0 124\n",
      "3.0 125\n",
      "5.0 126\n",
      "4.0 127\n",
      "5.0 128\n",
      "3.0 129\n",
      "4.0 130\n",
      "5.0 131\n",
      "1.0 132\n",
      "5.0 133\n",
      "1.0 134\n",
      "1.0 135\n",
      "1.0 136\n",
      "4.0 137\n",
      "1.0 138\n",
      "5.0 139\n",
      "2.0 140\n",
      "5.0 141\n",
      "4.0 142\n",
      "3.0 143\n",
      "4.0 144\n",
      "4.0 145\n",
      "5.0 146\n",
      "1.0 147\n",
      "1.0 148\n",
      "4.0 149\n",
      "5.0 150\n",
      "5.0 151\n",
      "5.0 152\n",
      "4.0 153\n",
      "5.0 154\n",
      "4.0 155\n",
      "4.0 156\n",
      "4.0 157\n",
      "5.0 158\n",
      "5.0 159\n",
      "4.0 160\n",
      "3.0 161\n",
      "5.0 162\n",
      "5.0 163\n",
      "4.0 164\n",
      "4.0 165\n",
      "1.0 166\n",
      "1.0 167\n",
      "3.0 168\n",
      "5.0 169\n",
      "5.0 170\n",
      "4.0 171\n",
      "3.0 172\n",
      "3.0 173\n",
      "3.0 174\n",
      "5.0 175\n",
      "4.0 176\n",
      "5.0 177\n",
      "5.0 178\n",
      "5.0 179\n",
      "5.0 180\n",
      "4.0 181\n",
      "5.0 182\n",
      "4.0 183\n",
      "4.0 184\n",
      "4.0 185\n",
      "5.0 186\n",
      "5.0 187\n",
      "5.0 188\n",
      "4.0 189\n",
      "5.0 190\n",
      "5.0 191\n",
      "5.0 192\n",
      "5.0 193\n",
      "4.0 194\n",
      "5.0 195\n",
      "5.0 196\n",
      "4.0 197\n",
      "5.0 198\n",
      "5.0 199\n",
      "5.0 200\n",
      "5.0 201\n",
      "4.0 202\n",
      "5.0 203\n",
      "5.0 204\n",
      "5.0 205\n",
      "5.0 206\n",
      "5.0 207\n",
      "4.0 208\n",
      "4.0 209\n",
      "5.0 210\n",
      "5.0 211\n",
      "4.0 212\n",
      "4.0 213\n",
      "4.0 214\n",
      "5.0 215\n",
      "3.0 216\n",
      "4.0 217\n",
      "5.0 218\n",
      "5.0 219\n",
      "4.0 220\n",
      "4.0 221\n",
      "4.0 222\n",
      "5.0 223\n",
      "4.0 224\n",
      "4.0 225\n",
      "4.5 226\n",
      "4.0 227\n",
      "4.0 228\n",
      "5.0 229\n",
      "4.0 230\n",
      "4.0 231\n",
      "5.0 232\n",
      "4.0 233\n",
      "4.0 234\n",
      "5.0 235\n",
      "1.0 236\n",
      "5.0 237\n",
      "4.0 238\n",
      "5.0 239\n",
      "2.0 240\n",
      "5.0 241\n",
      "5.0 242\n",
      "5.0 243\n",
      "4.0 244\n",
      "4.0 245\n",
      "5.0 246\n",
      "4.0 247\n",
      "4.0 248\n",
      "5.0 249\n",
      "4.0 250\n",
      "4.0 251\n",
      "4.0 252\n",
      "5.0 253\n",
      "5.0 254\n",
      "5.0 255\n",
      "4.0 256\n",
      "5.0 257\n",
      "5.0 258\n",
      "5.0 259\n",
      "4.0 260\n",
      "5.0 261\n",
      "4.0 262\n",
      "5.0 263\n",
      "5.0 264\n",
      "5.0 265\n",
      "4.0 266\n",
      "5.0 267\n",
      "4.0 268\n",
      "4.0 269\n",
      "5.0 270\n",
      "5.0 271\n",
      "5.0 272\n",
      "3.0 273\n",
      "5.0 274\n",
      "5.0 275\n",
      "5.0 276\n",
      "5.0 277\n",
      "5.0 278\n",
      "5.0 279\n",
      "5.0 280\n",
      "3.0 281\n",
      "4.0 282\n",
      "1.0 283\n",
      "5.0 284\n",
      "3.0 285\n",
      "4.0 286\n",
      "5.0 287\n",
      "3.0 288\n",
      "5.0 289\n",
      "4.0 290\n",
      "5.0 291\n",
      "5.0 292\n",
      "5.0 293\n",
      "4.0 294\n",
      "5.0 295\n",
      "3.0 296\n",
      "3.0 297\n",
      "3.0 298\n",
      "4.0 299\n",
      "5.0 300\n",
      "5.0 301\n",
      "5.0 302\n",
      "1.0 303\n",
      "5.0 304\n",
      "5.0 305\n",
      "3.0 306\n",
      "5.0 307\n",
      "5.0 308\n",
      "4.0 309\n",
      "5.0 310\n",
      "5.0 311\n",
      "1.0 312\n",
      "4.0 313\n",
      "1.0 314\n",
      "1.0 315\n",
      "5.0 316\n",
      "4.0 317\n",
      "5.0 318\n",
      "5.0 319\n",
      "5.0 320\n",
      "3.0 321\n",
      "4.0 322\n",
      "4.0 323\n",
      "3.0 324\n",
      "5.0 325\n",
      "4.5 326\n",
      "5.0 327\n",
      "3.0 328\n",
      "5.0 329\n",
      "4.0 330\n",
      "1.0 331\n",
      "5.0 332\n",
      "2.0 333\n",
      "5.0 334\n",
      "5.0 335\n",
      "4.0 336\n",
      "1.0 337\n",
      "5.0 338\n",
      "5.0 339\n",
      "5.0 340\n",
      "5.0 341\n",
      "2.0 342\n",
      "5.0 343\n",
      "4.0 344\n",
      "5.0 345\n",
      "1.0 346\n",
      "5.0 347\n",
      "4.0 348\n",
      "4.0 349\n",
      "4.0 350\n",
      "5.0 351\n",
      "5.0 352\n",
      "5.0 353\n",
      "5.0 354\n",
      "5.0 355\n",
      "5.0 356\n",
      "5.0 357\n",
      "5.0 358\n",
      "5.0 359\n",
      "5.0 360\n",
      "4.0 361\n",
      "3.0 362\n",
      "3.0 363\n",
      "4.0 364\n",
      "4.0 365\n",
      "4.0 366\n",
      "3.0 367\n",
      "2.0 368\n",
      "4.0 369\n",
      "4.0 370\n",
      "4.0 371\n",
      "4.0 372\n",
      "3.0 373\n",
      "2.0 374\n",
      "4.0 375\n",
      "4.0 376\n",
      "5.0 377\n",
      "5.0 378\n",
      "4.0 379\n",
      "4.0 380\n",
      "5.0 381\n",
      "5.0 382\n",
      "5.0 383\n",
      "4.0 384\n",
      "4.0 385\n",
      "5.0 386\n",
      "2.0 387\n",
      "5.0 388\n",
      "3.0 389\n",
      "5.0 390\n",
      "4.0 391\n",
      "1.0 392\n",
      "2.0 393\n",
      "5.0 394\n",
      "3.0 395\n",
      "5.0 396\n",
      "4.0 397\n",
      "1.0 398\n",
      "2.0 399\n",
      "5.0 400\n",
      "5.0 401\n",
      "5.0 402\n",
      "4.0 403\n",
      "4.0 404\n",
      "5.0 405\n",
      "2.0 406\n",
      "5.0 407\n",
      "5.0 408\n",
      "5.0 409\n",
      "4.0 410\n",
      "4.0 411\n",
      "5.0 412\n",
      "5.0 413\n",
      "1.0 414\n",
      "5.0 415\n",
      "4.0 416\n",
      "5.0 417\n",
      "1.0 418\n",
      "5.0 419\n",
      "5.0 420\n",
      "5.0 421\n",
      "4.0 422\n",
      "5.0 423\n",
      "5.0 424\n",
      "5.0 425\n",
      "4.0 426\n",
      "5.0 427\n",
      "5.0 428\n",
      "4.0 429\n",
      "5.0 430\n",
      "4.0 431\n",
      "4.0 432\n",
      "4.0 433\n",
      "5.0 434\n",
      "4.0 435\n",
      "5.0 436\n",
      "4.0 437\n",
      "4.0 438\n",
      "4.0 439\n",
      "3.0 440\n",
      "1.0 441\n",
      "5.0 442\n",
      "4.0 443\n",
      "5.0 444\n",
      "5.0 445\n",
      "5.0 446\n",
      "4.0 447\n",
      "5.0 448\n",
      "5.0 449\n",
      "5.0 450\n",
      "5.0 451\n",
      "5.0 452\n",
      "4.0 453\n",
      "5.0 454\n",
      "5.0 455\n",
      "5.0 456\n",
      "5.0 457\n",
      "5.0 458\n",
      "5.0 459\n",
      "5.0 460\n",
      "5.0 461\n",
      "4.0 462\n",
      "1.0 463\n",
      "5.0 464\n",
      "5.0 465\n",
      "5.0 466\n",
      "5.0 467\n",
      "4.0 468\n",
      "5.0 469\n",
      "5.0 470\n",
      "5.0 471\n",
      "5.0 472\n",
      "5.0 473\n",
      "5.0 474\n",
      "4.0 475\n",
      "5.0 476\n",
      "5.0 477\n",
      "4.0 478\n",
      "5.0 479\n",
      "4.0 480\n",
      "5.0 481\n",
      "5.0 482\n",
      "5.0 483\n",
      "5.0 484\n",
      "4.0 485\n",
      "4.0 486\n",
      "5.0 487\n",
      "5.0 488\n",
      "5.0 489\n",
      "5.0 490\n",
      "5.0 491\n",
      "5.0 492\n",
      "5.0 493\n",
      "5.0 494\n",
      "5.0 495\n",
      "5.0 496\n",
      "4.0 497\n",
      "5.0 498\n",
      "5.0 499\n",
      "4.0 500\n",
      "5.0 501\n",
      "4.0 502\n",
      "5.0 503\n",
      "5.0 504\n",
      "4.0 505\n",
      "4.0 506\n",
      "5.0 507\n",
      "5.0 508\n",
      "4.0 509\n",
      "5.0 510\n",
      "4.0 511\n",
      "5.0 512\n",
      "4.0 513\n",
      "4.0 514\n",
      "4.0 515\n",
      "5.0 516\n",
      "5.0 517\n",
      "5.0 518\n",
      "5.0 519\n",
      "5.0 520\n",
      "5.0 521\n",
      "3.0 522\n",
      "5.0 523\n",
      "4.0 524\n",
      "4.0 525\n",
      "3.0 526\n",
      "5.0 527\n",
      "4.0 528\n",
      "3.0 529\n",
      "5.0 530\n",
      "4.0 531\n",
      "4.0 532\n",
      "3.0 533\n",
      "5.0 534\n",
      "4.0 535\n",
      "3.0 536\n",
      "4.0 537\n",
      "5.0 538\n",
      "4.5 539\n",
      "4.0 540\n",
      "3.0 541\n",
      "5.0 542\n",
      "5.0 543\n",
      "5.0 544\n",
      "3.0 545\n",
      "4.0 546\n",
      "5.0 547\n",
      "4.0 548\n",
      "5.0 549\n",
      "4.0 550\n",
      "5.0 551\n",
      "2.0 552\n",
      "1.0 553\n",
      "4.0 554\n",
      "4.0 555\n",
      "3.0 556\n",
      "4.0 557\n",
      "2.0 558\n",
      "1.0 559\n",
      "4.0 560\n",
      "4.0 561\n",
      "3.0 562\n",
      "4.0 563\n",
      "4.0 564\n",
      "4.0 565\n",
      "3.0 566\n",
      "1.0 567\n",
      "3.0 568\n",
      "5.0 569\n",
      "4.0 570\n",
      "5.0 571\n",
      "5.0 572\n",
      "5.0 573\n",
      "5.0 574\n",
      "5.0 575\n",
      "4.0 576\n",
      "5.0 577\n",
      "5.0 578\n",
      "5.0 579\n",
      "3.0 580\n",
      "3.0 581\n",
      "4.0 582\n",
      "4.0 583\n",
      "5.0 584\n",
      "5.0 585\n",
      "2.0 586\n",
      "5.0 587\n",
      "1.0 588\n",
      "5.0 589\n",
      "5.0 590\n",
      "5.0 591\n",
      "5.0 592\n",
      "1.0 593\n",
      "5.0 594\n",
      "5.0 595\n",
      "5.0 596\n",
      "1.0 597\n",
      "5.0 598\n",
      "5.0 599\n",
      "5.0 600\n",
      "4.0 601\n",
      "2.0 602\n",
      "2.0 603\n",
      "4.0 604\n",
      "5.0 605\n",
      "1.0 606\n",
      "4.0 607\n",
      "3.0 608\n",
      "4.0 609\n",
      "5.0 610\n",
      "1.0 611\n",
      "4.0 612\n",
      "3.0 613\n",
      "3.0 614\n",
      "4.0 615\n",
      "4.0 616\n",
      "5.0 617\n",
      "5.0 618\n",
      "3.0 619\n",
      "4.0 620\n",
      "5.0 621\n",
      "4.0 622\n",
      "5.0 623\n",
      "1.0 624\n",
      "5.0 625\n",
      "4.0 626\n",
      "5.0 627\n",
      "1.0 628\n",
      "5.0 629\n",
      "4.0 630\n",
      "5.0 631\n",
      "4.0 632\n",
      "5.0 633\n",
      "1.0 634\n",
      "5.0 635\n",
      "4.0 636\n",
      "5.0 637\n",
      "4.0 638\n",
      "4.0 639\n",
      "5.0 640\n",
      "5.0 641\n",
      "2.0 642\n",
      "5.0 643\n",
      "5.0 644\n",
      "5.0 645\n",
      "5.0 646\n",
      "1.0 647\n",
      "4.0 648\n",
      "5.0 649\n",
      "5.0 650\n",
      "5.0 651\n",
      "3.0 652\n",
      "5.0 653\n",
      "5.0 654\n",
      "5.0 655\n",
      "5.0 656\n",
      "5.0 657\n",
      "5.0 658\n",
      "4.0 659\n",
      "4.5 660\n",
      "5.0 661\n",
      "4.0 662\n",
      "3.0 663\n",
      "3.0 664\n",
      "4.0 665\n",
      "1.0 666\n",
      "1.0 667\n",
      "3.0 668\n",
      "3.0 669\n",
      "4.0 670\n",
      "1.0 671\n",
      "1.0 672\n",
      "5.0 673\n",
      "4.0 674\n",
      "5.0 675\n",
      "3.0 676\n",
      "4.0 677\n",
      "5.0 678\n",
      "4.0 679\n",
      "5.0 680\n",
      "3.0 681\n",
      "4.0 682\n",
      "4.0 683\n",
      "4.0 684\n",
      "5.0 685\n",
      "4.0 686\n",
      "5.0 687\n",
      "5.0 688\n",
      "4.0 689\n",
      "4.0 690\n",
      "2.0 691\n",
      "5.0 692\n",
      "5.0 693\n",
      "5.0 694\n",
      "4.0 695\n",
      "4.0 696\n",
      "3.0 697\n",
      "4.0 698\n",
      "5.0 699\n",
      "4.0 700\n",
      "4.0 701\n",
      "3.0 702\n",
      "4.0 703\n",
      "5.0 704\n",
      "4.0 705\n",
      "3.0 706\n",
      "5.0 707\n",
      "4.0 708\n",
      "1.0 709\n",
      "1.0 710\n",
      "1.0 711\n",
      "1.0 712\n",
      "4.0 713\n",
      "4.0 714\n",
      "4.0 715\n",
      "4.0 716\n",
      "3.0 717\n",
      "5.0 718\n",
      "3.0 719\n",
      "2.0 720\n",
      "5.0 721\n",
      "5.0 722\n",
      "5.0 723\n",
      "5.0 724\n",
      "2.0 725\n",
      "4.0 726\n",
      "5.0 727\n",
      "4.0 728\n",
      "1.0 729\n",
      "2.0 730\n",
      "5.0 731\n",
      "4.0 732\n",
      "4.0 733\n",
      "4.0 734\n",
      "4.0 735\n",
      "4.0 736\n",
      "4.0 737\n",
      "3.0 738\n",
      "5.0 739\n",
      "5.0 740\n",
      "5.0 741\n",
      "5.0 742\n",
      "5.0 743\n",
      "5.0 744\n",
      "5.0 745\n",
      "5.0 746\n",
      "4.0 747\n",
      "5.0 748\n",
      "5.0 749\n",
      "4.0 750\n",
      "4.0 751\n",
      "5.0 752\n",
      "5.0 753\n",
      "5.0 754\n",
      "5.0 755\n",
      "4.0 756\n",
      "5.0 757\n",
      "5.0 758\n",
      "4.0 759\n",
      "4.0 760\n",
      "4.0 761\n",
      "5.0 762\n",
      "5.0 763\n",
      "5.0 764\n",
      "5.0 765\n",
      "4.0 766\n",
      "1.0 767\n",
      "1.0 768\n",
      "4.0 769\n",
      "5.0 770\n",
      "5.0 771\n",
      "5.0 772\n",
      "5.0 773\n",
      "5.0 774\n",
      "5.0 775\n",
      "4.0 776\n",
      "5.0 777\n",
      "5.0 778\n",
      "5.0 779\n",
      "4.0 780\n",
      "1.0 781\n",
      "5.0 782\n",
      "5.0 783\n",
      "5.0 784\n",
      "5.0 785\n",
      "4.0 786\n",
      "5.0 787\n",
      "5.0 788\n",
      "5.0 789\n",
      "5.0 790\n",
      "1.0 791\n",
      "4.5 792\n",
      "5.0 793\n",
      "4.0 794\n",
      "4.0 795\n",
      "5.0 796\n",
      "4.0 797\n",
      "5.0 798\n",
      "5.0 799\n",
      "4.0 800\n",
      "5.0 801\n",
      "4.0 802\n",
      "5.0 803\n",
      "5.0 804\n",
      "5.0 805\n",
      "5.0 806\n",
      "5.0 807\n",
      "4.0 808\n",
      "4.0 809\n",
      "3.0 810\n",
      "5.0 811\n",
      "5.0 812\n",
      "4.0 813\n",
      "5.0 814\n",
      "5.0 815\n",
      "5.0 816\n",
      "4.0 817\n",
      "5.0 818\n",
      "4.0 819\n",
      "5.0 820\n",
      "5.0 821\n",
      "5.0 822\n",
      "4.0 823\n",
      "3.0 824\n",
      "5.0 825\n",
      "5.0 826\n",
      "5.0 827\n",
      "3.0 828\n",
      "2.0 829\n",
      "5.0 830\n",
      "4.0 831\n",
      "5.0 832\n",
      "5.0 833\n",
      "5.0 834\n",
      "5.0 835\n",
      "5.0 836\n",
      "5.0 837\n",
      "5.0 838\n",
      "1.0 839\n",
      "1.0 840\n",
      "1.0 841\n",
      "2.0 842\n",
      "2.0 843\n",
      "4.0 844\n",
      "1.0 845\n",
      "4.0 846\n",
      "5.0 847\n",
      "5.0 848\n",
      "1.0 849\n",
      "1.0 850\n",
      "1.0 851\n",
      "2.0 852\n",
      "2.0 853\n",
      "4.0 854\n",
      "1.0 855\n",
      "4.0 856\n",
      "5.0 857\n",
      "5.0 858\n",
      "5.0 859\n",
      "4.0 860\n",
      "5.0 861\n",
      "4.0 862\n",
      "5.0 863\n",
      "4.0 864\n",
      "5.0 865\n",
      "5.0 866\n",
      "5.0 867\n",
      "5.0 868\n",
      "4.0 869\n",
      "5.0 870\n",
      "5.0 871\n",
      "5.0 872\n",
      "5.0 873\n",
      "1.0 874\n",
      "5.0 875\n",
      "5.0 876\n",
      "5.0 877\n",
      "4.0 878\n",
      "5.0 879\n",
      "5.0 880\n",
      "5.0 881\n",
      "5.0 882\n",
      "1.0 883\n",
      "5.0 884\n",
      "5.0 885\n",
      "5.0 886\n",
      "5.0 887\n",
      "5.0 888\n",
      "5.0 889\n",
      "4.0 890\n",
      "5.0 891\n",
      "5.0 892\n",
      "5.0 893\n",
      "5.0 894\n",
      "5.0 895\n",
      "5.0 896\n",
      "5.0 897\n",
      "5.0 898\n",
      "3.0 899\n",
      "5.0 900\n",
      "5.0 901\n",
      "5.0 902\n",
      "5.0 903\n",
      "5.0 904\n",
      "5.0 905\n",
      "5.0 906\n",
      "5.0 907\n",
      "5.0 908\n",
      "5.0 909\n",
      "5.0 910\n",
      "5.0 911\n",
      "5.0 912\n",
      "5.0 913\n",
      "4.0 914\n",
      "5.0 915\n",
      "5.0 916\n",
      "4.0 917\n",
      "5.0 918\n",
      "5.0 919\n",
      "5.0 920\n",
      "5.0 921\n",
      "5.0 922\n",
      "4.0 923\n",
      "1.0 924\n",
      "5.0 925\n",
      "4.0 926\n",
      "3.0 927\n",
      "5.0 928\n",
      "5.0 929\n",
      "5.0 930\n",
      "5.0 931\n",
      "5.0 932\n",
      "5.0 933\n",
      "5.0 934\n",
      "5.0 935\n",
      "5.0 936\n",
      "5.0 937\n",
      "5.0 938\n",
      "5.0 939\n",
      "4.0 940\n",
      "4.0 941\n",
      "5.0 942\n",
      "5.0 943\n",
      "4.0 944\n",
      "5.0 945\n",
      "4.0 946\n",
      "5.0 947\n",
      "5.0 948\n",
      "5.0 949\n",
      "4.0 950\n",
      "5.0 951\n",
      "4.0 952\n",
      "5.0 953\n",
      "5.0 954\n",
      "5.0 955\n",
      "3.0 956\n",
      "4.0 957\n",
      "5.0 958\n",
      "1.0 959\n",
      "1.0 960\n",
      "4.0 961\n",
      "3.0 962\n",
      "1.0 963\n",
      "3.0 964\n",
      "4.0 965\n",
      "5.0 966\n",
      "1.0 967\n",
      "1.0 968\n",
      "4.0 969\n",
      "3.0 970\n",
      "1.0 971\n",
      "5.0 972\n",
      "5.0 973\n",
      "5.0 974\n",
      "5.0 975\n",
      "5.0 976\n",
      "5.0 977\n",
      "5.0 978\n",
      "5.0 979\n",
      "5.0 980\n",
      "5.0 981\n",
      "5.0 982\n",
      "4.0 983\n",
      "4.0 984\n",
      "5.0 985\n",
      "5.0 986\n",
      "1.0 987\n",
      "5.0 988\n",
      "4.0 989\n",
      "5.0 990\n",
      "5.0 991\n",
      "5.0 992\n",
      "5.0 993\n",
      "5.0 994\n",
      "5.0 995\n",
      "5.0 996\n",
      "5.0 997\n",
      "3.0 998\n",
      "5.0 999\n",
      "5.0 1000\n",
      "5.0 1001\n",
      "1.0 1002\n",
      "5.0 1003\n",
      "4.0 1004\n",
      "4.0 1005\n",
      "4.0 1006\n",
      "5.0 1007\n",
      "5.0 1008\n",
      "5.0 1009\n",
      "5.0 1010\n",
      "5.0 1011\n",
      "5.0 1012\n",
      "5.0 1013\n",
      "5.0 1014\n",
      "5.0 1015\n",
      "5.0 1016\n",
      "5.0 1017\n",
      "5.0 1018\n",
      "5.0 1019\n",
      "5.0 1020\n",
      "4.0 1021\n",
      "4.0 1022\n",
      "5.0 1023\n",
      "5.0 1024\n",
      "2.0 1025\n",
      "5.0 1026\n",
      "5.0 1027\n",
      "5.0 1028\n",
      "5.0 1029\n",
      "5.0 1030\n",
      "5.0 1031\n",
      "5.0 1032\n",
      "5.0 1033\n",
      "5.0 1034\n",
      "5.0 1035\n",
      "5.0 1036\n",
      "5.0 1037\n",
      "5.0 1038\n",
      "5.0 1039\n",
      "5.0 1040\n",
      "5.0 1041\n",
      "5.0 1042\n",
      "5.0 1043\n",
      "5.0 1044\n",
      "1.0 1045\n",
      "5.0 1046\n",
      "5.0 1047\n",
      "5.0 1048\n",
      "5.0 1049\n",
      "5.0 1050\n",
      "5.0 1051\n",
      "4.0 1052\n",
      "5.0 1053\n",
      "5.0 1054\n",
      "5.0 1055\n",
      "5.0 1056\n",
      "5.0 1057\n",
      "5.0 1058\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5.0 1059\n",
      "5.0 1060\n",
      "5.0 1061\n",
      "3.0 1062\n",
      "4.0 1063\n",
      "1.0 1064\n",
      "1.0 1065\n",
      "1.0 1066\n",
      "4.0 1067\n",
      "4.0 1068\n",
      "5.0 1069\n",
      "5.0 1070\n",
      "4.0 1071\n",
      "4.0 1072\n",
      "5.0 1073\n",
      "5.0 1074\n",
      "4.5 1075\n",
      "5.0 1076\n",
      "4.0 1077\n",
      "4.0 1078\n",
      "5.0 1079\n",
      "5.0 1080\n",
      "5.0 1081\n",
      "5.0 1082\n",
      "5.0 1083\n",
      "5.0 1084\n",
      "5.0 1085\n",
      "5.0 1086\n",
      "5.0 1087\n",
      "5.0 1088\n",
      "5.0 1089\n",
      "5.0 1090\n",
      "5.0 1091\n",
      "5.0 1092\n",
      "5.0 1093\n",
      "5.0 1094\n",
      "5.0 1095\n",
      "5.0 1096\n",
      "5.0 1097\n",
      "5.0 1098\n",
      "5.0 1099\n",
      "5.0 1100\n",
      "5.0 1101\n",
      "2.0 1102\n",
      "5.0 1103\n",
      "5.0 1104\n",
      "2.0 1105\n",
      "5.0 1106\n",
      "4.0 1107\n",
      "5.0 1108\n",
      "5.0 1109\n",
      "3.0 1110\n",
      "5.0 1111\n",
      "5.0 1112\n",
      "5.0 1113\n",
      "5.0 1114\n",
      "5.0 1115\n",
      "5.0 1116\n",
      "4.0 1117\n",
      "5.0 1118\n",
      "1.0 1119\n",
      "4.0 1120\n",
      "1.0 1121\n",
      "5.0 1122\n",
      "5.0 1123\n",
      "3.0 1124\n",
      "3.0 1125\n",
      "4.0 1126\n",
      "3.0 1127\n",
      "5.0 1128\n",
      "3.0 1129\n",
      "5.0 1130\n",
      "5.0 1131\n",
      "3.0 1132\n",
      "5.0 1133\n",
      "5.0 1134\n",
      "5.0 1135\n",
      "5.0 1136\n",
      "5.0 1137\n",
      "5.0 1138\n",
      "5.0 1139\n",
      "5.0 1140\n",
      "5.0 1141\n",
      "4.0 1142\n",
      "5.0 1143\n",
      "5.0 1144\n",
      "5.0 1145\n",
      "5.0 1146\n",
      "5.0 1147\n",
      "5.0 1148\n",
      "5.0 1149\n",
      "5.0 1150\n",
      "5.0 1151\n",
      "5.0 1152\n",
      "5.0 1153\n",
      "5.0 1154\n",
      "5.0 1155\n",
      "5.0 1156\n",
      "5.0 1157\n",
      "5.0 1158\n",
      "5.0 1159\n",
      "5.0 1160\n",
      "5.0 1161\n",
      "5.0 1162\n",
      "5.0 1163\n",
      "4.0 1164\n",
      "5.0 1165\n",
      "5.0 1166\n",
      "5.0 1167\n",
      "5.0 1168\n",
      "5.0 1169\n",
      "5.0 1170\n",
      "1.0 1171\n",
      "4.5 1172\n",
      "4.0 1173\n",
      "5.0 1174\n",
      "4.0 1175\n",
      "1.0 1176\n",
      "5.0 1177\n",
      "5.0 1178\n",
      "5.0 1179\n",
      "5.0 1180\n",
      "5.0 1181\n",
      "5.0 1182\n",
      "4.0 1183\n",
      "5.0 1184\n",
      "5.0 1185\n",
      "4.0 1186\n",
      "5.0 1187\n",
      "4.0 1188\n",
      "4.0 1189\n",
      "5.0 1190\n",
      "5.0 1191\n",
      "5.0 1192\n",
      "5.0 1193\n",
      "4.0 1194\n",
      "4.0 1195\n",
      "1.0 1196\n",
      "5.0 1197\n",
      "2.0 1198\n",
      "5.0 1199\n",
      "4.0 1200\n",
      "4.0 1201\n",
      "5.0 1202\n",
      "1.0 1203\n",
      "4.0 1204\n",
      "5.0 1205\n",
      "5.0 1206\n",
      "1.0 1207\n",
      "5.0 1208\n",
      "1.0 1209\n",
      "5.0 1210\n",
      "5.0 1211\n",
      "5.0 1212\n",
      "5.0 1213\n",
      "5.0 1214\n",
      "5.0 1215\n",
      "1.0 1216\n",
      "5.0 1217\n",
      "1.0 1218\n",
      "5.0 1219\n",
      "5.0 1220\n",
      "5.0 1221\n",
      "5.0 1222\n",
      "5.0 1223\n",
      "4.0 1224\n",
      "5.0 1225\n",
      "5.0 1226\n",
      "5.0 1227\n",
      "5.0 1228\n",
      "3.0 1229\n",
      "4.0 1230\n",
      "5.0 1231\n",
      "5.0 1232\n",
      "4.0 1233\n",
      "5.0 1234\n",
      "1.0 1235\n",
      "5.0 1236\n",
      "5.0 1237\n",
      "5.0 1238\n",
      "5.0 1239\n",
      "5.0 1240\n",
      "5.0 1241\n",
      "5.0 1242\n",
      "5.0 1243\n",
      "3.0 1244\n",
      "5.0 1245\n",
      "4.0 1246\n",
      "5.0 1247\n",
      "5.0 1248\n",
      "5.0 1249\n",
      "5.0 1250\n",
      "5.0 1251\n",
      "5.0 1252\n",
      "3.0 1253\n",
      "5.0 1254\n",
      "4.0 1255\n",
      "5.0 1256\n",
      "5.0 1257\n",
      "5.0 1258\n",
      "1.0 1259\n",
      "2.0 1260\n",
      "5.0 1261\n",
      "5.0 1262\n",
      "3.0 1263\n",
      "5.0 1264\n",
      "5.0 1265\n",
      "4.0 1266\n",
      "5.0 1267\n",
      "5.0 1268\n",
      "5.0 1269\n",
      "1.0 1270\n",
      "5.0 1271\n",
      "5.0 1272\n",
      "4.0 1273\n",
      "1.0 1274\n",
      "5.0 1275\n",
      "5.0 1276\n",
      "5.0 1277\n",
      "5.0 1278\n",
      "5.0 1279\n",
      "5.0 1280\n",
      "5.0 1281\n",
      "5.0 1282\n",
      "4.0 1283\n",
      "1.0 1284\n",
      "5.0 1285\n",
      "5.0 1286\n",
      "5.0 1287\n",
      "5.0 1288\n",
      "5.0 1289\n",
      "5.0 1290\n",
      "5.0 1291\n",
      "5.0 1292\n",
      "4.0 1293\n",
      "1.0 1294\n",
      "1.0 1295\n",
      "1.0 1296\n",
      "5.0 1297\n",
      "5.0 1298\n",
      "4.0 1299\n",
      "5.0 1300\n",
      "5.0 1301\n",
      "1.0 1302\n",
      "5.0 1303\n",
      "5.0 1304\n",
      "5.0 1305\n",
      "5.0 1306\n",
      "1.0 1307\n",
      "5.0 1308\n",
      "5.0 1309\n",
      "4.0 1310\n",
      "5.0 1311\n",
      "4.0 1312\n",
      "4.0 1313\n",
      "5.0 1314\n",
      "2.0 1315\n",
      "3.0 1316\n",
      "5.0 1317\n",
      "5.0 1318\n",
      "5.0 1319\n",
      "5.0 1320\n",
      "4.0 1321\n",
      "4.0 1322\n",
      "5.0 1323\n",
      "2.0 1324\n",
      "3.0 1325\n",
      "5.0 1326\n",
      "5.0 1327\n",
      "5.0 1328\n",
      "5.0 1329\n",
      "5.0 1330\n",
      "5.0 1331\n",
      "5.0 1332\n",
      "5.0 1333\n",
      "5.0 1334\n",
      "5.0 1335\n",
      "5.0 1336\n",
      "1.0 1337\n",
      "1.0 1338\n",
      "5.0 1339\n",
      "1.0 1340\n",
      "5.0 1341\n",
      "4.0 1342\n",
      "5.0 1343\n",
      "5.0 1344\n",
      "5.0 1345\n",
      "5.0 1346\n",
      "5.0 1347\n",
      "5.0 1348\n",
      "5.0 1349\n",
      "5.0 1350\n",
      "5.0 1351\n",
      "5.0 1352\n",
      "5.0 1353\n",
      "5.0 1354\n",
      "5.0 1355\n",
      "5.0 1356\n",
      "5.0 1357\n",
      "5.0 1358\n",
      "5.0 1359\n",
      "5.0 1360\n",
      "5.0 1361\n",
      "5.0 1362\n",
      "5.0 1363\n",
      "4.0 1364\n",
      "4.0 1365\n",
      "5.0 1366\n",
      "5.0 1367\n",
      "5.0 1368\n",
      "3.0 1369\n",
      "4.0 1370\n",
      "4.0 1371\n",
      "5.0 1372\n",
      "4.0 1373\n",
      "2.0 1374\n",
      "5.0 1375\n",
      "4.0 1376\n",
      "5.0 1377\n",
      "5.0 1378\n",
      "3.0 1379\n",
      "3.0 1380\n",
      "2.0 1381\n",
      "5.0 1382\n",
      "4.0 1383\n",
      "1.0 1384\n",
      "5.0 1385\n",
      "5.0 1386\n",
      "1.0 1387\n",
      "1.0 1388\n",
      "1.0 1389\n",
      "1.0 1390\n",
      "1.0 1391\n",
      "5.0 1392\n",
      "5.0 1393\n",
      "1.0 1394\n",
      "1.0 1395\n",
      "1.0 1396\n",
      "1.0 1397\n",
      "1.0 1398\n",
      "5.0 1399\n",
      "5.0 1400\n",
      "1.0 1401\n",
      "2.0 1402\n",
      "4.0 1403\n",
      "1.0 1404\n",
      "5.0 1405\n",
      "4.0 1406\n",
      "4.0 1407\n",
      "4.0 1408\n",
      "5.0 1409\n",
      "4.0 1410\n",
      "3.0 1411\n",
      "5.0 1412\n",
      "4.0 1413\n",
      "4.0 1414\n",
      "4.0 1415\n",
      "5.0 1416\n",
      "4.0 1417\n",
      "3.0 1418\n",
      "5.0 1419\n",
      "5.0 1420\n",
      "5.0 1421\n",
      "5.0 1422\n",
      "5.0 1423\n",
      "5.0 1424\n",
      "5.0 1425\n",
      "5.0 1426\n",
      "2.0 1427\n",
      "4.0 1428\n",
      "4.0 1429\n",
      "5.0 1430\n",
      "5.0 1431\n",
      "5.0 1432\n",
      "4.0 1433\n",
      "5.0 1434\n",
      "1.0 1435\n",
      "5.0 1436\n",
      "2.0 1437\n",
      "4.0 1438\n",
      "4.0 1439\n",
      "5.0 1440\n",
      "5.0 1441\n",
      "5.0 1442\n",
      "4.0 1443\n",
      "5.0 1444\n",
      "1.0 1445\n",
      "1.0 1446\n",
      "4.0 1447\n",
      "5.0 1448\n",
      "4.0 1449\n",
      "5.0 1450\n",
      "5.0 1451\n",
      "1.0 1452\n",
      "5.0 1453\n",
      "4.0 1454\n",
      "5.0 1455\n",
      "4.0 1456\n",
      "4.0 1457\n",
      "4.0 1458\n",
      "5.0 1459\n",
      "5.0 1460\n",
      "5.0 1461\n",
      "3.0 1462\n",
      "1.0 1463\n",
      "5.0 1464\n",
      "1.0 1465\n",
      "4.0 1466\n",
      "1.0 1467\n",
      "5.0 1468\n",
      "5.0 1469\n",
      "5.0 1470\n",
      "4.0 1471\n",
      "5.0 1472\n",
      "4.0 1473\n",
      "5.0 1474\n",
      "5.0 1475\n",
      "3.0 1476\n",
      "3.0 1477\n",
      "4.0 1478\n",
      "1.0 1479\n",
      "5.0 1480\n",
      "3.0 1481\n",
      "4.0 1482\n",
      "4.0 1483\n",
      "1.0 1484\n",
      "1.0 1485\n",
      "1.0 1486\n",
      "4.0 1487\n",
      "4.0 1488\n",
      "5.0 1489\n",
      "2.0 1490\n",
      "5.0 1491\n",
      "4.0 1492\n",
      "1.0 1493\n",
      "1.0 1494\n",
      "1.0 1495\n",
      "4.0 1496\n",
      "4.0 1497\n",
      "5.0 1498\n",
      "2.0 1499\n",
      "5.0 1500\n",
      "5.0 1501\n",
      "5.0 1502\n",
      "5.0 1503\n",
      "5.0 1504\n",
      "5.0 1505\n",
      "5.0 1506\n",
      "5.0 1507\n",
      "1.0 1508\n",
      "4.0 1509\n",
      "5.0 1510\n",
      "5.0 1511\n",
      "4.0 1512\n",
      "4.0 1513\n",
      "5.0 1514\n",
      "5.0 1515\n",
      "1.0 1516\n",
      "4.0 1517\n",
      "5.0 1518\n",
      "5.0 1519\n",
      "4.0 1520\n",
      "4.0 1521\n",
      "5.0 1522\n",
      "5.0 1523\n",
      "5.0 1524\n",
      "3.0 1525\n",
      "5.0 1526\n",
      "5.0 1527\n",
      "3.0 1528\n",
      "5.0 1529\n",
      "4.0 1530\n",
      "5.0 1531\n",
      "5.0 1532\n",
      "5.0 1533\n",
      "5.0 1534\n",
      "4.0 1535\n",
      "1.0 1536\n",
      "5.0 1537\n",
      "5.0 1538\n",
      "5.0 1539\n",
      "5.0 1540\n",
      "4.0 1541\n",
      "5.0 1542\n",
      "1.0 1543\n",
      "4.0 1544\n",
      "5.0 1545\n",
      "1.0 1546\n",
      "5.0 1547\n",
      "5.0 1548\n",
      "5.0 1549\n",
      "4.0 1550\n",
      "5.0 1551\n",
      "5.0 1552\n",
      "5.0 1553\n",
      "5.0 1554\n",
      "4.0 1555\n",
      "2.0 1556\n",
      "4.0 1557\n",
      "5.0 1558\n",
      "2.0 1559\n",
      "4.0 1560\n",
      "2.0 1561\n",
      "4.0 1562\n",
      "5.0 1563\n",
      "2.0 1564\n",
      "5.0 1565\n",
      "3.0 1566\n",
      "5.0 1567\n",
      "4.0 1568\n",
      "5.0 1569\n",
      "5.0 1570\n",
      "4.0 1571\n",
      "5.0 1572\n",
      "5.0 1573\n",
      "5.0 1574\n",
      "5.0 1575\n",
      "1.0 1576\n",
      "5.0 1577\n",
      "5.0 1578\n",
      "5.0 1579\n",
      "4.0 1580\n",
      "5.0 1581\n",
      "4.0 1582\n",
      "5.0 1583\n",
      "4.0 1584\n",
      "4.0 1585\n",
      "5.0 1586\n",
      "4.0 1587\n",
      "4.0 1588\n",
      "4.0 1589\n",
      "4.0 1590\n",
      "2.0 1591\n",
      "4.0 1592\n",
      "1.0 1593\n",
      "1.0 1594\n",
      "5.0 1595\n",
      "1.0 1596\n",
      "1.0 1597\n",
      "4.0 1598\n",
      "4.0 1599\n",
      "4.0 1600\n",
      "5.0 1601\n",
      "1.0 1602\n",
      "4.0 1603\n",
      "5.0 1604\n",
      "4.0 1605\n",
      "5.0 1606\n",
      "5.0 1607\n",
      "5.0 1608\n",
      "5.0 1609\n",
      "5.0 1610\n",
      "4.0 1611\n",
      "1.0 1612\n",
      "4.0 1613\n",
      "5.0 1614\n",
      "4.0 1615\n",
      "5.0 1616\n",
      "5.0 1617\n",
      "3.0 1618\n",
      "5.0 1619\n",
      "5.0 1620\n",
      "5.0 1621\n",
      "5.0 1622\n",
      "5.0 1623\n",
      "1.0 1624\n",
      "1.0 1625\n",
      "3.0 1626\n",
      "5.0 1627\n",
      "5.0 1628\n",
      "4.0 1629\n",
      "3.0 1630\n",
      "5.0 1631\n",
      "5.0 1632\n",
      "5.0 1633\n",
      "5.0 1634\n",
      "5.0 1635\n",
      "5.0 1636\n",
      "2.0 1637\n",
      "2.0 1638\n",
      "4.0 1639\n",
      "5.0 1640\n",
      "5.0 1641\n",
      "5.0 1642\n",
      "5.0 1643\n",
      "5.0 1644\n",
      "5.0 1645\n",
      "5.0 1646\n",
      "5.0 1647\n",
      "5.0 1648\n",
      "5.0 1649\n",
      "5.0 1650\n",
      "5.0 1651\n",
      "5.0 1652\n",
      "5.0 1653\n",
      "5.0 1654\n",
      "5.0 1655\n",
      "3.0 1656\n",
      "3.0 1657\n",
      "2.0 1658\n",
      "5.0 1659\n",
      "5.0 1660\n",
      "5.0 1661\n",
      "5.0 1662\n",
      "5.0 1663\n",
      "5.0 1664\n",
      "3.0 1665\n",
      "3.0 1666\n",
      "2.0 1667\n",
      "5.0 1668\n",
      "5.0 1669\n",
      "5.0 1670\n",
      "5.0 1671\n",
      "5.0 1672\n",
      "5.0 1673\n",
      "5.0 1674\n",
      "5.0 1675\n",
      "5.0 1676\n",
      "5.0 1677\n",
      "1.0 1678\n",
      "4.0 1679\n",
      "5.0 1680\n",
      "5.0 1681\n",
      "5.0 1682\n",
      "2.0 1683\n",
      "5.0 1684\n",
      "5.0 1685\n",
      "5.0 1686\n",
      "5.0 1687\n",
      "5.0 1688\n",
      "5.0 1689\n",
      "5.0 1690\n",
      "5.0 1691\n",
      "4.0 1692\n",
      "5.0 1693\n",
      "5.0 1694\n",
      "5.0 1695\n",
      "5.0 1696\n",
      "5.0 1697\n",
      "5.0 1698\n",
      "5.0 1699\n",
      "5.0 1700\n",
      "5.0 1701\n",
      "4.0 1702\n",
      "5.0 1703\n",
      "5.0 1704\n",
      "5.0 1705\n",
      "5.0 1706\n",
      "5.0 1707\n",
      "1.0 1708\n",
      "5.0 1709\n",
      "5.0 1710\n",
      "5.0 1711\n",
      "5.0 1712\n",
      "5.0 1713\n",
      "5.0 1714\n",
      "5.0 1715\n",
      "5.0 1716\n",
      "5.0 1717\n",
      "1.0 1718\n",
      "5.0 1719\n",
      "5.0 1720\n",
      "4.0 1721\n",
      "1.0 1722\n",
      "5.0 1723\n",
      "5.0 1724\n",
      "5.0 1725\n",
      "5.0 1726\n",
      "5.0 1727\n",
      "1.0 1728\n",
      "5.0 1729\n",
      "5.0 1730\n",
      "4.0 1731\n",
      "1.0 1732\n",
      "5.0 1733\n",
      "5.0 1734\n",
      "5.0 1735\n",
      "5.0 1736\n",
      "5.0 1737\n",
      "4.0 1738\n",
      "4.0 1739\n",
      "2.0 1740\n",
      "5.0 1741\n",
      "5.0 1742\n",
      "5.0 1743\n",
      "5.0 1744\n",
      "5.0 1745\n",
      "5.0 1746\n",
      "5.0 1747\n",
      "1.0 1748\n",
      "5.0 1749\n",
      "5.0 1750\n",
      "5.0 1751\n",
      "5.0 1752\n",
      "5.0 1753\n",
      "5.0 1754\n",
      "5.0 1755\n",
      "5.0 1756\n",
      "5.0 1757\n",
      "1.0 1758\n",
      "5.0 1759\n",
      "5.0 1760\n",
      "5.0 1761\n",
      "5.0 1762\n",
      "5.0 1763\n",
      "5.0 1764\n",
      "5.0 1765\n",
      "5.0 1766\n",
      "5.0 1767\n",
      "1.0 1768\n",
      "5.0 1769\n",
      "5.0 1770\n",
      "5.0 1771\n",
      "5.0 1772\n",
      "5.0 1773\n",
      "5.0 1774\n",
      "5.0 1775\n",
      "5.0 1776\n",
      "5.0 1777\n",
      "2.0 1778\n",
      "5.0 1779\n",
      "5.0 1780\n",
      "5.0 1781\n",
      "5.0 1782\n",
      "5.0 1783\n",
      "5.0 1784\n",
      "4.0 1785\n",
      "5.0 1786\n",
      "5.0 1787\n",
      "5.0 1788\n",
      "5.0 1789\n",
      "5.0 1790\n",
      "1.0 1791\n",
      "5.0 1792\n",
      "5.0 1793\n",
      "4.0 1794\n",
      "1.0 1795\n",
      "2.0 1796\n",
      "1.0 1797\n",
      "1.0 1798\n",
      "1.0 1799\n",
      "1.0 1800\n",
      "5.0 1801\n",
      "5.0 1802\n",
      "5.0 1803\n",
      "2.0 1804\n",
      "1.0 1805\n",
      "5.0 1806\n",
      "5.0 1807\n",
      "5.0 1808\n",
      "4.0 1809\n",
      "5.0 1810\n",
      "5.0 1811\n",
      "5.0 1812\n",
      "5.0 1813\n",
      "3.0 1814\n",
      "5.0 1815\n",
      "5.0 1816\n",
      "5.0 1817\n",
      "5.0 1818\n",
      "5.0 1819\n",
      "5.0 1820\n",
      "5.0 1821\n",
      "5.0 1822\n",
      "4.0 1823\n",
      "5.0 1824\n",
      "5.0 1825\n",
      "5.0 1826\n",
      "5.0 1827\n",
      "1.0 1828\n",
      "4.0 1829\n",
      "5.0 1830\n",
      "5.0 1831\n",
      "5.0 1832\n",
      "5.0 1833\n",
      "5.0 1834\n",
      "3.0 1835\n",
      "5.0 1836\n",
      "5.0 1837\n",
      "5.0 1838\n",
      "5.0 1839\n",
      "5.0 1840\n",
      "5.0 1841\n",
      "5.0 1842\n",
      "5.0 1843\n",
      "5.0 1844\n",
      "5.0 1845\n",
      "4.0 1846\n",
      "5.0 1847\n",
      "5.0 1848\n",
      "5.0 1849\n",
      "5.0 1850\n",
      "5.0 1851\n",
      "5.0 1852\n",
      "5.0 1853\n",
      "4.0 1854\n",
      "5.0 1855\n",
      "5.0 1856\n",
      "5.0 1857\n",
      "5.0 1858\n",
      "5.0 1859\n",
      "5.0 1860\n",
      "4.0 1861\n",
      "5.0 1862\n",
      "5.0 1863\n",
      "4.0 1864\n",
      "4.5 1865\n",
      "4.0 1866\n",
      "4.0 1867\n",
      "5.0 1868\n",
      "5.0 1869\n",
      "4.0 1870\n",
      "5.0 1871\n",
      "5.0 1872\n",
      "5.0 1873\n",
      "5.0 1874\n",
      "1.0 1875\n",
      "5.0 1876\n",
      "5.0 1877\n",
      "5.0 1878\n",
      "5.0 1879\n",
      "5.0 1880\n",
      "4.0 1881\n",
      "4.0 1882\n",
      "4.0 1883\n",
      "5.0 1884\n",
      "4.0 1885\n",
      "5.0 1886\n",
      "5.0 1887\n",
      "4.0 1888\n",
      "5.0 1889\n",
      "5.0 1890\n",
      "5.0 1891\n",
      "5.0 1892\n",
      "5.0 1893\n",
      "5.0 1894\n",
      "5.0 1895\n",
      "5.0 1896\n",
      "5.0 1897\n",
      "5.0 1898\n",
      "5.0 1899\n",
      "5.0 1900\n",
      "3.0 1901\n",
      "5.0 1902\n",
      "5.0 1903\n",
      "5.0 1904\n",
      "5.0 1905\n",
      "5.0 1906\n",
      "5.0 1907\n",
      "5.0 1908\n",
      "5.0 1909\n",
      "5.0 1910\n",
      "3.0 1911\n",
      "5.0 1912\n",
      "5.0 1913\n",
      "5.0 1914\n",
      "5.0 1915\n",
      "5.0 1916\n",
      "5.0 1917\n",
      "5.0 1918\n",
      "4.0 1919\n",
      "5.0 1920\n",
      "5.0 1921\n",
      "2.0 1922\n",
      "5.0 1923\n",
      "1.0 1924\n",
      "5.0 1925\n",
      "1.0 1926\n",
      "4.0 1927\n",
      "4.0 1928\n",
      "5.0 1929\n",
      "4.0 1930\n",
      "5.0 1931\n",
      "5.0 1932\n",
      "5.0 1933\n",
      "5.0 1934\n",
      "5.0 1935\n",
      "5.0 1936\n",
      "5.0 1937\n",
      "5.0 1938\n",
      "1.0 1939\n",
      "5.0 1940\n",
      "5.0 1941\n",
      "5.0 1942\n",
      "5.0 1943\n",
      "5.0 1944\n",
      "5.0 1945\n",
      "4.0 1946\n",
      "5.0 1947\n",
      "4.0 1948\n",
      "5.0 1949\n",
      "5.0 1950\n",
      "5.0 1951\n",
      "5.0 1952\n",
      "5.0 1953\n",
      "1.0 1954\n",
      "5.0 1955\n",
      "4.0 1956\n",
      "5.0 1957\n",
      "4.0 1958\n",
      "5.0 1959\n",
      "5.0 1960\n",
      "5.0 1961\n",
      "5.0 1962\n",
      "5.0 1963\n",
      "1.0 1964\n",
      "5.0 1965\n",
      "5.0 1966\n",
      "2.0 1967\n",
      "1.0 1968\n",
      "5.0 1969\n",
      "5.0 1970\n",
      "5.0 1971\n",
      "5.0 1972\n",
      "4.0 1973\n",
      "5.0 1974\n",
      "5.0 1975\n",
      "5.0 1976\n",
      "5.0 1977\n",
      "5.0 1978\n",
      "5.0 1979\n",
      "5.0 1980\n",
      "5.0 1981\n",
      "5.0 1982\n",
      "5.0 1983\n",
      "5.0 1984\n",
      "5.0 1985\n",
      "5.0 1986\n",
      "4.0 1987\n",
      "5.0 1988\n",
      "5.0 1989\n",
      "5.0 1990\n",
      "5.0 1991\n",
      "5.0 1992\n",
      "5.0 1993\n",
      "5.0 1994\n",
      "5.0 1995\n",
      "5.0 1996\n",
      "4.0 1997\n",
      "5.0 1998\n",
      "5.0 1999\n",
      "5.0 2000\n",
      "5.0 2001\n",
      "5.0 2002\n",
      "5.0 2003\n",
      "5.0 2004\n",
      "5.0 2005\n",
      "5.0 2006\n",
      "5.0 2007\n",
      "5.0 2008\n",
      "5.0 2009\n",
      "1.0 2010\n",
      "4.0 2011\n",
      "5.0 2012\n",
      "5.0 2013\n",
      "5.0 2014\n",
      "5.0 2015\n",
      "5.0 2016\n",
      "5.0 2017\n",
      "4.0 2018\n",
      "5.0 2019\n",
      "3.0 2020\n",
      "4.0 2021\n",
      "5.0 2022\n",
      "5.0 2023\n",
      "5.0 2024\n",
      "5.0 2025\n",
      "5.0 2026\n",
      "2.0 2027\n",
      "5.0 2028\n",
      "5.0 2029\n",
      "5.0 2030\n",
      "5.0 2031\n",
      "5.0 2032\n",
      "5.0 2033\n",
      "5.0 2034\n",
      "5.0 2035\n",
      "5.0 2036\n",
      "5.0 2037\n",
      "5.0 2038\n",
      "5.0 2039\n",
      "5.0 2040\n",
      "5.0 2041\n",
      "5.0 2042\n",
      "4.0 2043\n",
      "5.0 2044\n",
      "1.0 2045\n",
      "5.0 2046\n",
      "5.0 2047\n",
      "4.0 2048\n",
      "5.0 2049\n",
      "5.0 2050\n",
      "5.0 2051\n",
      "5.0 2052\n",
      "4.0 2053\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4.0 2054\n",
      "4.0 2055\n",
      "5.0 2056\n",
      "5.0 2057\n",
      "5.0 2058\n",
      "5.0 2059\n",
      "5.0 2060\n",
      "5.0 2061\n",
      "4.0 2062\n",
      "4.0 2063\n",
      "4.0 2064\n",
      "5.0 2065\n",
      "5.0 2066\n",
      "5.0 2067\n",
      "5.0 2068\n",
      "5.0 2069\n",
      "3.0 2070\n",
      "3.0 2071\n",
      "4.0 2072\n",
      "5.0 2073\n",
      "4.0 2074\n",
      "4.0 2075\n",
      "1.0 2076\n",
      "4.0 2077\n",
      "5.0 2078\n",
      "4.0 2079\n",
      "5.0 2080\n",
      "5.0 2081\n",
      "5.0 2082\n",
      "4.0 2083\n",
      "5.0 2084\n",
      "5.0 2085\n",
      "4.0 2086\n",
      "5.0 2087\n",
      "5.0 2088\n",
      "5.0 2089\n",
      "5.0 2090\n",
      "4.0 2091\n",
      "5.0 2092\n",
      "5.0 2093\n",
      "4.0 2094\n",
      "5.0 2095\n",
      "5.0 2096\n",
      "5.0 2097\n",
      "5.0 2098\n",
      "5.0 2099\n",
      "5.0 2100\n",
      "5.0 2101\n",
      "4.0 2102\n",
      "5.0 2103\n",
      "5.0 2104\n",
      "5.0 2105\n",
      "5.0 2106\n",
      "5.0 2107\n",
      "5.0 2108\n",
      "4.0 2109\n",
      "5.0 2110\n",
      "5.0 2111\n",
      "5.0 2112\n",
      "5.0 2113\n",
      "5.0 2114\n",
      "5.0 2115\n",
      "4.0 2116\n",
      "5.0 2117\n",
      "5.0 2118\n",
      "5.0 2119\n",
      "3.0 2120\n",
      "5.0 2121\n",
      "5.0 2122\n",
      "5.0 2123\n",
      "5.0 2124\n",
      "5.0 2125\n",
      "5.0 2126\n",
      "5.0 2127\n",
      "5.0 2128\n",
      "5.0 2129\n",
      "3.0 2130\n",
      "5.0 2131\n",
      "5.0 2132\n",
      "5.0 2133\n",
      "5.0 2134\n",
      "5.0 2135\n",
      "5.0 2136\n",
      "3.0 2137\n",
      "2.0 2138\n",
      "5.0 2139\n",
      "5.0 2140\n",
      "5.0 2141\n",
      "1.0 2142\n",
      "1.0 2143\n",
      "1.0 2144\n",
      "5.0 2145\n",
      "5.0 2146\n",
      "3.0 2147\n",
      "2.0 2148\n",
      "5.0 2149\n",
      "5.0 2150\n",
      "5.0 2151\n",
      "1.0 2152\n",
      "1.0 2153\n",
      "1.0 2154\n",
      "5.0 2155\n",
      "5.0 2156\n",
      "4.0 2157\n",
      "5.0 2158\n",
      "5.0 2159\n",
      "5.0 2160\n",
      "5.0 2161\n",
      "5.0 2162\n",
      "5.0 2163\n",
      "4.0 2164\n",
      "3.0 2165\n",
      "5.0 2166\n",
      "5.0 2167\n",
      "5.0 2168\n",
      "5.0 2169\n",
      "5.0 2170\n",
      "5.0 2171\n",
      "5.0 2172\n",
      "5.0 2173\n",
      "4.0 2174\n",
      "4.0 2175\n",
      "5.0 2176\n",
      "2.0 2177\n",
      "5.0 2178\n",
      "5.0 2179\n",
      "4.0 2180\n",
      "5.0 2181\n",
      "5.0 2182\n",
      "3.0 2183\n",
      "1.0 2184\n",
      "5.0 2185\n",
      "4.0 2186\n",
      "3.0 2187\n",
      "5.0 2188\n",
      "5.0 2189\n",
      "5.0 2190\n",
      "5.0 2191\n",
      "5.0 2192\n",
      "5.0 2193\n",
      "1.0 2194\n",
      "5.0 2195\n",
      "4.0 2196\n",
      "3.0 2197\n",
      "5.0 2198\n",
      "5.0 2199\n",
      "5.0 2200\n",
      "5.0 2201\n",
      "5.0 2202\n",
      "5.0 2203\n",
      "4.0 2204\n",
      "4.0 2205\n",
      "3.0 2206\n",
      "5.0 2207\n",
      "4.0 2208\n",
      "4.0 2209\n",
      "5.0 2210\n",
      "4.0 2211\n",
      "5.0 2212\n",
      "4.0 2213\n",
      "5.0 2214\n",
      "5.0 2215\n",
      "4.0 2216\n",
      "5.0 2217\n",
      "4.0 2218\n",
      "5.0 2219\n",
      "5.0 2220\n",
      "4.0 2221\n",
      "5.0 2222\n",
      "5.0 2223\n",
      "4.0 2224\n",
      "4.0 2225\n",
      "5.0 2226\n",
      "4.0 2227\n",
      "5.0 2228\n",
      "1.0 2229\n",
      "5.0 2230\n",
      "5.0 2231\n",
      "5.0 2232\n",
      "1.0 2233\n",
      "3.0 2234\n",
      "4.0 2235\n",
      "5.0 2236\n",
      "5.0 2237\n",
      "5.0 2238\n",
      "5.0 2239\n",
      "5.0 2240\n",
      "5.0 2241\n",
      "5.0 2242\n",
      "5.0 2243\n",
      "3.0 2244\n",
      "5.0 2245\n",
      "4.0 2246\n",
      "5.0 2247\n",
      "5.0 2248\n",
      "5.0 2249\n",
      "5.0 2250\n",
      "5.0 2251\n",
      "5.0 2252\n",
      "3.0 2253\n",
      "5.0 2254\n",
      "4.0 2255\n",
      "3.0 2256\n",
      "4.0 2257\n",
      "4.0 2258\n",
      "5.0 2259\n",
      "5.0 2260\n",
      "5.0 2261\n",
      "3.0 2262\n",
      "5.0 2263\n",
      "5.0 2264\n",
      "5.0 2265\n",
      "5.0 2266\n",
      "4.0 2267\n",
      "5.0 2268\n",
      "5.0 2269\n",
      "5.0 2270\n",
      "5.0 2271\n",
      "5.0 2272\n",
      "5.0 2273\n",
      "5.0 2274\n",
      "5.0 2275\n",
      "5.0 2276\n",
      "5.0 2277\n",
      "5.0 2278\n",
      "5.0 2279\n",
      "5.0 2280\n",
      "5.0 2281\n",
      "5.0 2282\n",
      "4.0 2283\n",
      "5.0 2284\n",
      "5.0 2285\n",
      "5.0 2286\n",
      "5.0 2287\n",
      "4.0 2288\n",
      "3.0 2289\n",
      "1.0 2290\n",
      "5.0 2291\n",
      "5.0 2292\n",
      "5.0 2293\n",
      "4.0 2294\n",
      "5.0 2295\n",
      "5.0 2296\n",
      "1.0 2297\n",
      "5.0 2298\n",
      "4.0 2299\n",
      "5.0 2300\n",
      "5.0 2301\n",
      "2.0 2302\n",
      "5.0 2303\n",
      "1.0 2304\n",
      "2.0 2305\n",
      "5.0 2306\n",
      "4.0 2307\n",
      "3.0 2308\n",
      "5.0 2309\n",
      "3.0 2310\n",
      "4.0 2311\n",
      "5.0 2312\n",
      "2.0 2313\n",
      "4.0 2314\n",
      "4.0 2315\n",
      "4.0 2316\n",
      "5.0 2317\n",
      "3.0 2318\n",
      "5.0 2319\n",
      "4.0 2320\n",
      "5.0 2321\n",
      "4.0 2322\n",
      "1.0 2323\n",
      "5.0 2324\n",
      "3.0 2325\n",
      "5.0 2326\n",
      "4.0 2327\n",
      "4.0 2328\n",
      "5.0 2329\n",
      "5.0 2330\n",
      "5.0 2331\n",
      "5.0 2332\n",
      "5.0 2333\n",
      "4.0 2334\n",
      "5.0 2335\n",
      "5.0 2336\n",
      "5.0 2337\n",
      "5.0 2338\n",
      "5.0 2339\n",
      "5.0 2340\n",
      "5.0 2341\n",
      "5.0 2342\n",
      "5.0 2343\n",
      "5.0 2344\n",
      "5.0 2345\n",
      "5.0 2346\n",
      "5.0 2347\n",
      "1.0 2348\n",
      "5.0 2349\n",
      "5.0 2350\n",
      "5.0 2351\n",
      "5.0 2352\n",
      "5.0 2353\n",
      "5.0 2354\n",
      "4.0 2355\n",
      "5.0 2356\n",
      "1.0 2357\n",
      "5.0 2358\n",
      "5.0 2359\n",
      "5.0 2360\n",
      "5.0 2361\n",
      "5.0 2362\n",
      "5.0 2363\n",
      "4.0 2364\n",
      "5.0 2365\n",
      "3.0 2366\n",
      "5.0 2367\n",
      "3.0 2368\n",
      "5.0 2369\n",
      "5.0 2370\n",
      "3.0 2371\n",
      "5.0 2372\n",
      "5.0 2373\n",
      "5.0 2374\n",
      "5.0 2375\n",
      "1.0 2376\n",
      "5.0 2377\n",
      "4.0 2378\n",
      "1.0 2379\n",
      "4.0 2380\n",
      "5.0 2381\n",
      "3.0 2382\n",
      "5.0 2383\n",
      "4.0 2384\n",
      "5.0 2385\n",
      "3.0 2386\n",
      "4.0 2387\n",
      "4.0 2388\n",
      "5.0 2389\n",
      "5.0 2390\n",
      "5.0 2391\n",
      "4.0 2392\n",
      "5.0 2393\n",
      "5.0 2394\n",
      "4.0 2395\n",
      "5.0 2396\n",
      "3.0 2397\n",
      "5.0 2398\n",
      "5.0 2399\n",
      "5.0 2400\n",
      "5.0 2401\n",
      "5.0 2402\n",
      "5.0 2403\n",
      "5.0 2404\n",
      "4.0 2405\n",
      "4.0 2406\n",
      "2.0 2407\n",
      "5.0 2408\n",
      "4.0 2409\n",
      "4.0 2410\n",
      "4.0 2411\n",
      "4.0 2412\n",
      "1.0 2413\n",
      "1.0 2414\n",
      "1.0 2415\n",
      "5.0 2416\n",
      "4.0 2417\n",
      "4.0 2418\n",
      "4.0 2419\n",
      "5.0 2420\n",
      "5.0 2421\n",
      "5.0 2422\n",
      "5.0 2423\n",
      "5.0 2424\n",
      "5.0 2425\n",
      "5.0 2426\n",
      "5.0 2427\n",
      "5.0 2428\n",
      "5.0 2429\n",
      "5.0 2430\n",
      "5.0 2431\n",
      "3.0 2432\n",
      "5.0 2433\n",
      "3.0 2434\n",
      "4.0 2435\n",
      "5.0 2436\n",
      "4.0 2437\n",
      "5.0 2438\n",
      "5.0 2439\n",
      "3.0 2440\n",
      "3.0 2441\n",
      "1.0 2442\n",
      "4.0 2443\n",
      "5.0 2444\n",
      "5.0 2445\n",
      "5.0 2446\n",
      "5.0 2447\n",
      "5.0 2448\n",
      "5.0 2449\n",
      "5.0 2450\n",
      "5.0 2451\n",
      "5.0 2452\n",
      "5.0 2453\n",
      "5.0 2454\n",
      "5.0 2455\n",
      "5.0 2456\n",
      "5.0 2457\n",
      "5.0 2458\n",
      "5.0 2459\n",
      "5.0 2460\n",
      "5.0 2461\n",
      "5.0 2462\n",
      "5.0 2463\n",
      "5.0 2464\n",
      "4.0 2465\n",
      "1.0 2466\n",
      "4.0 2467\n",
      "5.0 2468\n",
      "5.0 2469\n",
      "5.0 2470\n",
      "5.0 2471\n",
      "1.0 2472\n",
      "5.0 2473\n",
      "5.0 2474\n",
      "5.0 2475\n",
      "5.0 2476\n",
      "5.0 2477\n",
      "5.0 2478\n",
      "5.0 2479\n",
      "5.0 2480\n",
      "3.0 2481\n",
      "5.0 2482\n",
      "4.0 2483\n",
      "3.0 2484\n",
      "4.0 2485\n",
      "5.0 2486\n",
      "3.0 2487\n",
      "5.0 2488\n",
      "4.0 2489\n",
      "3.0 2490\n",
      "4.0 2491\n",
      "5.0 2492\n",
      "4.0 2493\n",
      "1.0 2494\n",
      "5.0 2495\n",
      "2.0 2496\n",
      "1.0 2497\n",
      "4.0 2498\n",
      "5.0 2499\n",
      "5.0 2500\n",
      "5.0 2501\n",
      "5.0 2502\n",
      "5.0 2503\n",
      "4.0 2504\n",
      "5.0 2505\n",
      "5.0 2506\n",
      "5.0 2507\n",
      "5.0 2508\n",
      "5.0 2509\n",
      "1.0 2510\n",
      "5.0 2511\n",
      "5.0 2512\n",
      "5.0 2513\n",
      "5.0 2514\n",
      "2.0 2515\n",
      "5.0 2516\n",
      "3.0 2517\n",
      "4.0 2518\n",
      "1.0 2519\n",
      "2.0 2520\n",
      "5.0 2521\n",
      "3.0 2522\n",
      "4.0 2523\n",
      "1.0 2524\n",
      "5.0 2525\n",
      "4.0 2526\n",
      "5.0 2527\n",
      "5.0 2528\n",
      "5.0 2529\n",
      "5.0 2530\n",
      "5.0 2531\n",
      "1.0 2532\n",
      "5.0 2533\n",
      "4.0 2534\n",
      "5.0 2535\n",
      "4.0 2536\n",
      "5.0 2537\n",
      "4.0 2538\n",
      "3.0 2539\n",
      "1.0 2540\n",
      "2.0 2541\n",
      "1.0 2542\n",
      "5.0 2543\n",
      "2.0 2544\n",
      "4.0 2545\n",
      "5.0 2546\n",
      "5.0 2547\n",
      "5.0 2548\n",
      "5.0 2549\n",
      "5.0 2550\n",
      "5.0 2551\n",
      "5.0 2552\n",
      "5.0 2553\n",
      "5.0 2554\n",
      "4.0 2555\n",
      "5.0 2556\n",
      "5.0 2557\n",
      "5.0 2558\n",
      "5.0 2559\n",
      "5.0 2560\n",
      "5.0 2561\n",
      "5.0 2562\n",
      "5.0 2563\n",
      "5.0 2564\n",
      "5.0 2565\n",
      "1.0 2566\n",
      "5.0 2567\n",
      "1.0 2568\n",
      "3.0 2569\n",
      "4.0 2570\n",
      "4.0 2571\n",
      "5.0 2572\n",
      "5.0 2573\n",
      "4.0 2574\n",
      "4.0 2575\n",
      "5.0 2576\n",
      "4.0 2577\n",
      "4.0 2578\n",
      "5.0 2579\n",
      "5.0 2580\n",
      "4.0 2581\n",
      "4.0 2582\n",
      "5.0 2583\n",
      "5.0 2584\n",
      "5.0 2585\n",
      "5.0 2586\n",
      "5.0 2587\n",
      "5.0 2588\n",
      "5.0 2589\n",
      "4.0 2590\n",
      "5.0 2591\n",
      "5.0 2592\n",
      "5.0 2593\n",
      "4.0 2594\n",
      "5.0 2595\n",
      "5.0 2596\n",
      "5.0 2597\n",
      "5.0 2598\n",
      "1.0 2599\n",
      "3.0 2600\n",
      "5.0 2601\n",
      "5.0 2602\n",
      "1.0 2603\n",
      "4.0 2604\n",
      "5.0 2605\n",
      "5.0 2606\n",
      "5.0 2607\n",
      "5.0 2608\n",
      "1.0 2609\n",
      "3.0 2610\n",
      "5.0 2611\n",
      "5.0 2612\n",
      "1.0 2613\n",
      "5.0 2614\n",
      "5.0 2615\n",
      "4.0 2616\n",
      "4.0 2617\n",
      "4.0 2618\n",
      "5.0 2619\n",
      "1.0 2620\n",
      "1.0 2621\n",
      "5.0 2622\n",
      "4.0 2623\n",
      "5.0 2624\n",
      "5.0 2625\n",
      "5.0 2626\n",
      "5.0 2627\n",
      "5.0 2628\n",
      "5.0 2629\n",
      "5.0 2630\n",
      "5.0 2631\n",
      "5.0 2632\n",
      "2.0 2633\n",
      "5.0 2634\n",
      "5.0 2635\n",
      "4.0 2636\n",
      "5.0 2637\n",
      "4.0 2638\n",
      "1.0 2639\n",
      "5.0 2640\n",
      "5.0 2641\n",
      "1.0 2642\n",
      "1.0 2643\n",
      "1.0 2644\n",
      "1.0 2645\n",
      "1.0 2646\n",
      "3.0 2647\n",
      "5.0 2648\n",
      "5.0 2649\n",
      "5.0 2650\n",
      "5.0 2651\n",
      "5.0 2652\n",
      "4.0 2653\n",
      "4.0 2654\n",
      "4.0 2655\n",
      "1.0 2656\n",
      "3.0 2657\n",
      "2.0 2658\n",
      "5.0 2659\n",
      "5.0 2660\n",
      "5.0 2661\n",
      "5.0 2662\n",
      "4.0 2663\n",
      "4.0 2664\n",
      "4.0 2665\n",
      "5.0 2666\n",
      "5.0 2667\n",
      "5.0 2668\n",
      "5.0 2669\n",
      "4.0 2670\n",
      "5.0 2671\n",
      "5.0 2672\n",
      "5.0 2673\n",
      "5.0 2674\n",
      "5.0 2675\n",
      "5.0 2676\n",
      "5.0 2677\n",
      "4.0 2678\n",
      "1.0 2679\n",
      "4.0 2680\n",
      "5.0 2681\n",
      "4.0 2682\n",
      "1.0 2683\n",
      "4.0 2684\n",
      "4.0 2685\n",
      "5.0 2686\n",
      "1.0 2687\n",
      "4.0 2688\n",
      "4.0 2689\n",
      "5.0 2690\n",
      "4.0 2691\n",
      "5.0 2692\n",
      "5.0 2693\n",
      "5.0 2694\n",
      "5.0 2695\n",
      "5.0 2696\n",
      "5.0 2697\n",
      "1.0 2698\n",
      "1.0 2699\n",
      "1.0 2700\n",
      "5.0 2701\n",
      "2.0 2702\n",
      "3.0 2703\n",
      "5.0 2704\n",
      "5.0 2705\n",
      "5.0 2706\n",
      "5.0 2707\n",
      "5.0 2708\n",
      "5.0 2709\n",
      "5.0 2710\n",
      "4.0 2711\n",
      "5.0 2712\n",
      "5.0 2713\n",
      "5.0 2714\n",
      "5.0 2715\n",
      "4.0 2716\n",
      "4.0 2717\n",
      "5.0 2718\n",
      "5.0 2719\n",
      "5.0 2720\n",
      "5.0 2721\n",
      "5.0 2722\n",
      "3.0 2723\n",
      "5.0 2724\n",
      "5.0 2725\n",
      "5.0 2726\n",
      "5.0 2727\n",
      "5.0 2728\n",
      "5.0 2729\n",
      "5.0 2730\n",
      "5.0 2731\n",
      "5.0 2732\n",
      "5.0 2733\n",
      "5.0 2734\n",
      "5.0 2735\n",
      "5.0 2736\n",
      "5.0 2737\n",
      "4.0 2738\n",
      "5.0 2739\n",
      "5.0 2740\n",
      "5.0 2741\n",
      "5.0 2742\n",
      "5.0 2743\n",
      "3.0 2744\n",
      "5.0 2745\n",
      "5.0 2746\n",
      "1.0 2747\n",
      "4.0 2748\n",
      "5.0 2749\n",
      "5.0 2750\n",
      "5.0 2751\n",
      "5.0 2752\n",
      "1.0 2753\n",
      "4.0 2754\n",
      "5.0 2755\n",
      "5.0 2756\n",
      "5.0 2757\n",
      "4.0 2758\n",
      "5.0 2759\n",
      "5.0 2760\n",
      "5.0 2761\n",
      "5.0 2762\n",
      "5.0 2763\n",
      "4.0 2764\n",
      "4.0 2765\n",
      "4.0 2766\n",
      "5.0 2767\n",
      "4.0 2768\n",
      "2.0 2769\n",
      "4.0 2770\n",
      "4.0 2771\n",
      "2.0 2772\n",
      "4.0 2773\n",
      "5.0 2774\n",
      "4.0 2775\n",
      "1.0 2776\n",
      "4.0 2777\n",
      "5.0 2778\n",
      "5.0 2779\n",
      "5.0 2780\n",
      "5.0 2781\n",
      "5.0 2782\n",
      "5.0 2783\n",
      "5.0 2784\n",
      "5.0 2785\n",
      "5.0 2786\n",
      "5.0 2787\n",
      "4.0 2788\n",
      "3.0 2789\n",
      "5.0 2790\n",
      "4.0 2791\n",
      "1.0 2792\n",
      "5.0 2793\n",
      "5.0 2794\n",
      "5.0 2795\n",
      "5.0 2796\n",
      "4.0 2797\n",
      "5.0 2798\n",
      "5.0 2799\n",
      "5.0 2800\n",
      "5.0 2801\n",
      "5.0 2802\n",
      "5.0 2803\n",
      "5.0 2804\n",
      "5.0 2805\n",
      "4.0 2806\n",
      "5.0 2807\n",
      "5.0 2808\n",
      "1.0 2809\n",
      "5.0 2810\n",
      "5.0 2811\n",
      "2.0 2812\n",
      "5.0 2813\n",
      "5.0 2814\n",
      "4.0 2815\n",
      "5.0 2816\n",
      "5.0 2817\n",
      "5.0 2818\n",
      "5.0 2819\n",
      "5.0 2820\n",
      "5.0 2821\n",
      "4.0 2822\n",
      "2.0 2823\n",
      "2.0 2824\n",
      "5.0 2825\n",
      "2.0 2826\n",
      "1.0 2827\n",
      "5.0 2828\n",
      "2.0 2829\n",
      "5.0 2830\n",
      "5.0 2831\n",
      "1.0 2832\n",
      "5.0 2833\n",
      "5.0 2834\n",
      "5.0 2835\n",
      "5.0 2836\n",
      "5.0 2837\n",
      "5.0 2838\n",
      "5.0 2839\n",
      "5.0 2840\n",
      "5.0 2841\n",
      "5.0 2842\n",
      "4.0 2843\n",
      "4.0 2844\n",
      "3.0 2845\n",
      "4.0 2846\n",
      "4.0 2847\n",
      "5.0 2848\n",
      "2.0 2849\n",
      "1.0 2850\n",
      "5.0 2851\n",
      "5.0 2852\n",
      "3.0 2853\n",
      "5.0 2854\n",
      "4.0 2855\n",
      "5.0 2856\n",
      "3.0 2857\n",
      "5.0 2858\n",
      "5.0 2859\n",
      "5.0 2860\n",
      "5.0 2861\n",
      "4.0 2862\n",
      "5.0 2863\n",
      "4.0 2864\n",
      "3.0 2865\n",
      "5.0 2866\n",
      "4.0 2867\n",
      "5.0 2868\n",
      "5.0 2869\n",
      "5.0 2870\n",
      "5.0 2871\n",
      "5.0 2872\n",
      "5.0 2873\n",
      "5.0 2874\n",
      "4.0 2875\n",
      "5.0 2876\n",
      "5.0 2877\n",
      "5.0 2878\n",
      "5.0 2879\n",
      "5.0 2880\n",
      "5.0 2881\n",
      "5.0 2882\n",
      "4.0 2883\n",
      "5.0 2884\n",
      "5.0 2885\n",
      "5.0 2886\n",
      "1.0 2887\n",
      "2.0 2888\n",
      "5.0 2889\n",
      "5.0 2890\n",
      "4.0 2891\n",
      "5.0 2892\n",
      "3.0 2893\n",
      "5.0 2894\n",
      "5.0 2895\n",
      "2.0 2896\n",
      "4.0 2897\n",
      "5.0 2898\n",
      "5.0 2899\n",
      "5.0 2900\n",
      "3.0 2901\n",
      "5.0 2902\n",
      "5.0 2903\n",
      "5.0 2904\n",
      "5.0 2905\n",
      "5.0 2906\n",
      "1.0 2907\n",
      "4.0 2908\n",
      "4.0 2909\n",
      "4.0 2910\n",
      "3.0 2911\n",
      "3.0 2912\n",
      "5.0 2913\n",
      "5.0 2914\n",
      "5.0 2915\n",
      "5.0 2916\n",
      "5.0 2917\n",
      "1.0 2918\n",
      "4.0 2919\n",
      "5.0 2920\n",
      "4.0 2921\n",
      "3.0 2922\n",
      "4.0 2923\n",
      "5.0 2924\n",
      "4.0 2925\n",
      "1.0 2926\n",
      "4.0 2927\n",
      "1.0 2928\n",
      "1.0 2929\n",
      "5.0 2930\n",
      "5.0 2931\n",
      "5.0 2932\n",
      "4.0 2933\n",
      "4.0 2934\n",
      "5.0 2935\n",
      "4.0 2936\n",
      "4.0 2937\n",
      "3.0 2938\n",
      "5.0 2939\n",
      "1.0 2940\n",
      "3.0 2941\n",
      "4.0 2942\n",
      "5.0 2943\n",
      "5.0 2944\n",
      "1.0 2945\n",
      "4.0 2946\n",
      "5.0 2947\n",
      "4.0 2948\n",
      "5.0 2949\n",
      "5.0 2950\n",
      "5.0 2951\n",
      "5.0 2952\n",
      "1.0 2953\n",
      "4.0 2954\n",
      "4.0 2955\n",
      "5.0 2956\n",
      "4.0 2957\n",
      "1.0 2958\n",
      "4.0 2959\n",
      "4.0 2960\n",
      "5.0 2961\n",
      "5.0 2962\n",
      "5.0 2963\n",
      "5.0 2964\n",
      "4.0 2965\n",
      "5.0 2966\n",
      "5.0 2967\n",
      "2.0 2968\n",
      "5.0 2969\n",
      "5.0 2970\n",
      "5.0 2971\n",
      "5.0 2972\n",
      "4.0 2973\n",
      "5.0 2974\n",
      "5.0 2975\n",
      "5.0 2976\n",
      "5.0 2977\n",
      "1.0 2978\n",
      "2.0 2979\n",
      "1.0 2980\n",
      "1.0 2981\n",
      "1.0 2982\n",
      "4.0 2983\n",
      "4.0 2984\n",
      "5.0 2985\n",
      "4.0 2986\n",
      "4.0 2987\n",
      "5.0 2988\n",
      "5.0 2989\n",
      "4.0 2990\n",
      "4.0 2991\n",
      "5.0 2992\n",
      "5.0 2993\n",
      "2.0 2994\n",
      "1.0 2995\n",
      "2.0 2996\n",
      "4.0 2997\n",
      "2.0 2998\n",
      "3.0 2999\n",
      "5.0 3000\n",
      "5.0 3001\n",
      "5.0 3002\n",
      "3.0 3003\n",
      "5.0 3004\n",
      "5.0 3005\n",
      "5.0 3006\n",
      "5.0 3007\n",
      "4.0 3008\n",
      "3.0 3009\n",
      "5.0 3010\n",
      "5.0 3011\n",
      "4.0 3012\n",
      "4.0 3013\n",
      "5.0 3014\n",
      "5.0 3015\n",
      "4.0 3016\n",
      "5.0 3017\n",
      "5.0 3018\n",
      "4.0 3019\n",
      "5.0 3020\n",
      "1.0 3021\n",
      "4.0 3022\n",
      "5.0 3023\n",
      "2.0 3024\n",
      "5.0 3025\n",
      "4.0 3026\n",
      "5.0 3027\n",
      "4.0 3028\n",
      "5.0 3029\n",
      "5.0 3030\n",
      "5.0 3031\n",
      "3.0 3032\n",
      "1.0 3033\n",
      "5.0 3034\n",
      "4.0 3035\n",
      "4.0 3036\n",
      "4.0 3037\n",
      "5.0 3038\n",
      "4.0 3039\n",
      "1.0 3040\n",
      "4.0 3041\n",
      "1.0 3042\n",
      "4.0 3043\n",
      "5.0 3044\n",
      "4.0 3045\n",
      "5.0 3046\n",
      "4.0 3047\n",
      "4.0 3048\n",
      "4.0 3049\n",
      "4.0 3050\n",
      "3.0 3051\n",
      "3.0 3052\n",
      "3.0 3053\n",
      "4.0 3054\n",
      "5.0 3055\n",
      "5.0 3056\n",
      "5.0 3057\n",
      "5.0 3058\n",
      "4.0 3059\n",
      "5.0 3060\n",
      "5.0 3061\n",
      "5.0 3062\n",
      "3.0 3063\n",
      "4.0 3064\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5.0 3065\n",
      "5.0 3066\n",
      "5.0 3067\n",
      "1.0 3068\n",
      "5.0 3069\n",
      "1.0 3070\n",
      "5.0 3071\n",
      "4.0 3072\n",
      "2.0 3073\n",
      "5.0 3074\n",
      "5.0 3075\n",
      "5.0 3076\n",
      "1.0 3077\n",
      "4.0 3078\n",
      "4.0 3079\n",
      "4.0 3080\n",
      "5.0 3081\n",
      "4.0 3082\n",
      "4.0 3083\n",
      "5.0 3084\n",
      "4.0 3085\n",
      "3.0 3086\n",
      "5.0 3087\n",
      "1.0 3088\n",
      "5.0 3089\n",
      "4.0 3090\n",
      "5.0 3091\n",
      "4.0 3092\n",
      "5.0 3093\n",
      "5.0 3094\n",
      "5.0 3095\n",
      "5.0 3096\n",
      "3.0 3097\n",
      "5.0 3098\n",
      "5.0 3099\n",
      "5.0 3100\n",
      "5.0 3101\n",
      "4.0 3102\n",
      "5.0 3103\n",
      "5.0 3104\n",
      "4.0 3105\n",
      "5.0 3106\n",
      "4.0 3107\n",
      "5.0 3108\n",
      "4.0 3109\n",
      "5.0 3110\n",
      "5.0 3111\n",
      "5.0 3112\n",
      "5.0 3113\n",
      "4.0 3114\n",
      "5.0 3115\n",
      "5.0 3116\n",
      "5.0 3117\n",
      "1.0 3118\n",
      "4.0 3119\n",
      "2.0 3120\n",
      "5.0 3121\n",
      "5.0 3122\n",
      "5.0 3123\n",
      "4.0 3124\n",
      "4.0 3125\n",
      "4.0 3126\n",
      "5.0 3127\n",
      "1.0 3128\n",
      "5.0 3129\n",
      "5.0 3130\n",
      "5.0 3131\n",
      "3.0 3132\n",
      "5.0 3133\n",
      "5.0 3134\n",
      "4.0 3135\n",
      "5.0 3136\n",
      "5.0 3137\n",
      "4.0 3138\n",
      "5.0 3139\n",
      "5.0 3140\n",
      "4.0 3141\n",
      "5.0 3142\n",
      "5.0 3143\n",
      "5.0 3144\n",
      "5.0 3145\n",
      "5.0 3146\n",
      "5.0 3147\n",
      "5.0 3148\n",
      "5.0 3149\n",
      "5.0 3150\n",
      "5.0 3151\n",
      "5.0 3152\n",
      "5.0 3153\n",
      "5.0 3154\n",
      "5.0 3155\n",
      "1.0 3156\n",
      "5.0 3157\n",
      "5.0 3158\n",
      "5.0 3159\n",
      "5.0 3160\n",
      "5.0 3161\n",
      "5.0 3162\n",
      "5.0 3163\n",
      "5.0 3164\n",
      "5.0 3165\n",
      "5.0 3166\n",
      "5.0 3167\n",
      "5.0 3168\n",
      "4.0 3169\n",
      "5.0 3170\n",
      "5.0 3171\n",
      "4.0 3172\n",
      "5.0 3173\n",
      "3.0 3174\n",
      "4.0 3175\n",
      "1.0 3176\n",
      "5.0 3177\n",
      "1.0 3178\n",
      "1.0 3179\n",
      "2.0 3180\n",
      "5.0 3181\n",
      "5.0 3182\n",
      "5.0 3183\n",
      "5.0 3184\n",
      "5.0 3185\n",
      "5.0 3186\n",
      "5.0 3187\n",
      "4.0 3188\n",
      "4.0 3189\n",
      "4.0 3190\n",
      "4.0 3191\n",
      "5.0 3192\n",
      "5.0 3193\n",
      "3.0 3194\n",
      "5.0 3195\n",
      "5.0 3196\n",
      "5.0 3197\n",
      "4.0 3198\n",
      "4.0 3199\n",
      "5.0 3200\n",
      "1.0 3201\n",
      "5.0 3202\n",
      "4.0 3203\n",
      "5.0 3204\n",
      "5.0 3205\n",
      "5.0 3206\n",
      "5.0 3207\n",
      "5.0 3208\n",
      "5.0 3209\n",
      "5.0 3210\n",
      "4.0 3211\n",
      "1.0 3212\n",
      "1.0 3213\n",
      "3.0 3214\n",
      "2.0 3215\n",
      "5.0 3216\n",
      "4.0 3217\n",
      "1.0 3218\n",
      "5.0 3219\n",
      "3.0 3220\n",
      "5.0 3221\n",
      "4.0 3222\n",
      "5.0 3223\n",
      "5.0 3224\n",
      "4.0 3225\n",
      "5.0 3226\n",
      "5.0 3227\n",
      "5.0 3228\n",
      "5.0 3229\n",
      "5.0 3230\n",
      "3.0 3231\n",
      "5.0 3232\n",
      "5.0 3233\n",
      "5.0 3234\n",
      "5.0 3235\n",
      "4.0 3236\n",
      "5.0 3237\n",
      "5.0 3238\n",
      "5.0 3239\n",
      "5.0 3240\n",
      "5.0 3241\n",
      "5.0 3242\n",
      "5.0 3243\n",
      "5.0 3244\n",
      "5.0 3245\n",
      "5.0 3246\n",
      "5.0 3247\n",
      "5.0 3248\n",
      "1.0 3249\n",
      "1.0 3250\n",
      "1.0 3251\n",
      "5.0 3252\n",
      "5.0 3253\n",
      "5.0 3254\n",
      "5.0 3255\n",
      "5.0 3256\n",
      "5.0 3257\n",
      "5.0 3258\n",
      "5.0 3259\n",
      "4.0 3260\n",
      "5.0 3261\n",
      "5.0 3262\n",
      "4.0 3263\n",
      "5.0 3264\n",
      "5.0 3265\n",
      "5.0 3266\n",
      "5.0 3267\n",
      "4.0 3268\n",
      "1.0 3269\n",
      "4.0 3270\n",
      "5.0 3271\n",
      "5.0 3272\n",
      "5.0 3273\n",
      "5.0 3274\n",
      "5.0 3275\n",
      "5.0 3276\n",
      "5.0 3277\n",
      "4.0 3278\n",
      "5.0 3279\n",
      "5.0 3280\n",
      "5.0 3281\n",
      "5.0 3282\n",
      "3.0 3283\n",
      "3.0 3284\n",
      "4.0 3285\n",
      "5.0 3286\n",
      "5.0 3287\n",
      "5.0 3288\n",
      "5.0 3289\n",
      "5.0 3290\n",
      "5.0 3291\n",
      "5.0 3292\n",
      "5.0 3293\n",
      "5.0 3294\n",
      "5.0 3295\n",
      "5.0 3296\n",
      "5.0 3297\n",
      "5.0 3298\n",
      "5.0 3299\n",
      "5.0 3300\n",
      "5.0 3301\n",
      "5.0 3302\n",
      "5.0 3303\n",
      "5.0 3304\n",
      "5.0 3305\n",
      "5.0 3306\n",
      "4.0 3307\n",
      "2.0 3308\n",
      "5.0 3309\n",
      "5.0 3310\n",
      "5.0 3311\n",
      "1.0 3312\n",
      "5.0 3313\n",
      "4.0 3314\n",
      "5.0 3315\n",
      "5.0 3316\n",
      "2.0 3317\n",
      "4.0 3318\n",
      "5.0 3319\n",
      "4.0 3320\n",
      "4.0 3321\n",
      "4.0 3322\n",
      "3.0 3323\n",
      "4.0 3324\n",
      "4.0 3325\n",
      "4.0 3326\n",
      "5.0 3327\n",
      "5.0 3328\n",
      "5.0 3329\n",
      "5.0 3330\n",
      "5.0 3331\n",
      "4.0 3332\n",
      "5.0 3333\n",
      "5.0 3334\n",
      "5.0 3335\n",
      "5.0 3336\n",
      "4.0 3337\n",
      "5.0 3338\n",
      "5.0 3339\n",
      "5.0 3340\n",
      "5.0 3341\n",
      "4.0 3342\n",
      "4.0 3343\n",
      "5.0 3344\n",
      "5.0 3345\n",
      "5.0 3346\n",
      "5.0 3347\n",
      "5.0 3348\n",
      "5.0 3349\n",
      "4.0 3350\n",
      "5.0 3351\n",
      "5.0 3352\n",
      "4.0 3353\n",
      "5.0 3354\n",
      "5.0 3355\n",
      "5.0 3356\n",
      "5.0 3357\n",
      "4.0 3358\n",
      "4.0 3359\n",
      "5.0 3360\n",
      "5.0 3361\n",
      "5.0 3362\n",
      "5.0 3363\n",
      "4.0 3364\n",
      "4.0 3365\n",
      "1.0 3366\n",
      "5.0 3367\n",
      "5.0 3368\n",
      "5.0 3369\n",
      "5.0 3370\n",
      "5.0 3371\n",
      "5.0 3372\n",
      "5.0 3373\n",
      "2.0 3374\n",
      "1.0 3375\n",
      "3.0 3376\n",
      "5.0 3377\n",
      "4.0 3378\n",
      "5.0 3379\n",
      "4.0 3380\n",
      "5.0 3381\n",
      "1.0 3382\n",
      "4.0 3383\n",
      "3.0 3384\n",
      "5.0 3385\n",
      "4.0 3386\n",
      "5.0 3387\n",
      "5.0 3388\n",
      "5.0 3389\n",
      "5.0 3390\n",
      "5.0 3391\n",
      "5.0 3392\n",
      "5.0 3393\n",
      "4.0 3394\n",
      "4.0 3395\n",
      "2.0 3396\n",
      "4.0 3397\n",
      "4.0 3398\n",
      "5.0 3399\n",
      "3.0 3400\n",
      "5.0 3401\n",
      "5.0 3402\n",
      "5.0 3403\n",
      "4.0 3404\n",
      "5.0 3405\n",
      "4.0 3406\n",
      "5.0 3407\n",
      "5.0 3408\n",
      "5.0 3409\n",
      "5.0 3410\n",
      "5.0 3411\n",
      "1.0 3412\n",
      "3.0 3413\n",
      "4.0 3414\n",
      "2.0 3415\n",
      "4.0 3416\n",
      "5.0 3417\n",
      "5.0 3418\n",
      "5.0 3419\n",
      "5.0 3420\n",
      "5.0 3421\n",
      "5.0 3422\n",
      "5.0 3423\n",
      "5.0 3424\n",
      "5.0 3425\n",
      "5.0 3426\n",
      "5.0 3427\n",
      "5.0 3428\n",
      "5.0 3429\n",
      "5.0 3430\n",
      "5.0 3431\n",
      "4.0 3432\n",
      "5.0 3433\n",
      "5.0 3434\n",
      "5.0 3435\n",
      "3.0 3436\n",
      "5.0 3437\n",
      "4.0 3438\n",
      "5.0 3439\n",
      "5.0 3440\n",
      "3.0 3441\n",
      "4.0 3442\n",
      "5.0 3443\n",
      "4.0 3444\n",
      "4.0 3445\n",
      "5.0 3446\n",
      "4.0 3447\n",
      "5.0 3448\n",
      "5.0 3449\n",
      "4.0 3450\n",
      "5.0 3451\n",
      "1.0 3452\n",
      "5.0 3453\n",
      "5.0 3454\n",
      "5.0 3455\n",
      "5.0 3456\n",
      "4.0 3457\n",
      "4.0 3458\n",
      "1.0 3459\n",
      "1.0 3460\n",
      "5.0 3461\n",
      "5.0 3462\n",
      "3.0 3463\n",
      "4.0 3464\n",
      "4.0 3465\n",
      "4.0 3466\n",
      "4.0 3467\n",
      "4.0 3468\n",
      "3.0 3469\n",
      "5.0 3470\n",
      "5.0 3471\n",
      "4.0 3472\n",
      "1.0 3473\n",
      "5.0 3474\n",
      "5.0 3475\n",
      "5.0 3476\n",
      "4.0 3477\n",
      "5.0 3478\n",
      "4.0 3479\n",
      "5.0 3480\n",
      "5.0 3481\n",
      "1.0 3482\n",
      "1.0 3483\n",
      "1.0 3484\n",
      "5.0 3485\n",
      "1.0 3486\n",
      "5.0 3487\n",
      "5.0 3488\n",
      "1.0 3489\n",
      "5.0 3490\n",
      "4.0 3491\n",
      "4.0 3492\n",
      "5.0 3493\n",
      "4.0 3494\n",
      "5.0 3495\n",
      "4.0 3496\n",
      "5.0 3497\n",
      "4.0 3498\n",
      "5.0 3499\n",
      "5.0 3500\n",
      "5.0 3501\n",
      "5.0 3502\n",
      "5.0 3503\n",
      "4.0 3504\n",
      "1.0 3505\n",
      "5.0 3506\n",
      "5.0 3507\n",
      "5.0 3508\n",
      "5.0 3509\n",
      "3.0 3510\n",
      "5.0 3511\n",
      "4.0 3512\n",
      "4.0 3513\n",
      "3.0 3514\n",
      "5.0 3515\n",
      "4.0 3516\n",
      "5.0 3517\n",
      "5.0 3518\n",
      "5.0 3519\n",
      "5.0 3520\n",
      "2.0 3521\n",
      "5.0 3522\n",
      "1.0 3523\n",
      "3.0 3524\n",
      "4.0 3525\n",
      "4.0 3526\n",
      "4.0 3527\n",
      "3.0 3528\n",
      "4.0 3529\n",
      "4.0 3530\n",
      "5.0 3531\n",
      "5.0 3532\n",
      "5.0 3533\n",
      "5.0 3534\n",
      "5.0 3535\n",
      "4.0 3536\n",
      "5.0 3537\n",
      "5.0 3538\n",
      "4.0 3539\n",
      "5.0 3540\n",
      "4.0 3541\n",
      "1.0 3542\n",
      "5.0 3543\n",
      "5.0 3544\n",
      "3.0 3545\n",
      "5.0 3546\n",
      "5.0 3547\n",
      "5.0 3548\n",
      "5.0 3549\n",
      "5.0 3550\n",
      "5.0 3551\n",
      "5.0 3552\n",
      "5.0 3553\n",
      "5.0 3554\n",
      "5.0 3555\n",
      "3.0 3556\n",
      "5.0 3557\n",
      "5.0 3558\n",
      "5.0 3559\n",
      "4.0 3560\n",
      "1.0 3561\n",
      "5.0 3562\n",
      "5.0 3563\n",
      "5.0 3564\n",
      "5.0 3565\n",
      "3.0 3566\n",
      "3.0 3567\n",
      "1.0 3568\n",
      "4.0 3569\n",
      "5.0 3570\n",
      "3.0 3571\n",
      "5.0 3572\n",
      "2.0 3573\n",
      "3.0 3574\n",
      "5.0 3575\n",
      "4.0 3576\n",
      "5.0 3577\n",
      "5.0 3578\n",
      "5.0 3579\n",
      "5.0 3580\n",
      "1.0 3581\n",
      "5.0 3582\n",
      "1.0 3583\n",
      "5.0 3584\n",
      "1.0 3585\n",
      "5.0 3586\n",
      "5.0 3587\n",
      "5.0 3588\n",
      "4.0 3589\n",
      "3.0 3590\n",
      "4.0 3591\n",
      "5.0 3592\n",
      "5.0 3593\n",
      "4.0 3594\n",
      "5.0 3595\n",
      "5.0 3596\n",
      "3.0 3597\n",
      "4.0 3598\n",
      "5.0 3599\n",
      "5.0 3600\n",
      "3.0 3601\n",
      "5.0 3602\n",
      "5.0 3603\n",
      "4.0 3604\n",
      "5.0 3605\n",
      "4.0 3606\n",
      "5.0 3607\n",
      "3.0 3608\n",
      "4.0 3609\n",
      "3.0 3610\n",
      "5.0 3611\n",
      "2.0 3612\n",
      "4.0 3613\n",
      "4.0 3614\n",
      "4.0 3615\n",
      "5.0 3616\n",
      "5.0 3617\n",
      "4.0 3618\n",
      "4.0 3619\n",
      "5.0 3620\n",
      "5.0 3621\n",
      "4.0 3622\n",
      "4.0 3623\n",
      "5.0 3624\n",
      "5.0 3625\n",
      "4.0 3626\n",
      "4.0 3627\n",
      "4.0 3628\n",
      "4.0 3629\n",
      "5.0 3630\n",
      "4.0 3631\n",
      "5.0 3632\n",
      "4.0 3633\n",
      "5.0 3634\n",
      "5.0 3635\n",
      "5.0 3636\n",
      "1.0 3637\n",
      "5.0 3638\n",
      "5.0 3639\n",
      "5.0 3640\n",
      "5.0 3641\n",
      "5.0 3642\n",
      "4.0 3643\n",
      "4.0 3644\n",
      "5.0 3645\n",
      "5.0 3646\n",
      "5.0 3647\n",
      "5.0 3648\n",
      "5.0 3649\n",
      "5.0 3650\n",
      "5.0 3651\n",
      "5.0 3652\n",
      "4.0 3653\n",
      "5.0 3654\n",
      "3.0 3655\n",
      "1.0 3656\n",
      "5.0 3657\n",
      "4.0 3658\n",
      "5.0 3659\n",
      "4.0 3660\n",
      "3.0 3661\n",
      "3.0 3662\n",
      "1.0 3663\n",
      "1.0 3664\n",
      "4.0 3665\n",
      "5.0 3666\n",
      "5.0 3667\n",
      "5.0 3668\n",
      "5.0 3669\n",
      "1.0 3670\n",
      "5.0 3671\n",
      "5.0 3672\n",
      "5.0 3673\n",
      "5.0 3674\n",
      "5.0 3675\n",
      "1.0 3676\n",
      "2.0 3677\n",
      "4.0 3678\n",
      "5.0 3679\n",
      "5.0 3680\n",
      "5.0 3681\n",
      "5.0 3682\n",
      "5.0 3683\n",
      "5.0 3684\n",
      "5.0 3685\n",
      "5.0 3686\n",
      "4.0 3687\n",
      "5.0 3688\n",
      "5.0 3689\n",
      "4.0 3690\n",
      "4.0 3691\n",
      "5.0 3692\n",
      "5.0 3693\n",
      "3.0 3694\n",
      "1.0 3695\n",
      "1.0 3696\n",
      "5.0 3697\n",
      "4.0 3698\n",
      "4.0 3699\n",
      "1.0 3700\n",
      "5.0 3701\n",
      "3.0 3702\n",
      "5.0 3703\n",
      "4.0 3704\n",
      "5.0 3705\n",
      "5.0 3706\n",
      "5.0 3707\n",
      "5.0 3708\n",
      "5.0 3709\n",
      "1.0 3710\n",
      "3.0 3711\n",
      "5.0 3712\n",
      "1.0 3713\n",
      "3.0 3714\n",
      "5.0 3715\n",
      "5.0 3716\n",
      "5.0 3717\n",
      "4.0 3718\n",
      "3.0 3719\n",
      "1.0 3720\n",
      "3.0 3721\n",
      "5.0 3722\n",
      "5.0 3723\n",
      "5.0 3724\n",
      "4.0 3725\n",
      "4.0 3726\n",
      "4.0 3727\n",
      "5.0 3728\n",
      "4.0 3729\n",
      "5.0 3730\n",
      "4.0 3731\n",
      "1.0 3732\n",
      "2.0 3733\n",
      "1.0 3734\n",
      "4.0 3735\n",
      "1.0 3736\n",
      "5.0 3737\n",
      "4.0 3738\n",
      "4.0 3739\n",
      "3.0 3740\n",
      "5.0 3741\n",
      "5.0 3742\n",
      "1.0 3743\n",
      "5.0 3744\n",
      "5.0 3745\n",
      "5.0 3746\n",
      "5.0 3747\n",
      "1.0 3748\n",
      "4.0 3749\n",
      "5.0 3750\n",
      "5.0 3751\n",
      "4.0 3752\n",
      "5.0 3753\n",
      "4.0 3754\n",
      "4.0 3755\n",
      "5.0 3756\n",
      "4.0 3757\n",
      "2.0 3758\n",
      "1.0 3759\n",
      "1.0 3760\n",
      "5.0 3761\n",
      "5.0 3762\n",
      "3.0 3763\n",
      "5.0 3764\n",
      "4.0 3765\n",
      "2.0 3766\n",
      "5.0 3767\n",
      "3.0 3768\n",
      "1.0 3769\n",
      "4.0 3770\n",
      "5.0 3771\n",
      "5.0 3772\n",
      "5.0 3773\n",
      "4.0 3774\n",
      "1.0 3775\n",
      "1.0 3776\n",
      "1.0 3777\n",
      "2.0 3778\n",
      "1.0 3779\n",
      "4.0 3780\n",
      "4.0 3781\n",
      "4.0 3782\n",
      "1.0 3783\n",
      "5.0 3784\n",
      "5.0 3785\n",
      "4.0 3786\n",
      "5.0 3787\n",
      "5.0 3788\n",
      "5.0 3789\n",
      "4.0 3790\n",
      "4.0 3791\n",
      "3.0 3792\n",
      "5.0 3793\n",
      "4.0 3794\n",
      "5.0 3795\n",
      "5.0 3796\n",
      "5.0 3797\n",
      "5.0 3798\n",
      "5.0 3799\n",
      "4.0 3800\n",
      "1.0 3801\n",
      "5.0 3802\n",
      "4.0 3803\n",
      "5.0 3804\n",
      "3.0 3805\n",
      "5.0 3806\n",
      "4.0 3807\n",
      "4.0 3808\n",
      "4.0 3809\n",
      "5.0 3810\n",
      "5.0 3811\n",
      "5.0 3812\n",
      "5.0 3813\n",
      "5.0 3814\n",
      "2.0 3815\n",
      "4.0 3816\n",
      "5.0 3817\n",
      "5.0 3818\n",
      "4.0 3819\n",
      "4.0 3820\n",
      "3.0 3821\n",
      "5.0 3822\n",
      "3.0 3823\n",
      "5.0 3824\n",
      "5.0 3825\n",
      "5.0 3826\n",
      "4.0 3827\n",
      "5.0 3828\n",
      "5.0 3829\n",
      "5.0 3830\n",
      "2.0 3831\n",
      "5.0 3832\n",
      "4.0 3833\n",
      "5.0 3834\n",
      "5.0 3835\n",
      "5.0 3836\n",
      "5.0 3837\n",
      "5.0 3838\n",
      "2.0 3839\n",
      "4.0 3840\n",
      "1.0 3841\n",
      "5.0 3842\n",
      "5.0 3843\n",
      "5.0 3844\n",
      "5.0 3845\n",
      "5.0 3846\n",
      "5.0 3847\n",
      "5.0 3848\n",
      "3.0 3849\n",
      "1.0 3850\n",
      "4.0 3851\n",
      "1.0 3852\n",
      "1.0 3853\n",
      "5.0 3854\n",
      "4.0 3855\n",
      "3.0 3856\n",
      "5.0 3857\n",
      "5.0 3858\n",
      "4.0 3859\n",
      "5.0 3860\n",
      "5.0 3861\n",
      "5.0 3862\n",
      "3.0 3863\n",
      "5.0 3864\n",
      "1.0 3865\n",
      "5.0 3866\n",
      "5.0 3867\n",
      "4.0 3868\n",
      "1.0 3869\n",
      "4.0 3870\n",
      "1.0 3871\n",
      "5.0 3872\n",
      "5.0 3873\n",
      "3.0 3874\n",
      "5.0 3875\n",
      "5.0 3876\n",
      "1.0 3877\n",
      "1.0 3878\n",
      "5.0 3879\n",
      "5.0 3880\n",
      "5.0 3881\n",
      "4.0 3882\n",
      "5.0 3883\n",
      "1.0 3884\n",
      "1.0 3885\n",
      "2.0 3886\n",
      "1.0 3887\n",
      "4.0 3888\n",
      "4.0 3889\n",
      "5.0 3890\n",
      "3.0 3891\n",
      "5.0 3892\n",
      "4.0 3893\n",
      "4.0 3894\n",
      "5.0 3895\n",
      "5.0 3896\n",
      "4.0 3897\n",
      "5.0 3898\n",
      "5.0 3899\n",
      "5.0 3900\n",
      "5.0 3901\n",
      "5.0 3902\n",
      "4.0 3903\n",
      "1.0 3904\n",
      "3.0 3905\n",
      "1.0 3906\n",
      "1.0 3907\n",
      "5.0 3908\n",
      "5.0 3909\n",
      "5.0 3910\n",
      "4.0 3911\n",
      "5.0 3912\n",
      "1.0 3913\n",
      "4.0 3914\n",
      "5.0 3915\n",
      "5.0 3916\n",
      "5.0 3917\n",
      "5.0 3918\n",
      "5.0 3919\n",
      "5.0 3920\n",
      "1.0 3921\n",
      "5.0 3922\n",
      "3.0 3923\n",
      "1.0 3924\n",
      "3.0 3925\n",
      "4.0 3926\n",
      "3.0 3927\n",
      "5.0 3928\n",
      "4.0 3929\n",
      "5.0 3930\n",
      "4.0 3931\n",
      "4.0 3932\n",
      "2.0 3933\n",
      "3.0 3934\n",
      "4.0 3935\n",
      "5.0 3936\n",
      "1.0 3937\n",
      "1.0 3938\n",
      "5.0 3939\n",
      "3.0 3940\n",
      "1.0 3941\n",
      "5.0 3942\n",
      "5.0 3943\n",
      "4.0 3944\n",
      "2.0 3945\n",
      "1.0 3946\n",
      "4.0 3947\n",
      "1.0 3948\n",
      "4.0 3949\n",
      "4.0 3950\n",
      "4.0 3951\n",
      "1.0 3952\n",
      "5.0 3953\n",
      "5.0 3954\n",
      "5.0 3955\n",
      "5.0 3956\n",
      "5.0 3957\n",
      "5.0 3958\n",
      "3.0 3959\n",
      "5.0 3960\n",
      "4.0 3961\n",
      "4.0 3962\n",
      "3.0 3963\n",
      "5.0 3964\n",
      "5.0 3965\n",
      "4.0 3966\n",
      "5.0 3967\n",
      "5.0 3968\n",
      "2.0 3969\n",
      "5.0 3970\n",
      "5.0 3971\n",
      "1.0 3972\n",
      "1.0 3973\n",
      "1.0 3974\n",
      "4.0 3975\n",
      "5.0 3976\n",
      "5.0 3977\n",
      "5.0 3978\n",
      "5.0 3979\n",
      "5.0 3980\n",
      "5.0 3981\n",
      "5.0 3982\n",
      "5.0 3983\n",
      "4.0 3984\n",
      "5.0 3985\n",
      "4.0 3986\n",
      "1.0 3987\n",
      "5.0 3988\n",
      "3.0 3989\n",
      "1.0 3990\n",
      "5.0 3991\n",
      "4.0 3992\n",
      "5.0 3993\n",
      "5.0 3994\n",
      "5.0 3995\n",
      "5.0 3996\n",
      "5.0 3997\n",
      "1.0 3998\n",
      "5.0 3999\n",
      "3.0 4000\n",
      "1.0 4001\n",
      "5.0 4002\n",
      "4.0 4003\n",
      "2.0 4004\n",
      "4.0 4005\n",
      "5.0 4006\n",
      "5.0 4007\n",
      "5.0 4008\n",
      "5.0 4009\n",
      "4.0 4010\n",
      "5.0 4011\n",
      "5.0 4012\n",
      "4.0 4013\n",
      "4.0 4014\n",
      "4.0 4015\n",
      "5.0 4016\n",
      "5.0 4017\n",
      "5.0 4018\n",
      "4.0 4019\n",
      "4.0 4020\n",
      "4.0 4021\n",
      "5.0 4022\n",
      "5.0 4023\n",
      "5.0 4024\n",
      "4.5 4025\n",
      "5.0 4026\n",
      "4.0 4027\n",
      "5.0 4028\n",
      "3.0 4029\n",
      "5.0 4030\n",
      "4.0 4031\n",
      "3.0 4032\n",
      "1.0 4033\n",
      "4.0 4034\n",
      "5.0 4035\n",
      "1.0 4036\n",
      "5.0 4037\n",
      "4.0 4038\n",
      "5.0 4039\n",
      "3.0 4040\n",
      "4.0 4041\n",
      "4.0 4042\n",
      "4.0 4043\n",
      "5.0 4044\n",
      "4.0 4045\n",
      "1.0 4046\n",
      "5.0 4047\n",
      "4.0 4048\n",
      "5.0 4049\n",
      "4.0 4050\n",
      "5.0 4051\n",
      "5.0 4052\n",
      "4.0 4053\n",
      "4.0 4054\n",
      "5.0 4055\n",
      "5.0 4056\n",
      "5.0 4057\n",
      "4.0 4058\n",
      "4.0 4059\n",
      "4.0 4060\n",
      "1.0 4061\n",
      "1.0 4062\n",
      "2.0 4063\n",
      "1.0 4064\n",
      "5.0 4065\n",
      "1.0 4066\n",
      "3.0 4067\n",
      "4.0 4068\n",
      "3.0 4069\n",
      "2.0 4070\n",
      "2.0 4071\n",
      "5.0 4072\n",
      "1.0 4073\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4.0 4074\n",
      "1.0 4075\n",
      "5.0 4076\n",
      "1.0 4077\n",
      "5.0 4078\n",
      "5.0 4079\n",
      "5.0 4080\n",
      "3.0 4081\n",
      "5.0 4082\n",
      "5.0 4083\n",
      "5.0 4084\n",
      "4.0 4085\n",
      "5.0 4086\n",
      "4.0 4087\n",
      "3.0 4088\n",
      "5.0 4089\n",
      "5.0 4090\n",
      "5.0 4091\n",
      "5.0 4092\n",
      "5.0 4093\n",
      "5.0 4094\n",
      "5.0 4095\n",
      "1.0 4096\n",
      "4.0 4097\n",
      "1.0 4098\n",
      "4.0 4099\n",
      "4.0 4100\n",
      "5.0 4101\n",
      "4.0 4102\n",
      "4.0 4103\n",
      "4.0 4104\n",
      "1.0 4105\n",
      "5.0 4106\n",
      "5.0 4107\n",
      "3.0 4108\n",
      "5.0 4109\n",
      "5.0 4110\n",
      "1.0 4111\n",
      "3.0 4112\n",
      "1.0 4113\n",
      "1.0 4114\n",
      "5.0 4115\n",
      "5.0 4116\n",
      "5.0 4117\n",
      "5.0 4118\n",
      "5.0 4119\n",
      "5.0 4120\n",
      "3.0 4121\n",
      "1.0 4122\n",
      "3.0 4123\n",
      "1.0 4124\n",
      "1.0 4125\n",
      "5.0 4126\n",
      "4.0 4127\n",
      "5.0 4128\n",
      "1.0 4129\n",
      "4.0 4130\n",
      "1.0 4131\n",
      "3.0 4132\n",
      "1.0 4133\n",
      "5.0 4134\n",
      "1.0 4135\n",
      "2.0 4136\n",
      "5.0 4137\n",
      "4.0 4138\n",
      "4.0 4139\n",
      "5.0 4140\n",
      "5.0 4141\n",
      "5.0 4142\n",
      "5.0 4143\n",
      "1.0 4144\n",
      "4.0 4145\n",
      "1.0 4146\n",
      "3.0 4147\n",
      "4.0 4148\n",
      "1.0 4149\n",
      "4.0 4150\n",
      "1.0 4151\n",
      "2.0 4152\n",
      "5.0 4153\n",
      "5.0 4154\n",
      "1.0 4155\n",
      "5.0 4156\n",
      "5.0 4157\n",
      "4.0 4158\n",
      "1.0 4159\n",
      "5.0 4160\n",
      "4.0 4161\n",
      "5.0 4162\n",
      "5.0 4163\n",
      "5.0 4164\n",
      "5.0 4165\n",
      "5.0 4166\n",
      "5.0 4167\n",
      "3.0 4168\n",
      "3.0 4169\n",
      "4.0 4170\n",
      "4.0 4171\n",
      "3.0 4172\n",
      "4.0 4173\n",
      "1.0 4174\n",
      "3.0 4175\n",
      "4.0 4176\n",
      "4.0 4177\n",
      "4.0 4178\n",
      "5.0 4179\n",
      "4.0 4180\n",
      "4.0 4181\n",
      "1.0 4182\n",
      "5.0 4183\n",
      "1.0 4184\n",
      "1.0 4185\n",
      "5.0 4186\n",
      "4.0 4187\n",
      "5.0 4188\n",
      "5.0 4189\n",
      "5.0 4190\n",
      "1.0 4191\n",
      "4.0 4192\n",
      "5.0 4193\n",
      "5.0 4194\n",
      "5.0 4195\n",
      "5.0 4196\n",
      "5.0 4197\n",
      "4.0 4198\n",
      "5.0 4199\n",
      "5.0 4200\n",
      "5.0 4201\n",
      "5.0 4202\n",
      "5.0 4203\n",
      "5.0 4204\n",
      "5.0 4205\n",
      "1.0 4206\n",
      "1.0 4207\n",
      "4.0 4208\n",
      "4.0 4209\n",
      "1.0 4210\n",
      "5.0 4211\n",
      "1.0 4212\n",
      "5.0 4213\n",
      "5.0 4214\n",
      "5.0 4215\n",
      "4.0 4216\n",
      "5.0 4217\n",
      "3.0 4218\n",
      "5.0 4219\n",
      "5.0 4220\n",
      "4.0 4221\n",
      "4.0 4222\n",
      "5.0 4223\n",
      "5.0 4224\n",
      "4.0 4225\n",
      "5.0 4226\n",
      "5.0 4227\n",
      "5.0 4228\n",
      "3.0 4229\n",
      "5.0 4230\n",
      "4.0 4231\n",
      "4.0 4232\n",
      "4.0 4233\n",
      "4.0 4234\n",
      "3.0 4235\n",
      "5.0 4236\n",
      "5.0 4237\n",
      "5.0 4238\n",
      "5.0 4239\n",
      "4.0 4240\n",
      "5.0 4241\n",
      "5.0 4242\n",
      "4.0 4243\n",
      "5.0 4244\n",
      "3.0 4245\n",
      "3.0 4246\n",
      "5.0 4247\n",
      "1.0 4248\n",
      "1.0 4249\n",
      "5.0 4250\n",
      "5.0 4251\n",
      "4.0 4252\n",
      "5.0 4253\n",
      "3.0 4254\n",
      "4.0 4255\n",
      "5.0 4256\n",
      "5.0 4257\n",
      "5.0 4258\n",
      "5.0 4259\n",
      "5.0 4260\n",
      "1.0 4261\n",
      "4.0 4262\n",
      "3.0 4263\n",
      "5.0 4264\n",
      "5.0 4265\n",
      "1.0 4266\n",
      "1.0 4267\n",
      "4.0 4268\n",
      "1.0 4269\n",
      "3.0 4270\n",
      "4.0 4271\n",
      "5.0 4272\n",
      "4.0 4273\n",
      "4.0 4274\n",
      "4.0 4275\n",
      "5.0 4276\n",
      "5.0 4277\n",
      "4.0 4278\n",
      "5.0 4279\n",
      "4.5 4280\n",
      "5.0 4281\n",
      "5.0 4282\n",
      "5.0 4283\n",
      "2.0 4284\n",
      "5.0 4285\n",
      "1.0 4286\n",
      "1.0 4287\n",
      "4.0 4288\n",
      "1.0 4289\n",
      "1.0 4290\n",
      "1.0 4291\n",
      "1.0 4292\n",
      "5.0 4293\n",
      "4.0 4294\n",
      "5.0 4295\n",
      "5.0 4296\n",
      "4.0 4297\n",
      "1.0 4298\n",
      "1.0 4299\n",
      "4.0 4300\n",
      "5.0 4301\n",
      "1.0 4302\n",
      "4.0 4303\n",
      "5.0 4304\n",
      "1.0 4305\n",
      "2.0 4306\n",
      "1.0 4307\n",
      "1.0 4308\n",
      "4.0 4309\n",
      "5.0 4310\n",
      "1.0 4311\n",
      "1.0 4312\n",
      "1.0 4313\n",
      "1.0 4314\n",
      "1.0 4315\n",
      "4.0 4316\n",
      "2.0 4317\n",
      "1.0 4318\n",
      "1.0 4319\n",
      "1.0 4320\n",
      "1.0 4321\n",
      "4.0 4322\n",
      "4.0 4323\n",
      "2.0 4324\n",
      "4.0 4325\n",
      "1.0 4326\n",
      "5.0 4327\n",
      "5.0 4328\n",
      "5.0 4329\n",
      "5.0 4330\n",
      "4.0 4331\n",
      "5.0 4332\n",
      "4.0 4333\n",
      "5.0 4334\n",
      "4.5 4335\n",
      "4.0 4336\n",
      "5.0 4337\n",
      "4.0 4338\n",
      "1.0 4339\n",
      "4.0 4340\n",
      "1.0 4341\n",
      "4.0 4342\n",
      "5.0 4343\n",
      "5.0 4344\n",
      "1.0 4345\n",
      "4.0 4346\n",
      "1.0 4347\n",
      "3.0 4348\n",
      "1.0 4349\n",
      "1.0 4350\n",
      "4.0 4351\n",
      "5.0 4352\n",
      "1.0 4353\n",
      "5.0 4354\n",
      "5.0 4355\n",
      "3.0 4356\n",
      "3.0 4357\n",
      "1.0 4358\n",
      "4.0 4359\n",
      "1.0 4360\n",
      "5.0 4361\n",
      "5.0 4362\n",
      "1.0 4363\n",
      "5.0 4364\n",
      "4.0 4365\n",
      "4.0 4366\n",
      "1.0 4367\n",
      "4.0 4368\n",
      "5.0 4369\n",
      "1.0 4370\n",
      "5.0 4371\n",
      "1.0 4372\n",
      "1.0 4373\n",
      "4.0 4374\n",
      "3.0 4375\n",
      "5.0 4376\n",
      "1.0 4377\n",
      "5.0 4378\n",
      "5.0 4379\n",
      "5.0 4380\n",
      "3.0 4381\n",
      "1.0 4382\n",
      "1.0 4383\n",
      "2.0 4384\n",
      "5.0 4385\n",
      "5.0 4386\n",
      "5.0 4387\n",
      "1.0 4388\n",
      "5.0 4389\n",
      "5.0 4390\n",
      "4.0 4391\n",
      "1.0 4392\n",
      "1.0 4393\n",
      "5.0 4394\n",
      "1.0 4395\n",
      "5.0 4396\n",
      "5.0 4397\n",
      "5.0 4398\n",
      "5.0 4399\n",
      "5.0 4400\n",
      "5.0 4401\n",
      "5.0 4402\n",
      "3.0 4403\n",
      "4.0 4404\n",
      "1.0 4405\n",
      "5.0 4406\n",
      "3.0 4407\n",
      "5.0 4408\n",
      "1.0 4409\n",
      "4.0 4410\n",
      "3.0 4411\n",
      "4.0 4412\n",
      "3.0 4413\n",
      "2.0 4414\n",
      "4.5 4415\n",
      "4.0 4416\n",
      "1.0 4417\n",
      "4.5 4418\n",
      "5.0 4419\n",
      "1.0 4420\n",
      "1.0 4421\n",
      "1.0 4422\n",
      "4.0 4423\n",
      "5.0 4424\n",
      "5.0 4425\n",
      "5.0 4426\n",
      "5.0 4427\n",
      "4.0 4428\n",
      "1.0 4429\n",
      "5.0 4430\n",
      "5.0 4431\n",
      "4.0 4432\n",
      "5.0 4433\n",
      "5.0 4434\n",
      "1.0 4435\n",
      "5.0 4436\n",
      "3.0 4437\n",
      "1.0 4438\n",
      "1.0 4439\n",
      "4.0 4440\n",
      "4.0 4441\n",
      "4.0 4442\n",
      "4.0 4443\n",
      "3.0 4444\n",
      "4.0 4445\n",
      "4.0 4446\n",
      "4.0 4447\n",
      "4.0 4448\n",
      "4.0 4449\n",
      "5.0 4450\n",
      "5.0 4451\n",
      "5.0 4452\n",
      "1.0 4453\n",
      "5.0 4454\n",
      "4.0 4455\n",
      "5.0 4456\n",
      "5.0 4457\n",
      "3.0 4458\n",
      "5.0 4459\n",
      "4.0 4460\n",
      "4.0 4461\n",
      "5.0 4462\n",
      "1.0 4463\n",
      "2.0 4464\n",
      "5.0 4465\n",
      "1.0 4466\n",
      "3.0 4467\n",
      "1.0 4468\n",
      "5.0 4469\n",
      "4.0 4470\n",
      "1.0 4471\n",
      "1.0 4472\n",
      "1.0 4473\n",
      "5.0 4474\n",
      "1.0 4475\n",
      "1.0 4476\n",
      "1.0 4477\n",
      "2.0 4478\n",
      "1.0 4479\n",
      "1.0 4480\n",
      "4.0 4481\n",
      "1.0 4482\n",
      "3.0 4483\n",
      "1.0 4484\n",
      "3.0 4485\n",
      "4.0 4486\n",
      "2.0 4487\n",
      "2.0 4488\n",
      "5.0 4489\n",
      "1.0 4490\n",
      "5.0 4491\n",
      "5.0 4492\n",
      "4.0 4493\n",
      "5.0 4494\n",
      "5.0 4495\n",
      "5.0 4496\n",
      "4.0 4497\n",
      "5.0 4498\n",
      "4.0 4499\n",
      "5.0 4500\n",
      "2.0 4501\n",
      "3.0 4502\n",
      "4.0 4503\n",
      "5.0 4504\n",
      "1.0 4505\n",
      "5.0 4506\n",
      "5.0 4507\n",
      "5.0 4508\n",
      "1.0 4509\n",
      "1.0 4510\n",
      "4.0 4511\n",
      "1.0 4512\n",
      "1.0 4513\n",
      "5.0 4514\n",
      "5.0 4515\n",
      "5.0 4516\n",
      "5.0 4517\n",
      "4.0 4518\n",
      "2.0 4519\n",
      "1.0 4520\n",
      "4.0 4521\n",
      "4.0 4522\n",
      "3.0 4523\n",
      "1.0 4524\n",
      "1.0 4525\n",
      "4.0 4526\n",
      "5.0 4527\n",
      "4.0 4528\n",
      "4.0 4529\n",
      "1.0 4530\n",
      "1.0 4531\n",
      "4.0 4532\n",
      "5.0 4533\n",
      "2.0 4534\n",
      "2.0 4535\n",
      "4.0 4536\n",
      "2.0 4537\n",
      "5.0 4538\n",
      "1.0 4539\n",
      "1.0 4540\n",
      "5.0 4541\n",
      "4.0 4542\n",
      "1.0 4543\n",
      "2.0 4544\n",
      "4.0 4545\n",
      "4.0 4546\n",
      "2.0 4547\n",
      "2.0 4548\n",
      "3.0 4549\n",
      "4.0 4550\n",
      "2.0 4551\n",
      "5.0 4552\n",
      "3.0 4553\n",
      "5.0 4554\n",
      "5.0 4555\n",
      "1.0 4556\n",
      "4.0 4557\n",
      "4.0 4558\n",
      "5.0 4559\n",
      "1.0 4560\n",
      "5.0 4561\n",
      "3.0 4562\n",
      "4.0 4563\n",
      "1.0 4564\n",
      "1.0 4565\n",
      "1.0 4566\n",
      "1.0 4567\n",
      "1.0 4568\n",
      "3.0 4569\n",
      "5.0 4570\n",
      "1.0 4571\n",
      "1.0 4572\n",
      "5.0 4573\n",
      "2.0 4574\n",
      "5.0 4575\n",
      "1.0 4576\n",
      "1.0 4577\n",
      "2.0 4578\n",
      "1.0 4579\n",
      "3.5 4580\n",
      "5.0 4581\n",
      "4.5 4582\n",
      "5.0 4583\n",
      "4.0 4584\n",
      "2.0 4585\n",
      "5.0 4586\n",
      "5.0 4587\n",
      "5.0 4588\n",
      "5.0 4589\n",
      "5.0 4590\n",
      "4.0 4591\n",
      "1.0 4592\n",
      "5.0 4593\n",
      "1.0 4594\n",
      "5.0 4595\n",
      "5.0 4596\n",
      "5.0 4597\n",
      "1.0 4598\n",
      "1.0 4599\n",
      "5.0 4600\n",
      "1.0 4601\n",
      "1.0 4602\n",
      "2.0 4603\n",
      "4.0 4604\n",
      "2.0 4605\n",
      "3.0 4606\n",
      "4.0 4607\n",
      "3.0 4608\n",
      "4.0 4609\n",
      "4.0 4610\n",
      "2.0 4611\n",
      "5.0 4612\n",
      "1.0 4613\n",
      "5.0 4614\n",
      "4.0 4615\n",
      "1.0 4616\n",
      "2.0 4617\n",
      "5.0 4618\n",
      "4.0 4619\n",
      "3.0 4620\n",
      "1.0 4621\n",
      "4.0 4622\n",
      "1.0 4623\n",
      "1.0 4624\n",
      "4.0 4625\n",
      "5.0 4626\n",
      "1.0 4627\n",
      "1.0 4628\n",
      "4.0 4629\n",
      "5.0 4630\n",
      "1.0 4631\n",
      "1.0 4632\n",
      "5.0 4633\n",
      "5.0 4634\n",
      "1.0 4635\n",
      "5.0 4636\n",
      "5.0 4637\n",
      "4.0 4638\n",
      "5.0 4639\n",
      "1.0 4640\n",
      "1.0 4641\n",
      "4.0 4642\n",
      "3.0 4643\n",
      "4.0 4644\n",
      "1.0 4645\n",
      "5.0 4646\n",
      "1.0 4647\n",
      "5.0 4648\n",
      "3.0 4649\n",
      "3.0 4650\n",
      "5.0 4651\n",
      "3.0 4652\n",
      "2.0 4653\n",
      "5.0 4654\n",
      "5.0 4655\n",
      "4.0 4656\n",
      "1.0 4657\n",
      "1.0 4658\n",
      "1.0 4659\n",
      "4.0 4660\n",
      "5.0 4661\n",
      "1.0 4662\n",
      "4.0 4663\n",
      "4.0 4664\n",
      "1.0 4665\n",
      "4.0 4666\n",
      "5.0 4667\n",
      "5.0 4668\n",
      "5.0 4669\n",
      "4.0 4670\n",
      "5.0 4671\n",
      "1.0 4672\n",
      "2.0 4673\n",
      "1.0 4674\n",
      "3.0 4675\n",
      "1.0 4676\n",
      "1.0 4677\n",
      "5.0 4678\n",
      "4.0 4679\n",
      "1.0 4680\n",
      "1.0 4681\n",
      "3.0 4682\n",
      "4.0 4683\n",
      "3.0 4684\n",
      "3.0 4685\n",
      "1.0 4686\n",
      "1.0 4687\n",
      "3.0 4688\n",
      "5.0 4689\n",
      "5.0 4690\n",
      "5.0 4691\n",
      "3.0 4692\n",
      "1.0 4693\n",
      "2.0 4694\n",
      "2.0 4695\n",
      "4.0 4696\n",
      "5.0 4697\n",
      "5.0 4698\n",
      "1.0 4699\n",
      "2.0 4700\n",
      "1.0 4701\n",
      "5.0 4702\n",
      "5.0 4703\n",
      "5.0 4704\n",
      "5.0 4705\n",
      "5.0 4706\n",
      "5.0 4707\n",
      "4.0 4708\n",
      "5.0 4709\n",
      "5.0 4710\n",
      "5.0 4711\n",
      "2.0 4712\n",
      "1.0 4713\n",
      "1.0 4714\n",
      "2.0 4715\n",
      "4.0 4716\n",
      "4.0 4717\n",
      "5.0 4718\n",
      "3.0 4719\n",
      "5.0 4720\n",
      "3.0 4721\n",
      "3.0 4722\n",
      "1.0 4723\n",
      "1.0 4724\n",
      "4.0 4725\n",
      "5.0 4726\n",
      "1.0 4727\n",
      "1.0 4728\n",
      "1.0 4729\n",
      "1.0 4730\n",
      "1.0 4731\n",
      "5.0 4732\n",
      "1.0 4733\n",
      "5.0 4734\n",
      "4.0 4735\n",
      "4.0 4736\n",
      "3.0 4737\n",
      "5.0 4738\n",
      "4.0 4739\n",
      "5.0 4740\n",
      "5.0 4741\n",
      "1.0 4742\n",
      "1.0 4743\n",
      "1.0 4744\n",
      "1.0 4745\n",
      "3.0 4746\n",
      "4.0 4747\n",
      "4.0 4748\n",
      "5.0 4749\n",
      "5.0 4750\n",
      "5.0 4751\n",
      "5.0 4752\n",
      "4.0 4753\n",
      "4.0 4754\n",
      "2.0 4755\n",
      "5.0 4756\n",
      "5.0 4757\n",
      "5.0 4758\n",
      "3.0 4759\n",
      "3.0 4760\n",
      "5.0 4761\n",
      "5.0 4762\n",
      "4.0 4763\n",
      "5.0 4764\n",
      "5.0 4765\n",
      "2.0 4766\n",
      "5.0 4767\n",
      "1.0 4768\n",
      "4.0 4769\n",
      "4.0 4770\n",
      "5.0 4771\n",
      "5.0 4772\n",
      "5.0 4773\n",
      "1.0 4774\n",
      "5.0 4775\n",
      "4.0 4776\n",
      "5.0 4777\n",
      "1.0 4778\n",
      "4.0 4779\n",
      "5.0 4780\n",
      "3.0 4781\n",
      "5.0 4782\n",
      "2.0 4783\n",
      "3.0 4784\n",
      "5.0 4785\n",
      "2.0 4786\n",
      "5.0 4787\n",
      "1.0 4788\n",
      "4.0 4789\n",
      "5.0 4790\n",
      "5.0 4791\n",
      "5.0 4792\n",
      "5.0 4793\n",
      "5.0 4794\n",
      "5.0 4795\n",
      "5.0 4796\n",
      "2.0 4797\n",
      "4.0 4798\n",
      "1.0 4799\n",
      "3.0 4800\n",
      "3.0 4801\n",
      "5.0 4802\n",
      "4.0 4803\n",
      "4.0 4804\n",
      "5.0 4805\n",
      "5.0 4806\n",
      "5.0 4807\n",
      "5.0 4808\n",
      "5.0 4809\n",
      "3.0 4810\n",
      "5.0 4811\n",
      "1.0 4812\n",
      "5.0 4813\n",
      "1.0 4814\n",
      "3.0 4815\n",
      "5.0 4816\n",
      "4.0 4817\n",
      "1.0 4818\n",
      "4.0 4819\n",
      "3.0 4820\n",
      "4.0 4821\n",
      "5.0 4822\n",
      "5.0 4823\n",
      "1.0 4824\n",
      "5.0 4825\n",
      "5.0 4826\n",
      "1.0 4827\n",
      "1.0 4828\n",
      "1.0 4829\n",
      "5.0 4830\n",
      "5.0 4831\n",
      "4.0 4832\n",
      "1.0 4833\n",
      "1.0 4834\n",
      "5.0 4835\n",
      "5.0 4836\n",
      "5.0 4837\n",
      "3.0 4838\n",
      "5.0 4839\n",
      "5.0 4840\n",
      "4.0 4841\n",
      "4.0 4842\n",
      "4.0 4843\n",
      "1.0 4844\n",
      "1.0 4845\n",
      "1.0 4846\n",
      "4.0 4847\n",
      "4.0 4848\n",
      "1.0 4849\n",
      "1.0 4850\n",
      "1.0 4851\n",
      "1.0 4852\n",
      "5.0 4853\n",
      "1.0 4854\n",
      "1.0 4855\n",
      "4.0 4856\n",
      "3.0 4857\n",
      "3.0 4858\n",
      "4.0 4859\n",
      "5.0 4860\n",
      "5.0 4861\n",
      "1.0 4862\n",
      "5.0 4863\n",
      "1.0 4864\n",
      "5.0 4865\n",
      "4.0 4866\n",
      "3.0 4867\n",
      "4.0 4868\n",
      "5.0 4869\n",
      "2.0 4870\n",
      "1.0 4871\n",
      "4.0 4872\n",
      "4.0 4873\n",
      "3.0 4874\n",
      "4.0 4875\n",
      "5.0 4876\n",
      "4.0 4877\n",
      "1.0 4878\n",
      "4.0 4879\n",
      "3.0 4880\n",
      "5.0 4881\n",
      "5.0 4882\n",
      "5.0 4883\n",
      "5.0 4884\n",
      "4.0 4885\n",
      "4.0 4886\n",
      "4.0 4887\n",
      "5.0 4888\n",
      "1.0 4889\n",
      "5.0 4890\n",
      "1.0 4891\n",
      "5.0 4892\n",
      "5.0 4893\n",
      "5.0 4894\n",
      "5.0 4895\n",
      "2.0 4896\n",
      "3.0 4897\n",
      "1.0 4898\n",
      "2.0 4899\n",
      "2.0 4900\n",
      "4.0 4901\n",
      "3.0 4902\n",
      "5.0 4903\n",
      "5.0 4904\n",
      "5.0 4905\n",
      "5.0 4906\n",
      "4.0 4907\n",
      "5.0 4908\n",
      "4.0 4909\n",
      "2.0 4910\n",
      "5.0 4911\n",
      "5.0 4912\n",
      "5.0 4913\n",
      "5.0 4914\n",
      "2.0 4915\n",
      "1.0 4916\n",
      "5.0 4917\n",
      "5.0 4918\n",
      "1.0 4919\n",
      "5.0 4920\n",
      "5.0 4921\n",
      "5.0 4922\n",
      "5.0 4923\n",
      "5.0 4924\n",
      "5.0 4925\n",
      "5.0 4926\n",
      "1.0 4927\n",
      "5.0 4928\n",
      "4.0 4929\n",
      "4.0 4930\n",
      "5.0 4931\n",
      "3.0 4932\n",
      "2.0 4933\n",
      "1.0 4934\n",
      "3.0 4935\n",
      "4.0 4936\n",
      "1.0 4937\n",
      "1.0 4938\n",
      "4.0 4939\n",
      "5.0 4940\n",
      "4.0 4941\n",
      "5.0 4942\n",
      "5.0 4943\n",
      "1.0 4944\n",
      "5.0 4945\n",
      "5.0 4946\n",
      "4.0 4947\n",
      "4.0 4948\n",
      "3.0 4949\n",
      "1.0 4950\n",
      "5.0 4951\n",
      "1.0 4952\n",
      "5.0 4953\n",
      "5.0 4954\n",
      "4.0 4955\n",
      "1.0 4956\n",
      "5.0 4957\n",
      "3.0 4958\n",
      "5.0 4959\n",
      "1.0 4960\n",
      "1.0 4961\n",
      "4.0 4962\n",
      "1.0 4963\n",
      "5.0 4964\n",
      "3.0 4965\n",
      "4.0 4966\n",
      "1.0 4967\n",
      "5.0 4968\n",
      "2.0 4969\n",
      "1.0 4970\n",
      "1.0 4971\n",
      "4.0 4972\n",
      "1.0 4973\n",
      "1.0 4974\n",
      "4.0 4975\n",
      "1.0 4976\n",
      "1.0 4977\n",
      "2.0 4978\n",
      "5.0 4979\n",
      "5.0 4980\n",
      "5.0 4981\n",
      "4.0 4982\n",
      "5.0 4983\n",
      "1.0 4984\n",
      "5.0 4985\n",
      "2.0 4986\n",
      "1.0 4987\n",
      "1.0 4988\n",
      "3.0 4989\n",
      "1.0 4990\n",
      "5.0 4991\n",
      "3.0 4992\n",
      "1.0 4993\n",
      "4.0 4994\n",
      "1.0 4995\n",
      "4.0 4996\n",
      "4.0 4997\n",
      "4.0 4998\n",
      "5.0 4999\n",
      "3.0 5000\n",
      "1.0 5001\n",
      "4.0 5002\n",
      "1.0 5003\n",
      "1.0 5004\n",
      "3.0 5005\n",
      "1.0 5006\n",
      "1.0 5007\n",
      "5.0 5008\n",
      "5.0 5009\n",
      "5.0 5010\n",
      "1.0 5011\n",
      "1.0 5012\n",
      "1.0 5013\n",
      "1.0 5014\n",
      "3.0 5015\n",
      "5.0 5016\n",
      "4.0 5017\n",
      "3.0 5018\n",
      "1.0 5019\n",
      "5.0 5020\n",
      "4.0 5021\n",
      "5.0 5022\n",
      "1.0 5023\n",
      "4.0 5024\n",
      "5.0 5025\n",
      "5.0 5026\n",
      "5.0 5027\n",
      "1.0 5028\n",
      "1.0 5029\n",
      "5.0 5030\n",
      "5.0 5031\n",
      "1.0 5032\n",
      "4.0 5033\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0 5034\n",
      "1.0 5035\n",
      "4.0 5036\n",
      "1.0 5037\n",
      "4.0 5038\n",
      "4.0 5039\n",
      "1.0 5040\n",
      "4.0 5041\n",
      "4.0 5042\n",
      "5.0 5043\n",
      "5.0 5044\n",
      "4.0 5045\n",
      "1.0 5046\n",
      "1.0 5047\n",
      "5.0 5048\n",
      "2.0 5049\n",
      "2.0 5050\n",
      "5.0 5051\n",
      "5.0 5052\n",
      "5.0 5053\n",
      "1.0 5054\n",
      "2.0 5055\n",
      "4.0 5056\n",
      "5.0 5057\n",
      "2.0 5058\n",
      "1.0 5059\n",
      "1.0 5060\n",
      "1.0 5061\n",
      "5.0 5062\n",
      "1.0 5063\n",
      "1.0 5064\n",
      "4.0 5065\n",
      "5.0 5066\n",
      "1.0 5067\n",
      "1.0 5068\n",
      "1.0 5069\n",
      "1.0 5070\n",
      "5.0 5071\n",
      "4.0 5072\n",
      "4.0 5073\n",
      "5.0 5074\n",
      "4.0 5075\n",
      "5.0 5076\n",
      "4.0 5077\n",
      "1.0 5078\n",
      "2.0 5079\n",
      "3.0 5080\n",
      "5.0 5081\n",
      "2.0 5082\n",
      "1.0 5083\n",
      "1.0 5084\n",
      "3.0 5085\n",
      "1.0 5086\n",
      "5.0 5087\n",
      "3.0 5088\n",
      "1.0 5089\n",
      "5.0 5090\n",
      "1.0 5091\n",
      "5.0 5092\n",
      "5.0 5093\n",
      "5.0 5094\n",
      "1.0 5095\n",
      "4.0 5096\n",
      "1.0 5097\n",
      "3.0 5098\n",
      "5.0 5099\n",
      "1.0 5100\n",
      "5.0 5101\n",
      "1.0 5102\n",
      "1.0 5103\n",
      "3.0 5104\n",
      "5.0 5105\n",
      "1.0 5106\n",
      "1.0 5107\n",
      "2.0 5108\n",
      "1.0 5109\n",
      "1.0 5110\n",
      "5.0 5111\n",
      "5.0 5112\n",
      "1.0 5113\n",
      "4.0 5114\n",
      "2.0 5115\n",
      "1.0 5116\n",
      "5.0 5117\n",
      "4.0 5118\n",
      "2.0 5119\n",
      "1.0 5120\n",
      "5.0 5121\n",
      "1.0 5122\n",
      "2.0 5123\n",
      "1.0 5124\n",
      "3.0 5125\n",
      "4.0 5126\n",
      "5.0 5127\n",
      "2.0 5128\n",
      "4.0 5129\n",
      "5.0 5130\n",
      "1.0 5131\n",
      "4.0 5132\n",
      "2.0 5133\n",
      "1.0 5134\n",
      "4.0 5135\n",
      "2.0 5136\n",
      "3.0 5137\n",
      "4.0 5138\n",
      "4.0 5139\n",
      "3.0 5140\n",
      "3.0 5141\n",
      "4.0 5142\n",
      "5.0 5143\n",
      "5.0 5144\n",
      "1.0 5145\n",
      "4.0 5146\n",
      "5.0 5147\n",
      "1.0 5148\n",
      "5.0 5149\n",
      "5.0 5150\n",
      "4.0 5151\n",
      "1.0 5152\n",
      "5.0 5153\n",
      "5.0 5154\n",
      "4.0 5155\n",
      "2.0 5156\n",
      "5.0 5157\n",
      "5.0 5158\n",
      "5.0 5159\n",
      "5.0 5160\n",
      "1.0 5161\n",
      "3.0 5162\n",
      "1.0 5163\n",
      "2.0 5164\n",
      "3.0 5165\n",
      "5.0 5166\n",
      "5.0 5167\n",
      "5.0 5168\n",
      "4.0 5169\n",
      "5.0 5170\n",
      "1.0 5171\n",
      "1.0 5172\n",
      "5.0 5173\n",
      "1.0 5174\n",
      "1.0 5175\n",
      "4.0 5176\n",
      "4.0 5177\n",
      "5.0 5178\n",
      "3.0 5179\n",
      "4.0 5180\n",
      "5.0 5181\n",
      "5.0 5182\n",
      "3.0 5183\n",
      "3.0 5184\n",
      "5.0 5185\n",
      "4.0 5186\n",
      "4.0 5187\n",
      "4.0 5188\n",
      "5.0 5189\n",
      "5.0 5190\n",
      "3.0 5191\n",
      "3.0 5192\n",
      "4.0 5193\n",
      "5.0 5194\n",
      "5.0 5195\n",
      "1.0 5196\n",
      "5.0 5197\n",
      "5.0 5198\n",
      "5.0 5199\n",
      "5.0 5200\n",
      "1.0 5201\n",
      "4.0 5202\n",
      "4.0 5203\n",
      "2.0 5204\n",
      "3.0 5205\n",
      "3.0 5206\n",
      "3.0 5207\n",
      "4.0 5208\n",
      "4.0 5209\n",
      "2.0 5210\n",
      "5.0 5211\n",
      "5.0 5212\n",
      "1.0 5213\n",
      "5.0 5214\n",
      "4.0 5215\n",
      "5.0 5216\n",
      "1.0 5217\n",
      "5.0 5218\n",
      "5.0 5219\n",
      "4.0 5220\n",
      "2.0 5221\n",
      "2.0 5222\n",
      "1.0 5223\n",
      "5.0 5224\n",
      "3.0 5225\n",
      "3.0 5226\n",
      "1.0 5227\n",
      "1.0 5228\n",
      "4.0 5229\n",
      "2.0 5230\n",
      "3.0 5231\n",
      "3.0 5232\n",
      "5.0 5233\n",
      "4.0 5234\n",
      "5.0 5235\n",
      "3.0 5236\n",
      "3.0 5237\n",
      "5.0 5238\n",
      "3.0 5239\n",
      "2.0 5240\n",
      "4.0 5241\n",
      "1.0 5242\n",
      "5.0 5243\n",
      "1.0 5244\n",
      "4.0 5245\n",
      "5.0 5246\n",
      "4.0 5247\n",
      "4.0 5248\n",
      "5.0 5249\n",
      "5.0 5250\n",
      "5.0 5251\n",
      "3.0 5252\n",
      "3.0 5253\n",
      "4.0 5254\n",
      "3.0 5255\n",
      "5.0 5256\n",
      "4.0 5257\n",
      "5.0 5258\n",
      "4.0 5259\n",
      "5.0 5260\n",
      "4.0 5261\n",
      "5.0 5262\n",
      "5.0 5263\n",
      "5.0 5264\n",
      "5.0 5265\n",
      "5.0 5266\n",
      "4.0 5267\n",
      "5.0 5268\n",
      "1.0 5269\n",
      "3.0 5270\n",
      "4.0 5271\n",
      "3.0 5272\n",
      "1.0 5273\n",
      "1.0 5274\n",
      "5.0 5275\n",
      "5.0 5276\n",
      "5.0 5277\n",
      "5.0 5278\n",
      "1.0 5279\n",
      "4.0 5280\n",
      "4.0 5281\n",
      "1.0 5282\n",
      "1.0 5283\n",
      "4.0 5284\n",
      "5.0 5285\n",
      "4.0 5286\n",
      "4.0 5287\n",
      "3.0 5288\n",
      "1.0 5289\n",
      "5.0 5290\n",
      "5.0 5291\n",
      "5.0 5292\n",
      "1.0 5293\n",
      "1.0 5294\n",
      "5.0 5295\n",
      "3.0 5296\n",
      "4.0 5297\n",
      "4.0 5298\n",
      "4.0 5299\n",
      "5.0 5300\n",
      "3.0 5301\n",
      "1.0 5302\n",
      "5.0 5303\n",
      "3.0 5304\n",
      "5.0 5305\n",
      "1.0 5306\n",
      "5.0 5307\n",
      "4.0 5308\n",
      "1.0 5309\n",
      "1.0 5310\n",
      "5.0 5311\n",
      "4.0 5312\n",
      "4.0 5313\n",
      "5.0 5314\n",
      "5.0 5315\n",
      "5.0 5316\n",
      "5.0 5317\n",
      "5.0 5318\n",
      "5.0 5319\n",
      "4.0 5320\n",
      "4.0 5321\n",
      "3.0 5322\n",
      "5.0 5323\n",
      "5.0 5324\n",
      "5.0 5325\n",
      "5.0 5326\n",
      "5.0 5327\n",
      "4.0 5328\n",
      "1.0 5329\n",
      "4.0 5330\n",
      "4.0 5331\n",
      "5.0 5332\n",
      "4.0 5333\n",
      "5.0 5334\n",
      "4.0 5335\n",
      "1.0 5336\n",
      "1.0 5337\n",
      "3.0 5338\n",
      "5.0 5339\n",
      "4.0 5340\n",
      "3.0 5341\n",
      "1.0 5342\n",
      "1.0 5343\n",
      "2.0 5344\n",
      "5.0 5345\n",
      "1.0 5346\n",
      "4.0 5347\n",
      "4.0 5348\n",
      "5.0 5349\n",
      "5.0 5350\n",
      "4.0 5351\n",
      "5.0 5352\n",
      "4.0 5353\n",
      "1.0 5354\n",
      "2.0 5355\n",
      "5.0 5356\n",
      "3.0 5357\n",
      "4.0 5358\n",
      "5.0 5359\n",
      "1.0 5360\n",
      "5.0 5361\n",
      "3.0 5362\n",
      "3.0 5363\n",
      "3.0 5364\n",
      "5.0 5365\n",
      "5.0 5366\n",
      "4.0 5367\n",
      "4.0 5368\n",
      "3.0 5369\n",
      "5.0 5370\n",
      "1.0 5371\n",
      "2.0 5372\n",
      "5.0 5373\n",
      "2.0 5374\n",
      "1.0 5375\n",
      "4.0 5376\n",
      "4.0 5377\n",
      "5.0 5378\n",
      "4.0 5379\n",
      "1.0 5380\n",
      "1.0 5381\n",
      "5.0 5382\n",
      "3.0 5383\n",
      "3.0 5384\n",
      "1.0 5385\n",
      "5.0 5386\n",
      "1.0 5387\n",
      "3.0 5388\n",
      "5.0 5389\n",
      "1.0 5390\n",
      "5.0 5391\n",
      "2.0 5392\n",
      "1.0 5393\n",
      "5.0 5394\n",
      "1.0 5395\n",
      "1.0 5396\n",
      "4.0 5397\n",
      "1.0 5398\n",
      "3.0 5399\n",
      "5.0 5400\n",
      "5.0 5401\n",
      "1.0 5402\n",
      "1.0 5403\n",
      "4.0 5404\n",
      "1.0 5405\n",
      "5.0 5406\n",
      "4.0 5407\n",
      "4.0 5408\n",
      "5.0 5409\n",
      "5.0 5410\n",
      "1.0 5411\n",
      "4.0 5412\n",
      "3.0 5413\n",
      "1.0 5414\n",
      "2.0 5415\n",
      "5.0 5416\n",
      "5.0 5417\n",
      "4.0 5418\n",
      "5.0 5419\n",
      "3.0 5420\n",
      "4.0 5421\n",
      "5.0 5422\n",
      "3.0 5423\n",
      "5.0 5424\n",
      "5.0 5425\n",
      "5.0 5426\n",
      "1.0 5427\n",
      "3.0 5428\n",
      "4.0 5429\n",
      "3.0 5430\n",
      "1.0 5431\n",
      "1.0 5432\n",
      "4.0 5433\n",
      "4.0 5434\n",
      "2.0 5435\n",
      "1.0 5436\n",
      "5.0 5437\n",
      "1.0 5438\n",
      "2.0 5439\n",
      "1.0 5440\n",
      "5.0 5441\n",
      "1.0 5442\n",
      "1.0 5443\n",
      "4.0 5444\n",
      "1.0 5445\n",
      "5.0 5446\n",
      "1.0 5447\n",
      "5.0 5448\n",
      "4.0 5449\n",
      "4.0 5450\n",
      "5.0 5451\n",
      "4.0 5452\n",
      "5.0 5453\n",
      "5.0 5454\n",
      "4.0 5455\n",
      "5.0 5456\n",
      "5.0 5457\n",
      "1.0 5458\n",
      "4.0 5459\n",
      "1.0 5460\n",
      "1.0 5461\n",
      "1.0 5462\n",
      "5.0 5463\n",
      "1.0 5464\n",
      "1.0 5465\n",
      "2.0 5466\n",
      "5.0 5467\n",
      "3.0 5468\n",
      "5.0 5469\n",
      "5.0 5470\n",
      "3.0 5471\n",
      "5.0 5472\n",
      "4.0 5473\n",
      "5.0 5474\n",
      "5.0 5475\n",
      "1.0 5476\n",
      "5.0 5477\n",
      "4.0 5478\n",
      "4.0 5479\n",
      "1.0 5480\n",
      "1.0 5481\n",
      "5.0 5482\n",
      "4.0 5483\n",
      "3.0 5484\n",
      "4.0 5485\n",
      "5.0 5486\n",
      "5.0 5487\n",
      "1.0 5488\n",
      "2.0 5489\n",
      "4.0 5490\n",
      "1.0 5491\n",
      "3.0 5492\n",
      "4.0 5493\n",
      "1.0 5494\n",
      "5.0 5495\n",
      "5.0 5496\n",
      "5.0 5497\n",
      "4.0 5498\n",
      "5.0 5499\n",
      "3.0 5500\n",
      "3.0 5501\n",
      "5.0 5502\n",
      "3.0 5503\n",
      "5.0 5504\n",
      "5.0 5505\n",
      "5.0 5506\n",
      "1.0 5507\n",
      "1.0 5508\n",
      "4.0 5509\n",
      "4.0 5510\n",
      "4.0 5511\n",
      "4.0 5512\n",
      "4.0 5513\n",
      "4.0 5514\n",
      "5.0 5515\n",
      "5.0 5516\n",
      "4.0 5517\n",
      "3.0 5518\n",
      "5.0 5519\n",
      "5.0 5520\n",
      "5.0 5521\n",
      "3.0 5522\n",
      "3.0 5523\n",
      "5.0 5524\n",
      "5.0 5525\n",
      "5.0 5526\n",
      "5.0 5527\n",
      "3.0 5528\n",
      "5.0 5529\n",
      "4.0 5530\n",
      "1.0 5531\n",
      "5.0 5532\n",
      "4.0 5533\n",
      "5.0 5534\n",
      "4.0 5535\n",
      "4.0 5536\n",
      "3.0 5537\n",
      "5.0 5538\n",
      "4.0 5539\n",
      "4.0 5540\n",
      "4.0 5541\n",
      "4.0 5542\n",
      "4.0 5543\n",
      "3.0 5544\n",
      "3.0 5545\n",
      "5.0 5546\n",
      "4.0 5547\n",
      "4.0 5548\n",
      "4.0 5549\n",
      "1.0 5550\n",
      "1.0 5551\n",
      "2.0 5552\n",
      "5.0 5553\n",
      "5.0 5554\n",
      "4.0 5555\n",
      "1.0 5556\n",
      "5.0 5557\n",
      "5.0 5558\n",
      "4.0 5559\n",
      "5.0 5560\n",
      "5.0 5561\n",
      "5.0 5562\n",
      "5.0 5563\n",
      "5.0 5564\n",
      "3.0 5565\n",
      "3.0 5566\n",
      "4.0 5567\n",
      "4.0 5568\n",
      "4.0 5569\n",
      "4.0 5570\n",
      "4.0 5571\n",
      "1.0 5572\n",
      "4.0 5573\n",
      "5.0 5574\n",
      "5.0 5575\n",
      "5.0 5576\n",
      "5.0 5577\n",
      "5.0 5578\n",
      "5.0 5579\n",
      "3.0 5580\n",
      "4.0 5581\n",
      "4.0 5582\n",
      "1.0 5583\n",
      "4.0 5584\n",
      "1.0 5585\n",
      "4.0 5586\n",
      "4.0 5587\n",
      "5.0 5588\n",
      "1.0 5589\n",
      "3.0 5590\n",
      "5.0 5591\n",
      "5.0 5592\n",
      "1.0 5593\n",
      "4.0 5594\n",
      "1.0 5595\n",
      "1.0 5596\n",
      "5.0 5597\n",
      "4.0 5598\n",
      "4.0 5599\n",
      "2.0 5600\n"
     ]
    }
   ],
   "source": [
    "positive=0\n",
    "negative=0\n",
    "count=0\n",
    "#polarity and star graph \n",
    "\n",
    "for i in df['text']:\n",
    "    blob = TextBlob(i)\n",
    "    feel = blob.sentiment\n",
    "    if feel.polarity > 0:\n",
    "        positive = positive + 1\n",
    "    else:\n",
    "        negative = negative + 1\n",
    "    if(feel.polarity<0.10):\n",
    "        df.loc[count, 'sentiment'] = int(1)\n",
    "    elif(feel.polarity>=0.10 and feel.polarity < 0.16):\n",
    "        df.loc[count, 'sentiment'] = int(2)\n",
    "    elif(feel.polarity>=0.16 and feel.polarity < 0.025):\n",
    "        df.loc[count, 'sentiment'] = int(3)\n",
    "    elif(feel.polarity>=0.025 and feel.polarity < 0.35):\n",
    "        df.loc[count, 'sentiment'] = int(4)\n",
    "    else:\n",
    "        df.loc[count, 'sentiment'] = int(5)\n",
    "    w=df.loc[count, 'stars']\n",
    "#     print(w)\n",
    "    if w=='NaN':\n",
    "        print(i)\n",
    "    print(w,count)\n",
    "    dict1[int(w)].append(feel.polarity)\n",
    "    count=count+1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 532,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>stars</th>\n",
       "      <th>oneword</th>\n",
       "      <th>length</th>\n",
       "      <th>sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [text, stars, oneword, length, sentiment]\n",
       "Index: []"
      ]
     },
     "execution_count": 532,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nan_rows = df[df['stars'].isnull()]\n",
    "nan_rows"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 533,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[0.13333333333333333,\n",
       "  -0.2,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  3.700743415417188e-17,\n",
       "  0.5,\n",
       "  -0.15000000000000002,\n",
       "  0.0,\n",
       "  0.09999999999999999,\n",
       "  -0.625,\n",
       "  0.13636363636363635,\n",
       "  0.0,\n",
       "  0.13333333333333333,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  -0.5999999999999999,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.7,\n",
       "  -0.05000000000000002,\n",
       "  -0.05000000000000002,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.3,\n",
       "  0.1,\n",
       "  0.1,\n",
       "  -0.8499999999999999,\n",
       "  0.4000000000000001,\n",
       "  0.4000000000000001,\n",
       "  -0.125,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.16428571428571426,\n",
       "  0.1,\n",
       "  -0.15,\n",
       "  0.1,\n",
       "  -0.15,\n",
       "  -0.4,\n",
       "  -0.25,\n",
       "  -0.6,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.2583333333333333,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.4,\n",
       "  -0.49999999999999994,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.4,\n",
       "  -0.49999999999999994,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  -0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.3499999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.3499999999999999,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  -0.5,\n",
       "  4.4408920985006264e-17,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.425,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  -0.4,\n",
       "  0.4333333333333333,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  -0.16666666666666666,\n",
       "  -0.2916666666666667,\n",
       "  0.14999999999999997,\n",
       "  0.14999999999999997,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  -0.22222222222222218,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  -1.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.4,\n",
       "  0.2,\n",
       "  -1.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.4,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.7142857142857143,\n",
       "  -0.16666666666666666,\n",
       "  -0.16666666666666666,\n",
       "  0.0,\n",
       "  -0.6,\n",
       "  -0.15000000000000002,\n",
       "  0.05333333333333334,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  -0.25,\n",
       "  0.7,\n",
       "  5.551115123125783e-17,\n",
       "  0.0,\n",
       "  -0.025,\n",
       "  0.0,\n",
       "  -0.4,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.35,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  5.551115123125783e-17,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.65,\n",
       "  0.0,\n",
       "  -0.75,\n",
       "  0.0,\n",
       "  -0.12499999999999994,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.18333333333333332,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.1,\n",
       "  0.0,\n",
       "  0.16,\n",
       "  0.08333333333333333,\n",
       "  0.5,\n",
       "  -0.4,\n",
       "  0.08333333333333333,\n",
       "  0.5,\n",
       "  -0.4,\n",
       "  0.44999999999999996,\n",
       "  0.44999999999999996,\n",
       "  0.0,\n",
       "  -0.6,\n",
       "  0.0,\n",
       "  -0.325,\n",
       "  0.0,\n",
       "  -0.14999999999999994,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.6999999999999998,\n",
       "  -0.6499999999999999,\n",
       "  -0.41538461538461535,\n",
       "  -0.4,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  -0.25,\n",
       "  0.06458333333333334,\n",
       "  0.4,\n",
       "  -0.2888888888888889,\n",
       "  5.551115123125783e-17,\n",
       "  5.551115123125783e-17,\n",
       "  -0.5,\n",
       "  -0.3,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.09999999999999998,\n",
       "  -0.4,\n",
       "  0.09999999999999998,\n",
       "  -0.4,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.35714285714285715,\n",
       "  -0.6,\n",
       "  -0.7666666666666666,\n",
       "  -0.6,\n",
       "  -0.3,\n",
       "  0.7333333333333333,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  0.16666666666666666,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  -0.25,\n",
       "  -0.3,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.3,\n",
       "  0.35714285714285715,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.13333333333333333,\n",
       "  0.2,\n",
       "  0.15625,\n",
       "  -0.1466666666666666,\n",
       "  0.225,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6499999999999999,\n",
       "  -0.5,\n",
       "  -1.0,\n",
       "  -0.25,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.325,\n",
       "  0.0,\n",
       "  -0.19999999999999996,\n",
       "  0.20500000000000002,\n",
       "  -0.6999999999999998,\n",
       "  0.25,\n",
       "  -0.35000000000000003,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.5,\n",
       "  0.0,\n",
       "  -0.6,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  -0.5,\n",
       "  0.1,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.175,\n",
       "  -0.5,\n",
       "  -0.2699999999999999,\n",
       "  -0.6999999999999998,\n",
       "  0.35000000000000003,\n",
       "  0.275,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.3499999999999999,\n",
       "  0.125,\n",
       "  -0.5,\n",
       "  0.7,\n",
       "  -0.5638888888888888,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.35714285714285715,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  -0.19999999999999993,\n",
       "  -0.2749999999999999,\n",
       "  0.25,\n",
       "  -0.3666666666666667,\n",
       "  -0.1,\n",
       "  -0.3,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  -0.19333333333333322,\n",
       "  0.225,\n",
       "  -0.4,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  -0.2,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.18333333333333335,\n",
       "  -0.6999999999999998,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.7142857142857143,\n",
       "  0.7,\n",
       "  -0.32499999999999996,\n",
       "  -0.125,\n",
       "  -0.30000000000000004,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.2857142857142857,\n",
       "  -0.5999999999999999,\n",
       "  0.0,\n",
       "  -0.22499999999999995,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.0634920634920635,\n",
       "  0.7,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.04999999999999999,\n",
       "  -0.7999999999999998,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.3333333333333333,\n",
       "  -0.39583333333333337,\n",
       "  0.2,\n",
       "  0.6,\n",
       "  -0.07142857142857142,\n",
       "  -0.5,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.5857142857142857,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  -0.6999999999999998,\n",
       "  -0.025,\n",
       "  0.4285714285714286,\n",
       "  -1.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.11666666666666668,\n",
       "  -1.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.4,\n",
       "  -0.1,\n",
       "  -0.6999999999999998,\n",
       "  0.7,\n",
       "  -0.8499999999999999,\n",
       "  0.0,\n",
       "  -0.4,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  -0.2916666666666667,\n",
       "  0.35714285714285715,\n",
       "  -0.20000000000000004,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.375,\n",
       "  -0.12142857142857141,\n",
       "  -0.4,\n",
       "  0.0,\n",
       "  -0.23076923076923078,\n",
       "  0.13636363636363635,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  -0.2,\n",
       "  -0.23076923076923078,\n",
       "  -0.5799999999999998,\n",
       "  -0.14583333333333334,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.19999999999999996,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.46666666666666656,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  -1.0,\n",
       "  0.35714285714285715,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.14999999999999994,\n",
       "  -0.05,\n",
       "  0.0,\n",
       "  -0.07500000000000001,\n",
       "  0.7,\n",
       "  -0.6999999999999998,\n",
       "  0.475,\n",
       "  0.7,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.3499999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  -0.16666666666666666,\n",
       "  0.0,\n",
       "  -0.8,\n",
       "  5.551115123125783e-17,\n",
       "  0.13636363636363635,\n",
       "  0.21428571428571427,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.17499999999999996,\n",
       "  -0.6999999999999998,\n",
       "  5.551115123125783e-17,\n",
       "  0.0,\n",
       "  -0.31999999999999995,\n",
       "  0.0,\n",
       "  0.26666666666666666,\n",
       "  0.125,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  0.26666666666666666,\n",
       "  0.0,\n",
       "  -0.6,\n",
       "  -0.35,\n",
       "  -0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.3499999999999999,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  -0.4,\n",
       "  0.7,\n",
       "  5.551115123125783e-17,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.4,\n",
       "  0.2,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.11666666666666668,\n",
       "  0.0,\n",
       "  -0.16666666666666666,\n",
       "  -0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  -0.3499999999999999,\n",
       "  0.4166666666666667,\n",
       "  -0.02500000000000001,\n",
       "  0.23333333333333336,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.4,\n",
       "  0.475,\n",
       "  -0.6499999999999999,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.18333333333333332,\n",
       "  5.551115123125783e-17,\n",
       "  -0.4,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.1,\n",
       "  -0.4833333333333332,\n",
       "  0.125,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.31666666666666665,\n",
       "  0.35,\n",
       "  -0.025000000000000022,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.4000000000000001,\n",
       "  -0.19999999999999998,\n",
       "  0.7,\n",
       "  -0.43333333333333324,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.6999999999999998,\n",
       "  -0.16666666666666666,\n",
       "  -0.24999999999999992,\n",
       "  -0.16666666666666666,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  -0.6999999999999998,\n",
       "  -0.05,\n",
       "  0.0,\n",
       "  0.125,\n",
       "  -1.0,\n",
       "  -0.3375,\n",
       "  -0.1,\n",
       "  0.35,\n",
       "  -0.1,\n",
       "  -0.08333333333333333,\n",
       "  5.551115123125783e-17,\n",
       "  0.35714285714285715,\n",
       "  0.0,\n",
       "  0.35714285714285715,\n",
       "  0.5285714285714286,\n",
       "  0.0,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  -0.4,\n",
       "  0.0,\n",
       "  -0.08333333333333333,\n",
       "  -0.6,\n",
       "  0.0,\n",
       "  -0.3499999999999999,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.2,\n",
       "  -0.012499999999999997,\n",
       "  0.7,\n",
       "  0.06666666666666667,\n",
       "  0.7,\n",
       "  -0.8499999999999999,\n",
       "  0.7,\n",
       "  -0.19999999999999993,\n",
       "  0.08333333333333337,\n",
       "  -0.2,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  -0.3333333333333333,\n",
       "  -0.5,\n",
       "  -0.17499999999999993,\n",
       "  0.25,\n",
       "  -1.0,\n",
       "  0.0,\n",
       "  0.35714285714285715,\n",
       "  -1.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.30000000000000004,\n",
       "  0.0,\n",
       "  -0.2,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  0.35,\n",
       "  -0.3,\n",
       "  -0.16666666666666666,\n",
       "  -0.36249999999999993,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.1,\n",
       "  5.551115123125783e-17,\n",
       "  0.0,\n",
       "  -0.1333333333333333,\n",
       "  -0.22499999999999992,\n",
       "  -0.16666666666666666,\n",
       "  -1.0,\n",
       "  0.5,\n",
       "  -0.11833333333333336,\n",
       "  0.2,\n",
       "  -0.4,\n",
       "  -0.4357142857142856,\n",
       "  -0.6999999999999998,\n",
       "  0.22499999999999998,\n",
       "  3.700743415417188e-17,\n",
       "  -0.3799999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5666666666666667,\n",
       "  5.551115123125783e-17,\n",
       "  0.0,\n",
       "  -0.012499999999999997,\n",
       "  -1.1102230246251566e-17,\n",
       "  0.03333333333333333,\n",
       "  -0.6,\n",
       "  -0.175,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  -0.2916666666666667,\n",
       "  -0.6999999999999998,\n",
       "  -0.02500000000000001,\n",
       "  -0.6,\n",
       "  -0.4,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  -0.6,\n",
       "  -0.44999999999999996,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6071428571428572,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  0.8,\n",
       "  -0.6999999999999998,\n",
       "  -0.6999999999999998,\n",
       "  0.13636363636363635,\n",
       "  0.0,\n",
       "  -0.1499999999999999,\n",
       "  -0.16666666666666666,\n",
       "  0.0,\n",
       "  -0.16333333333333333,\n",
       "  0.0,\n",
       "  -0.3,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.06666666666666672,\n",
       "  0.35,\n",
       "  -0.17499999999999993,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.049999999999999996,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  -0.125,\n",
       "  -1.0,\n",
       "  -0.14999999999999994,\n",
       "  0.7,\n",
       "  -0.3166666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.25,\n",
       "  5.551115123125783e-17,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.75,\n",
       "  0.7,\n",
       "  -0.15,\n",
       "  0.0,\n",
       "  -0.425,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.05952380952380952,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  -0.3,\n",
       "  -0.7142857142857143,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  -1.0,\n",
       "  -0.45,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -1.0,\n",
       "  5.551115123125783e-17,\n",
       "  0.0,\n",
       "  0.125,\n",
       "  0.39999999999999997,\n",
       "  -0.4666666666666666,\n",
       "  -0.4333333333333333,\n",
       "  -0.4,\n",
       "  0.0,\n",
       "  0.7999999999999999,\n",
       "  0.7,\n",
       "  -0.6499999999999999,\n",
       "  -0.4,\n",
       "  -0.08333333333333333,\n",
       "  0.032142857142857126,\n",
       "  -0.2722222222222222,\n",
       "  -0.16666666666666666,\n",
       "  0.04999999999999999,\n",
       "  -0.5499999999999999,\n",
       "  -0.38499999999999995,\n",
       "  0.0],\n",
       " [0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.1,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  -0.09999999999999998,\n",
       "  -0.09999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.8,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.15000000000000002,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.16,\n",
       "  -0.08333333333333333,\n",
       "  0.0,\n",
       "  -0.08333333333333333,\n",
       "  0.0,\n",
       "  0.16,\n",
       "  0.3,\n",
       "  0.275,\n",
       "  0.6,\n",
       "  0.3708333333333333,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.44999999999999996,\n",
       "  0.75,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.05,\n",
       "  0.25,\n",
       "  0.25,\n",
       "  -0.125,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.175,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  -0.3499999999999999,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.4,\n",
       "  0.6,\n",
       "  -0.5,\n",
       "  0.05555555555555555,\n",
       "  0.16,\n",
       "  0.0,\n",
       "  0.43,\n",
       "  0.43,\n",
       "  0.35,\n",
       "  -0.6999999999999998,\n",
       "  -0.125,\n",
       "  0.7,\n",
       "  0.06250000000000004,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  -0.11666666666666668,\n",
       "  0.1333333333333333,\n",
       "  0.0,\n",
       "  -0.4749999999999999,\n",
       "  0.35,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  -0.03125,\n",
       "  -0.08333333333333333,\n",
       "  0.0,\n",
       "  0.26249999999999996,\n",
       "  0.15,\n",
       "  -0.0666666666666667,\n",
       "  0.25,\n",
       "  0.48333333333333334,\n",
       "  0.7000000000000001,\n",
       "  0.7,\n",
       "  -0.2,\n",
       "  0.0,\n",
       "  -0.3,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.08166666666666668,\n",
       "  0.47857142857142854,\n",
       "  0.0,\n",
       "  -0.75,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.5249999999999999,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  -0.7142857142857143,\n",
       "  -0.3333333333333333,\n",
       "  -0.6999999999999998,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  -0.4333333333333332,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.125,\n",
       "  0.4000000000000001,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  -0.6,\n",
       "  0.35,\n",
       "  -0.30000000000000004,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  3.700743415417188e-17,\n",
       "  -0.45,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  -0.16666666666666666,\n",
       "  0.475,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.35,\n",
       "  -0.16666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.4,\n",
       "  0.4266666666666667,\n",
       "  0.16666666666666666,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  0.5375,\n",
       "  0.7,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  5.551115123125783e-17,\n",
       "  -0.30000000000000004,\n",
       "  0.1,\n",
       "  0.0,\n",
       "  0.2333333333333333,\n",
       "  0.65,\n",
       "  -0.6,\n",
       "  0.5166666666666666,\n",
       "  0.2,\n",
       "  0.13142857142857145,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.10000000000000003,\n",
       "  0.35,\n",
       "  -0.09523809523809522,\n",
       "  -0.6999999999999998,\n",
       "  -0.06666666666666661,\n",
       "  -3.700743415417188e-17,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.19999999999999998,\n",
       "  5.551115123125783e-17,\n",
       "  0.2333333333333333,\n",
       "  0.2,\n",
       "  0.35,\n",
       "  0.3,\n",
       "  0.16,\n",
       "  -0.4,\n",
       "  0.35,\n",
       "  0.225,\n",
       "  -0.5,\n",
       "  0.0,\n",
       "  0.10625000000000001,\n",
       "  0.07999999999999999,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.05,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -1.0,\n",
       "  0.0],\n",
       " [0.8,\n",
       "  0.7,\n",
       "  -0.125,\n",
       "  0.0,\n",
       "  -0.125,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.1,\n",
       "  0.29545454545454547,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.20666666666666664,\n",
       "  0.61,\n",
       "  0.5249999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  0.23116883116883116,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  0.2,\n",
       "  0.32499999999999996,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.4,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.2333333333333333,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6000000000000001,\n",
       "  0.3,\n",
       "  0.2333333333333333,\n",
       "  0.3,\n",
       "  0.2333333333333333,\n",
       "  0.6333333333333333,\n",
       "  0.0,\n",
       "  -0.6,\n",
       "  0.2,\n",
       "  0.2,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.024999999999999994,\n",
       "  0.14999999999999997,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.19318181818181818,\n",
       "  0.16,\n",
       "  0.06666666666666665,\n",
       "  5.551115123125783e-17,\n",
       "  0.06666666666666665,\n",
       "  5.551115123125783e-17,\n",
       "  -0.25,\n",
       "  -0.25,\n",
       "  -0.25000000000000006,\n",
       "  -0.25000000000000006,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.1,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.15,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.475,\n",
       "  0.7,\n",
       "  0.2833333333333334,\n",
       "  0.7,\n",
       "  0.2833333333333334,\n",
       "  0.35555555555555557,\n",
       "  0.55,\n",
       "  0.425,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.5833333333333334,\n",
       "  0.5833333333333334,\n",
       "  0.7,\n",
       "  0.7083333333333335,\n",
       "  0.7083333333333335,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.35555555555555557,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.3833333333333333,\n",
       "  0.0,\n",
       "  0.19166666666666665,\n",
       "  -0.05,\n",
       "  0.0,\n",
       "  -0.05,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.4625,\n",
       "  0.4625,\n",
       "  -0.16666666666666666,\n",
       "  0.1,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.41818181818181815,\n",
       "  0.6,\n",
       "  0.35833333333333334,\n",
       "  0.35833333333333334,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.8,\n",
       "  0.31666666666666665,\n",
       "  0.7,\n",
       "  0.4333333333333333,\n",
       "  0.625,\n",
       "  0.45,\n",
       "  0.7,\n",
       "  0.3666666666666667,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.39999999999999997,\n",
       "  0.7,\n",
       "  0.49999999999999994,\n",
       "  0.2,\n",
       "  0.8,\n",
       "  -0.12222222222222219,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.575,\n",
       "  0.6,\n",
       "  0.575,\n",
       "  0.30999999999999994,\n",
       "  0.30999999999999994,\n",
       "  0.7,\n",
       "  0.4,\n",
       "  0.1,\n",
       "  0.1,\n",
       "  0.24999999999999997,\n",
       "  0.225,\n",
       "  0.6749999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.51,\n",
       "  0.1,\n",
       "  0.18,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.24999999999999994,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.6,\n",
       "  0.65,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.3,\n",
       "  0.6333333333333333,\n",
       "  0.7,\n",
       "  0.525,\n",
       "  0.4333333333333333,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.16666666666666666,\n",
       "  -0.05000000000000002,\n",
       "  0.4,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.13333333333333333,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.16666666666666666,\n",
       "  -0.3,\n",
       "  -0.3499999999999999,\n",
       "  0.33749999999999997,\n",
       "  0.475,\n",
       "  0.65,\n",
       "  0.3,\n",
       "  0.32333333333333336,\n",
       "  0.625,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  0.5666666666666667,\n",
       "  0.1481818181818182,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.4,\n",
       "  -0.17499999999999993,\n",
       "  -0.05,\n",
       "  0.4285714285714286,\n",
       "  0.7,\n",
       "  0.10000000000000005,\n",
       "  0.16,\n",
       "  0.7,\n",
       "  0.6749999999999999,\n",
       "  -0.05,\n",
       "  0.27499999999999997,\n",
       "  0.375,\n",
       "  0.7333333333333334,\n",
       "  0.2846153846153846,\n",
       "  0.5,\n",
       "  0.6499999999999999,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.275,\n",
       "  -0.050000000000000044,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.03333333333333329,\n",
       "  0.3,\n",
       "  0.1611111111111111,\n",
       "  0.0,\n",
       "  0.3416666666666667,\n",
       "  0.35,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.6590909090909092,\n",
       "  5.551115123125783e-17,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.75,\n",
       "  0.37,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.39166666666666666,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  -0.05,\n",
       "  0.25,\n",
       "  0.7333333333333334,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.6499999999999999,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.16395833333333334,\n",
       "  0.4333333333333333,\n",
       "  0.0,\n",
       "  -0.1,\n",
       "  0.35,\n",
       "  -0.09999999999999996,\n",
       "  0.475,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.31,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.75,\n",
       "  0.0,\n",
       "  0.27499999999999997,\n",
       "  1.0,\n",
       "  0.4000000000000001,\n",
       "  0.2666666666666666,\n",
       "  0.45555555555555555,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  0.16666666666666666,\n",
       "  0.2,\n",
       "  0.1166666666666667,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.25,\n",
       "  0.3,\n",
       "  0.5666666666666667,\n",
       "  0.16666666666666666,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.35714285714285715,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  -0.4749999999999999,\n",
       "  0.7,\n",
       "  -0.6,\n",
       "  0.7,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.10000000000000003,\n",
       "  0.7,\n",
       "  0.40238095238095234,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.6666666666666666,\n",
       "  0.7,\n",
       "  -0.6499999999999999,\n",
       "  0.4166666666666667,\n",
       "  0.7,\n",
       "  1.0,\n",
       "  0.5777777777777777,\n",
       "  0.8,\n",
       "  0.35,\n",
       "  0.7,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.44999999999999996,\n",
       "  0.35,\n",
       "  0.38999999999999996,\n",
       "  0.0,\n",
       "  0.26666666666666666,\n",
       "  0.0,\n",
       "  0.16666666666666666,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.26666666666666666,\n",
       "  -0.14583333333333334,\n",
       "  0.32222222222222224,\n",
       "  0.7,\n",
       "  0.14285714285714285,\n",
       "  0.22499999999999998,\n",
       "  0.7,\n",
       "  0.15,\n",
       "  0.6,\n",
       "  0.15,\n",
       "  0.25,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  -0.30952380952380953,\n",
       "  0.55,\n",
       "  0.0,\n",
       "  0.4533333333333333,\n",
       "  -0.08333333333333333,\n",
       "  0.26666666666666666,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.44999999999999996,\n",
       "  0.7,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.3333333333333333,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.35,\n",
       "  0.19999999999999996,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  0.3,\n",
       "  0.6,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  -1.0,\n",
       "  0.7,\n",
       "  0.6333333333333333,\n",
       "  0.6999999999999998,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.16,\n",
       "  0.6333333333333333,\n",
       "  0.6999999999999998,\n",
       "  0.5666666666666667,\n",
       "  -0.6999999999999998,\n",
       "  0.6499999999999999,\n",
       "  0.175,\n",
       "  0.44999999999999996,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.2916666666666667,\n",
       "  -0.3,\n",
       "  0.0],\n",
       " [0.4666666666666666,\n",
       "  -0.05,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6055194805194805,\n",
       "  0.5,\n",
       "  0.8,\n",
       "  -0.05,\n",
       "  0.09999999999999998,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.20000000000000004,\n",
       "  0.13636363636363635,\n",
       "  -0.25,\n",
       "  0.7,\n",
       "  0.5285714285714286,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.65,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.5,\n",
       "  0.5,\n",
       "  0.35714285714285715,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.575,\n",
       "  -0.07500000000000001,\n",
       "  0.19999999999999998,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.15,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.44999999999999996,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  -0.03333333333333331,\n",
       "  0.425,\n",
       "  0.5285714285714286,\n",
       "  0.5121212121212121,\n",
       "  0.0,\n",
       "  0.5121212121212121,\n",
       "  0.0,\n",
       "  0.9,\n",
       "  0.75,\n",
       "  0.09999999999999998,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  0.39999999999999997,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  1.0,\n",
       "  0.08333333333333337,\n",
       "  0.55,\n",
       "  0.15357142857142858,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.4166666666666667,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  0.4,\n",
       "  0.125,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  -0.05,\n",
       "  -0.13888888888888887,\n",
       "  -0.25,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  -0.05,\n",
       "  0.3,\n",
       "  0.55,\n",
       "  0.8,\n",
       "  0.5666666666666667,\n",
       "  0.5428571428571429,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0625,\n",
       "  0.6,\n",
       "  0.35714285714285715,\n",
       "  0.7,\n",
       "  0.6333333333333333,\n",
       "  0.7,\n",
       "  0.8,\n",
       "  -0.05,\n",
       "  0.0,\n",
       "  0.2857142857142857,\n",
       "  0.3181818181818182,\n",
       "  0.05,\n",
       "  0.75,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5666666666666667,\n",
       "  0.0,\n",
       "  0.5666666666666667,\n",
       "  0.9,\n",
       "  0.9,\n",
       "  0.25,\n",
       "  0.7,\n",
       "  0.25,\n",
       "  0.7,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.125,\n",
       "  0.0,\n",
       "  0.27499999999999997,\n",
       "  0.13636363636363635,\n",
       "  0.125,\n",
       "  0.0,\n",
       "  0.27499999999999997,\n",
       "  0.13636363636363635,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.15,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.1,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  0.3333333333333333,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.33888888888888885,\n",
       "  0.2583333333333333,\n",
       "  0.35,\n",
       "  -0.08333333333333333,\n",
       "  0.3375,\n",
       "  0.0,\n",
       "  -0.08333333333333333,\n",
       "  0.3375,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.25,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.3523809523809524,\n",
       "  0.5222222222222223,\n",
       "  -0.16666666666666674,\n",
       "  0.3523809523809524,\n",
       "  0.5222222222222223,\n",
       "  -0.16666666666666674,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  -0.05,\n",
       "  -0.25,\n",
       "  0.41818181818181815,\n",
       "  0.7,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.7,\n",
       "  0.16666666666666663,\n",
       "  0.25,\n",
       "  0.7,\n",
       "  0.5,\n",
       "  -0.25,\n",
       "  0.5,\n",
       "  -0.25,\n",
       "  0.35714285714285715,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.4166666666666667,\n",
       "  0.8,\n",
       "  0.16,\n",
       "  -0.25,\n",
       "  0.3833333333333333,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.4333333333333333,\n",
       "  0.35,\n",
       "  -0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7666666666666666,\n",
       "  0.0,\n",
       "  0.1875,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.225,\n",
       "  0.6666666666666666,\n",
       "  0.75,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5454545454545454,\n",
       "  0.0,\n",
       "  0.16666666666666666,\n",
       "  0.55,\n",
       "  0.85,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.421590909090909,\n",
       "  0.5249999999999999,\n",
       "  0.421590909090909,\n",
       "  0.5249999999999999,\n",
       "  0.13636363636363635,\n",
       "  0.3,\n",
       "  0.6,\n",
       "  -0.5,\n",
       "  -0.5,\n",
       "  0.11666666666666671,\n",
       "  0.2857142857142857,\n",
       "  0.8,\n",
       "  -0.05,\n",
       "  0.0,\n",
       "  0.9,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.6,\n",
       "  0.35,\n",
       "  0.6,\n",
       "  -0.05555555555555556,\n",
       "  0.16666666666666666,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.2857142857142857,\n",
       "  0.5,\n",
       "  0.475,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  0.7,\n",
       "  0.1,\n",
       "  0.06666666666666665,\n",
       "  0.3,\n",
       "  0.7,\n",
       "  -0.25,\n",
       "  0.1,\n",
       "  0.7,\n",
       "  0.13636363636363635,\n",
       "  0.0,\n",
       "  0.75,\n",
       "  0.15,\n",
       "  0.3,\n",
       "  0.21666666666666667,\n",
       "  0.6999999999999998,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.5,\n",
       "  0.16666666666666666,\n",
       "  -0.25,\n",
       "  0.19318181818181818,\n",
       "  0.19318181818181818,\n",
       "  0.7,\n",
       "  0.04999999999999993,\n",
       "  -0.15,\n",
       "  -0.15,\n",
       "  0.4000000000000001,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.3,\n",
       "  0.8,\n",
       "  0.3,\n",
       "  0.475,\n",
       "  0.7,\n",
       "  0.175,\n",
       "  0.5,\n",
       "  -0.175,\n",
       "  0.0,\n",
       "  0.5199999999999999,\n",
       "  -0.6,\n",
       "  0.6833333333333332,\n",
       "  0.7250000000000001,\n",
       "  -0.16666666666666666,\n",
       "  0.0,\n",
       "  0.7333333333333334,\n",
       "  0.7250000000000001,\n",
       "  -0.16666666666666666,\n",
       "  0.0,\n",
       "  0.7333333333333334,\n",
       "  0.16,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.16,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.16666666666666666,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.11818181818181818,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.16666666666666666,\n",
       "  0.7333333333333334,\n",
       "  0.0,\n",
       "  0.16666666666666666,\n",
       "  0.7333333333333334,\n",
       "  0.3,\n",
       "  0.25,\n",
       "  0.6499999999999999,\n",
       "  0.3,\n",
       "  0.25,\n",
       "  0.6499999999999999,\n",
       "  0.3333333333333333,\n",
       "  0.8,\n",
       "  -0.4,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  -0.3125,\n",
       "  0.0,\n",
       "  -0.3125,\n",
       "  -0.05,\n",
       "  0.8,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.8,\n",
       "  0.5,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.14285714285714285,\n",
       "  0.3,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.7666666666666666,\n",
       "  0.5523809523809523,\n",
       "  0.15555555555555559,\n",
       "  0.0,\n",
       "  0.7333333333333334,\n",
       "  0.7333333333333334,\n",
       "  0.6499999999999999,\n",
       "  0.6499999999999999,\n",
       "  0.2,\n",
       "  -0.05,\n",
       "  0.7,\n",
       "  0.5,\n",
       "  0.5166666666666666,\n",
       "  0.5888888888888889,\n",
       "  0.0,\n",
       "  0.76,\n",
       "  0.7,\n",
       "  0.5833333333333334,\n",
       "  -0.2,\n",
       "  0.7,\n",
       "  0.20000000000000004,\n",
       "  0.30000000000000004,\n",
       "  0.4333333333333333,\n",
       "  0.7,\n",
       "  0.75,\n",
       "  0.6,\n",
       "  0.6,\n",
       "  -0.4,\n",
       "  -0.16249999999999998,\n",
       "  0.7333333333333334,\n",
       "  0.35,\n",
       "  0.5555555555555555,\n",
       "  0.6,\n",
       "  0.75,\n",
       "  0.6,\n",
       "  0.75,\n",
       "  0.7,\n",
       "  0.2333333333333333,\n",
       "  0.2333333333333333,\n",
       "  0.5499999999999999,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.5166666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.3857142857142857,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.3857142857142857,\n",
       "  0.5066666666666667,\n",
       "  0.0,\n",
       "  0.55,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.5121212121212121,\n",
       "  0.3,\n",
       "  0.7,\n",
       "  0.45333333333333325,\n",
       "  0.7,\n",
       "  0.3035714285714286,\n",
       "  0.55,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.75,\n",
       "  0.43333333333333335,\n",
       "  0.4,\n",
       "  0.75,\n",
       "  0.75,\n",
       "  0.35714285714285715,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5249999999999999,\n",
       "  0.5833333333333333,\n",
       "  0.0,\n",
       "  0.7333333333333334,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.25,\n",
       "  0.5166666666666666,\n",
       "  0.35,\n",
       "  0.35,\n",
       "  0.6,\n",
       "  0.12222222222222223,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.48333333333333334,\n",
       "  -0.30000000000000004,\n",
       "  0.09999999999999998,\n",
       "  0.75,\n",
       "  0.6499999999999999,\n",
       "  -0.2166666666666666,\n",
       "  0.25,\n",
       "  0.41111111111111104,\n",
       "  0.44999999999999996,\n",
       "  0.20416666666666664,\n",
       "  -0.30000000000000004,\n",
       "  0.44285714285714284,\n",
       "  0.7,\n",
       "  -0.16666666666666666,\n",
       "  -0.16666666666666666,\n",
       "  0.26785714285714285,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.4,\n",
       "  0.7,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.4333333333333333,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.23333333333333336,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  5.551115123125783e-17,\n",
       "  0.4333333333333333,\n",
       "  0.4333333333333333,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.6833333333333332,\n",
       "  0.6833333333333332,\n",
       "  0.7,\n",
       "  0.225,\n",
       "  0.6083333333333333,\n",
       "  0.3,\n",
       "  0.7,\n",
       "  0.225,\n",
       "  0.6083333333333333,\n",
       "  0.3,\n",
       "  0.7,\n",
       "  0.16666666666666666,\n",
       "  0.16666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.05,\n",
       "  0.42708333333333326,\n",
       "  0.9,\n",
       "  0.7333333333333333,\n",
       "  0.4333333333333333,\n",
       "  0.08,\n",
       "  0.33,\n",
       "  0.11249999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.3174242424242424,\n",
       "  0.05,\n",
       "  0.7,\n",
       "  0.9,\n",
       "  0.5642857142857143,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.25,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.6333333333333333,\n",
       "  0.5249999999999999,\n",
       "  0.6,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.5249999999999999,\n",
       "  0.0,\n",
       "  0.47857142857142854,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.75,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.6999999999999998,\n",
       "  0.5,\n",
       "  0.35714285714285715,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.41111111111111115,\n",
       "  0.2864285714285714,\n",
       "  0.5666666666666668,\n",
       "  0.25,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.2527777777777777,\n",
       "  0.39999999999999997,\n",
       "  0.5,\n",
       "  0.4,\n",
       "  0.7,\n",
       "  -0.04166666666666667,\n",
       "  0.65,\n",
       "  0.2,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.6071428571428571,\n",
       "  0.43333333333333335,\n",
       "  0.26545454545454544,\n",
       "  0.21666666666666665,\n",
       "  0.8,\n",
       "  0.35,\n",
       "  0.65,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.75,\n",
       "  0.35333333333333333,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.5,\n",
       "  0.6499999999999999,\n",
       "  0.25,\n",
       "  0.4499999999999999,\n",
       "  0.2,\n",
       "  0.7,\n",
       "  0.6599999999999999,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.06818181818181818,\n",
       "  0.0,\n",
       "  0.22499999999999998,\n",
       "  0.5,\n",
       "  0.26666666666666666,\n",
       "  0.75,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.025,\n",
       "  0.4666666666666666,\n",
       "  0.7333333333333333,\n",
       "  0.7333333333333334,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6416666666666666,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.6,\n",
       "  0.4,\n",
       "  0.7,\n",
       "  0.5,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.4166666666666667,\n",
       "  0.0,\n",
       "  0.15333333333333332,\n",
       "  0.0,\n",
       "  0.4333333333333333,\n",
       "  0.7,\n",
       "  0.05833333333333333,\n",
       "  0.0,\n",
       "  -0.04999999999999999,\n",
       "  0.5874999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7666666666666666,\n",
       "  0.6499999999999999,\n",
       "  -0.6,\n",
       "  0.9,\n",
       "  -0.08181818181818182,\n",
       "  0.4666666666666666,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.3333333333333333,\n",
       "  0.0,\n",
       "  0.36818181818181817,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.1,\n",
       "  0.4666666666666667,\n",
       "  0.5166666666666666,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.6666666666666666,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.5199999999999999,\n",
       "  0.55,\n",
       "  -0.6999999999999998,\n",
       "  0.04999999999999999,\n",
       "  0.8,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.6777777777777777,\n",
       "  0.35,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.525,\n",
       "  0.55,\n",
       "  0.6,\n",
       "  0.4333333333333333,\n",
       "  0.4999999999999999,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7666666666666666,\n",
       "  0.42857142857142855,\n",
       "  0.0,\n",
       "  -0.5000000000000001,\n",
       "  0.38888888888888884,\n",
       "  0.9,\n",
       "  0.6,\n",
       "  -0.17857142857142858,\n",
       "  0.125,\n",
       "  0.4000000000000001,\n",
       "  0.04999999999999999,\n",
       "  0.0,\n",
       "  0.39999999999999997,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  0.35,\n",
       "  0.7333333333333334,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.32499999999999996,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  -0.06250000000000001,\n",
       "  0.7,\n",
       "  0.25,\n",
       "  0.48333333333333334,\n",
       "  0.6499999999999999,\n",
       "  0.7,\n",
       "  -0.16666666666666666,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  -0.3,\n",
       "  0.27878787878787875,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.44999999999999996,\n",
       "  0.7,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.7500000000000001,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.09999999999999998,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.3333333333333333,\n",
       "  -0.03333333333333335,\n",
       "  0.7666666666666666,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.2,\n",
       "  0.07500000000000001,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.75,\n",
       "  0.0,\n",
       "  0.6999999999999998,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.6,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  0.3333333333333333,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.475,\n",
       "  0.13636363636363635,\n",
       "  0.30277777777777776,\n",
       "  0.7,\n",
       "  -0.25,\n",
       "  0.6499999999999999,\n",
       "  0.6499999999999999,\n",
       "  0.4000000000000001,\n",
       "  0.35,\n",
       "  0.5047619047619046,\n",
       "  0.4,\n",
       "  0.35,\n",
       "  0.5249999999999999,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.4454545454545455,\n",
       "  0.6133333333333333,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.5666666666666667,\n",
       "  0.0,\n",
       "  0.45416666666666666,\n",
       "  0.5166666666666666,\n",
       "  0.7,\n",
       "  0.25,\n",
       "  0.16,\n",
       "  0.16666666666666666,\n",
       "  0.8,\n",
       "  0.6,\n",
       "  0.6999999999999998,\n",
       "  0.0,\n",
       "  0.4,\n",
       "  0.6,\n",
       "  0.7999999999999999,\n",
       "  0.35,\n",
       "  -0.026666666666666637,\n",
       "  0.6,\n",
       "  0.8,\n",
       "  0.44999999999999996,\n",
       "  0.0,\n",
       "  0.4333333333333333,\n",
       "  0.0,\n",
       "  0.39166666666666666,\n",
       "  0.8,\n",
       "  5.551115123125783e-17,\n",
       "  0.0,\n",
       "  0.1,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.07499999999999998,\n",
       "  0.3,\n",
       "  0.12222222222222222,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.3,\n",
       "  0.4681818181818182,\n",
       "  0.3333333333333333,\n",
       "  0.0,\n",
       "  0.4000000000000001,\n",
       "  -0.4,\n",
       "  0.7,\n",
       "  0.6,\n",
       "  0.8,\n",
       "  -0.16666666666666666,\n",
       "  0.04166666666666666,\n",
       "  0.4166666666666667,\n",
       "  0.33333333333333326,\n",
       "  0.6999999999999998,\n",
       "  0.7,\n",
       "  0.4166666666666667,\n",
       "  0.4681818181818182,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.75,\n",
       "  0.7,\n",
       "  0.35,\n",
       "  0.7,\n",
       "  0.4,\n",
       "  0.5333333333333333,\n",
       "  0.8,\n",
       "  0.6999999999999998,\n",
       "  0.49999999999999994,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.07840909090909087,\n",
       "  0.44999999999999996,\n",
       "  0.0,\n",
       "  0.21212121212121213,\n",
       "  0.6499999999999999,\n",
       "  0.6666666666666666,\n",
       "  0.05000000000000001,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.4,\n",
       "  0.5666666666666667,\n",
       "  0.0,\n",
       "  0.1,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.7666666666666666,\n",
       "  0.25,\n",
       "  0.4666666666666666,\n",
       "  0.5666666666666668,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.6000000000000001,\n",
       "  0.7,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.6999999999999998,\n",
       "  0.6121212121212121,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.41818181818181815,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.5166666666666666,\n",
       "  0.2,\n",
       "  0.1333333333333333,\n",
       "  0.24999999999999997,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.525,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  -0.05,\n",
       "  0.3333333333333333,\n",
       "  0.8,\n",
       "  -0.02121212121212122,\n",
       "  0.0,\n",
       "  0.225,\n",
       "  0.6,\n",
       "  0.6,\n",
       "  0.4444444444444444,\n",
       "  0.16666666666666666,\n",
       "  0.2,\n",
       "  0.7,\n",
       "  0.2,\n",
       "  0.31212121212121213,\n",
       "  1.0,\n",
       "  0.19047619047619047,\n",
       "  0.16666666666666669,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.06666666666666665,\n",
       "  0.15000000000000008,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  0.44999999999999996,\n",
       "  0.4000000000000001,\n",
       "  0.22499999999999998,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  -0.05,\n",
       "  ...],\n",
       " [-0.25,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.016666666666666677,\n",
       "  0.6000000000000001,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  1.0,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  -0.025,\n",
       "  0.35,\n",
       "  0.8,\n",
       "  0.5,\n",
       "  1.0,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  -0.025,\n",
       "  0.35,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.675,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  -0.04999999999999999,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  -0.10714285714285715,\n",
       "  0.4,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.75,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.16666666666666666,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.16666666666666666,\n",
       "  0.25,\n",
       "  0.35714285714285715,\n",
       "  -0.25,\n",
       "  0.6166666666666667,\n",
       "  0.35714285714285715,\n",
       "  0.0,\n",
       "  -0.6,\n",
       "  0.7,\n",
       "  0.35714285714285715,\n",
       "  0.0,\n",
       "  -0.6,\n",
       "  0.7,\n",
       "  0.8,\n",
       "  0.41428571428571426,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  -0.02777777777777778,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.20000000000000004,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.20000000000000004,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.32500000000000007,\n",
       "  -0.625,\n",
       "  0.4333333333333333,\n",
       "  0.11212121212121214,\n",
       "  0.3333333333333333,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.3333333333333333,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.1,\n",
       "  0.5,\n",
       "  0.6,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.7666666666666666,\n",
       "  0.38749999999999996,\n",
       "  0.6999999999999998,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  -0.05,\n",
       "  -0.25,\n",
       "  0.6999999999999998,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  -0.05,\n",
       "  -0.25,\n",
       "  0.6,\n",
       "  0.65,\n",
       "  0.4,\n",
       "  0.3,\n",
       "  0.75,\n",
       "  0.6541666666666667,\n",
       "  0.47878787878787876,\n",
       "  0.6999999999999998,\n",
       "  0.475,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.16666666666666666,\n",
       "  1.0,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.09444444444444444,\n",
       "  0.25,\n",
       "  0.16666666666666666,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  0.16666666666666666,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.1,\n",
       "  0.0,\n",
       "  0.6333333333333333,\n",
       "  0.6,\n",
       "  0.3160714285714286,\n",
       "  0.25,\n",
       "  0.8,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.11818181818181818,\n",
       "  0.6,\n",
       "  0.8,\n",
       "  0.25,\n",
       "  -0.05,\n",
       "  0.35714285714285715,\n",
       "  0.35714285714285715,\n",
       "  0.475,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.21666666666666667,\n",
       "  -0.25,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.4000000000000001,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.5785714285714286,\n",
       "  0.7666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  -0.15,\n",
       "  -0.2318181818181818,\n",
       "  0.25,\n",
       "  -0.3,\n",
       "  -0.18181818181818182,\n",
       "  0.13636363636363635,\n",
       "  0.4000000000000001,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.75,\n",
       "  0.375,\n",
       "  0.25,\n",
       "  0.75,\n",
       "  0.375,\n",
       "  0.5,\n",
       "  0.3666666666666667,\n",
       "  0.5,\n",
       "  0.3666666666666667,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.5285714285714286,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.5285714285714286,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  0.75,\n",
       "  0.8,\n",
       "  0.25,\n",
       "  0.7,\n",
       "  0.2,\n",
       "  0.75,\n",
       "  0.8,\n",
       "  0.25,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.29,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.29,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.4333333333333333,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.05833333333333333,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  -0.15,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.5222222222222221,\n",
       "  0.0,\n",
       "  0.85,\n",
       "  0.5222222222222221,\n",
       "  0.0,\n",
       "  0.85,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.05,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.6083333333333334,\n",
       "  0.0,\n",
       "  0.26666666666666666,\n",
       "  0.9,\n",
       "  0.26666666666666666,\n",
       "  0.9,\n",
       "  0.0,\n",
       "  0.19999999999999998,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  -0.2666666666666666,\n",
       "  0.0,\n",
       "  0.9,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  0.35,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.35714285714285715,\n",
       "  0.3,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.35714285714285715,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.3333333333333333,\n",
       "  0.55,\n",
       "  0.55,\n",
       "  0.47878787878787876,\n",
       "  0.2,\n",
       "  0.7,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.16666666666666666,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.16666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.16666666666666666,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  -0.03333333333333333,\n",
       "  -0.25,\n",
       "  -0.03333333333333333,\n",
       "  -0.25,\n",
       "  0.8,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.375,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.35,\n",
       "  0.7000000000000001,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6666666666666666,\n",
       "  0.4,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5249999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.4,\n",
       "  -0.16666666666666666,\n",
       "  0.5285714285714286,\n",
       "  0.08333333333333333,\n",
       "  0.7,\n",
       "  0.6000000000000001,\n",
       "  0.3375,\n",
       "  0.1,\n",
       "  0.4333333333333333,\n",
       "  0.5371428571428571,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.4000000000000001,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.09523809523809523,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.15,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.07,\n",
       "  0.6333333333333333,\n",
       "  0.25,\n",
       "  0.6499999999999999,\n",
       "  0.7,\n",
       "  0.75,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.75,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.5333333333333334,\n",
       "  0.0,\n",
       "  0.8000000000000002,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.1,\n",
       "  0.4666666666666666,\n",
       "  0.44999999999999996,\n",
       "  0.35,\n",
       "  0.7,\n",
       "  0.8,\n",
       "  0.525,\n",
       "  0.31785714285714284,\n",
       "  0.525,\n",
       "  0.31785714285714284,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.13636363636363635,\n",
       "  0.8,\n",
       "  0.7999999999999999,\n",
       "  0.75,\n",
       "  0.6749999999999999,\n",
       "  0.56,\n",
       "  0.6,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.13636363636363635,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.56,\n",
       "  0.6,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.13636363636363635,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  -0.15,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  -0.25,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.4,\n",
       "  0.050000000000000044,\n",
       "  0.6499999999999999,\n",
       "  0.4,\n",
       "  0.7333333333333334,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.5166666666666666,\n",
       "  0.8,\n",
       "  0.8000000000000002,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.2,\n",
       "  0.25,\n",
       "  0.3,\n",
       "  0.1481818181818182,\n",
       "  0.0,\n",
       "  -0.05,\n",
       "  0.0,\n",
       "  0.34545454545454546,\n",
       "  -0.5000000000000001,\n",
       "  0.09523809523809525,\n",
       "  0.0,\n",
       "  0.2875,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6166666666666667,\n",
       "  0.38,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.3,\n",
       "  0.3166666666666667,\n",
       "  1.0,\n",
       "  0.7666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.8666666666666667,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.8666666666666667,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6142857142857142,\n",
       "  0.2,\n",
       "  0.8,\n",
       "  0.39166666666666666,\n",
       "  0.8,\n",
       "  0.36190476190476195,\n",
       "  0.016666666666666666,\n",
       "  0.75,\n",
       "  0.03333333333333333,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.4583333333333333,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7249999999999999,\n",
       "  0.5111111111111112,\n",
       "  0.5833333333333333,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.35714285714285715,\n",
       "  0.2,\n",
       "  0.3333333333333333,\n",
       "  0.8,\n",
       "  0.5,\n",
       "  0.7999999999999999,\n",
       "  -0.2,\n",
       "  0.2,\n",
       "  0.5499999999999999,\n",
       "  0.4833333333333334,\n",
       "  0.65,\n",
       "  0.4166666666666667,\n",
       "  0.75,\n",
       "  0.32999999999999996,\n",
       "  0.8,\n",
       "  0.35,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.4,\n",
       "  0.30000000000000004,\n",
       "  0.3333333333333333,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.4,\n",
       "  0.30000000000000004,\n",
       "  0.3333333333333333,\n",
       "  0.0,\n",
       "  -0.05,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6000000000000001,\n",
       "  0.0,\n",
       "  0.6666666666666666,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.75,\n",
       "  0.024999999999999994,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.19285714285714284,\n",
       "  0.5681818181818181,\n",
       "  0.55,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.6499999999999999,\n",
       "  0.5249999999999999,\n",
       "  0.06818181818181818,\n",
       "  0.0,\n",
       "  -0.2916666666666667,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.8,\n",
       "  0.16666666666666666,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.8,\n",
       "  0.16666666666666666,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  -0.2,\n",
       "  0.45,\n",
       "  0.8,\n",
       "  0.4619047619047619,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.14285714285714285,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.2,\n",
       "  0.0,\n",
       "  0.13636363636363635,\n",
       "  0.6999999999999998,\n",
       "  0.0,\n",
       "  0.475,\n",
       "  0.3,\n",
       "  0.375,\n",
       "  0.0,\n",
       "  0.76,\n",
       "  0.75,\n",
       "  0.3333333333333333,\n",
       "  0.21428571428571427,\n",
       "  0.4571428571428571,\n",
       "  0.62,\n",
       "  0.825,\n",
       "  0.3333333333333333,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.27499999999999997,\n",
       "  0.8,\n",
       "  0.825,\n",
       "  0.3333333333333333,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.27499999999999997,\n",
       "  0.8,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.3,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.5777777777777777,\n",
       "  0.6,\n",
       "  0.5,\n",
       "  0.5,\n",
       "  0.8,\n",
       "  0.6,\n",
       "  0.5,\n",
       "  -0.25,\n",
       "  0.7,\n",
       "  0.4666666666666666,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  0.3333333333333333,\n",
       "  0.7,\n",
       "  0.5285714285714286,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.13636363636363635,\n",
       "  0.14285714285714285,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6375,\n",
       "  0.0,\n",
       "  0.13636363636363635,\n",
       "  0.14285714285714285,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6375,\n",
       "  0.7,\n",
       "  0.37708333333333327,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.16,\n",
       "  0.3,\n",
       "  0.7000000000000001,\n",
       "  0.7000000000000001,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.4666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.4666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.65,\n",
       "  0.0,\n",
       "  0.35714285714285715,\n",
       "  0.0,\n",
       "  0.7333333333333334,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.13636363636363635,\n",
       "  -0.1,\n",
       "  0.7083333333333334,\n",
       "  -0.15,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.08333333333333333,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7083333333333334,\n",
       "  -0.15,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.08333333333333333,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.44999999999999996,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.4,\n",
       "  0.5,\n",
       "  0.5,\n",
       "  0.5787878787878787,\n",
       "  0.0,\n",
       "  0.5166666666666666,\n",
       "  -0.25,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.5166666666666666,\n",
       "  -0.25,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.7444444444444445,\n",
       "  0.8,\n",
       "  0.33,\n",
       "  0.8,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.2958333333333333,\n",
       "  0.7,\n",
       "  -0.0029761904761904934,\n",
       "  1.0,\n",
       "  0.65,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.2857142857142857,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.35,\n",
       "  -0.1,\n",
       "  0.13636363636363635,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.24545454545454545,\n",
       "  0.2857142857142857,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.16666666666666666,\n",
       "  0.31212121212121213,\n",
       "  0.5083333333333333,\n",
       "  -0.09999999999999999,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.425,\n",
       "  0.44999999999999996,\n",
       "  0.8,\n",
       "  -0.06666666666666667,\n",
       "  0.8,\n",
       "  -0.06666666666666667,\n",
       "  0.3,\n",
       "  1.0,\n",
       "  0.9,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  1.0,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.75,\n",
       "  0.7,\n",
       "  0.35,\n",
       "  0.6666666666666666,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.35,\n",
       "  0.0,\n",
       "  0.175,\n",
       "  0.25,\n",
       "  0.625,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.525,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.5333333333333333,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.3,\n",
       "  -0.17000000000000004,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.3,\n",
       "  -0.17000000000000004,\n",
       "  0.0,\n",
       "  0.25,\n",
       "  0.0,\n",
       "  -0.17857142857142858,\n",
       "  0.36190476190476195,\n",
       "  0.0,\n",
       "  0.19318181818181818,\n",
       "  0.15000000000000002,\n",
       "  0.6499999999999999,\n",
       "  0.0,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.25555555555555554,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.0,\n",
       "  0.38571428571428573,\n",
       "  0.35714285714285715,\n",
       "  0.8,\n",
       "  0.5,\n",
       "  0.5,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.6499999999999999,\n",
       "  0.8,\n",
       "  0.3333333333333333,\n",
       "  0.4666666666666666,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.6333333333333333,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8000000000000002,\n",
       "  0.6,\n",
       "  0.7,\n",
       "  -0.6999999999999998,\n",
       "  0.0,\n",
       "  -0.125,\n",
       "  0.5,\n",
       "  -0.3,\n",
       "  0.0,\n",
       "  0.6499999999999999,\n",
       "  0.3181818181818182,\n",
       "  0.5,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.55,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.8,\n",
       "  0.7,\n",
       "  0.5681818181818181,\n",
       "  0.5833333333333334,\n",
       "  0.5,\n",
       "  0.5,\n",
       "  0.75,\n",
       "  0.3,\n",
       "  0.6999999999999998,\n",
       "  0.7,\n",
       "  0.7000000000000001,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.4,\n",
       "  0.0,\n",
       "  0.1,\n",
       "  0.0,\n",
       "  -0.3,\n",
       "  0.45000000000000007,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.1,\n",
       "  0.0,\n",
       "  -0.3,\n",
       "  0.45000000000000007,\n",
       "  0.0,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.0,\n",
       "  0.75,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.2,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.7,\n",
       "  0.7,\n",
       "  0.6,\n",
       "  0.0,\n",
       "  0.3,\n",
       "  0.7,\n",
       "  0.5777777777777777,\n",
       "  0.6666666666666666,\n",
       "  0.8,\n",
       "  0.20000000000000004,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.8,\n",
       "  0.8,\n",
       "  0.2785714285714286,\n",
       "  0.3333333333333333,\n",
       "  0.5833333333333334,\n",
       "  0.8,\n",
       "  0.20000000000000004,\n",
       "  0.0,\n",
       "  1.0,\n",
       "  0.8,\n",
       "  0.8,\n",
       "  0.2785714285714286,\n",
       "  0.3333333333333333,\n",
       "  0.5833333333333334,\n",
       "  0.5499999999999999,\n",
       "  0.9,\n",
       "  ...]]"
      ]
     },
     "execution_count": 533,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "x = [1,2,3,4,5]\n",
    "y = list(dict1.values())\n",
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 534,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.xlabel(\"X-axis\")\n",
    "plt.ylabel(\"Y-axis\")\n",
    "plt.title(\"Polarity Graph\")\n",
    "for i in range(len(y)):\n",
    "    plt.plot(x,[pt[i] for pt in y],label = 'id %s'%(i+1))\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### REASON FOR BALANCING CAN BE SEEN BY GRAPH SHOWING VARIATION IN COLLECTION OF RATING COUNT ###$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 535,
   "metadata": {},
   "outputs": [],
   "source": [
    "# stars_1 = df[df['stars']==1]\n",
    "# stars_2 = df[df['stars']==2]\n",
    "# stars_3 = df[df['stars']==3]\n",
    "# stars_5 = df[df['stars']==5]\n",
    "# stars_4 = df[df['stars']==4]\n",
    "\n",
    "# df_minority_upsampled_1 = resample(stars_1, \n",
    "#                                  replace=True, \n",
    "#                                  n_samples=3526,\n",
    "#                                  random_state=123)\n",
    "# df_minority_upsampled_2 = resample(stars_2, \n",
    "#                                  replace=True, \n",
    "#                                  n_samples=3526,\n",
    "#                                  random_state=123)\n",
    "# df_minority_upsampled_3 = resample(stars_3, \n",
    "#                                  replace=True, \n",
    "#                                  n_samples=3526,\n",
    "#                                  random_state=123)\n",
    "# df_minority_upsampled_5 = resample(stars_5, \n",
    "#                                  replace=True, \n",
    "#                                  n_samples=3526,\n",
    "#                                  random_state=123)\n",
    "# df_upsampled = pd.concat([stars_4, df_minority_upsampled_1,df_minority_upsampled_2,df_minority_upsampled_3,df_minority_upsampled_5])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "we’ll get less biased predictions if we train the system on balanced data. This means that ideally we should have the same number of examples of each review type."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 536,
   "metadata": {},
   "outputs": [],
   "source": [
    "# N_star_categories = 5\n",
    "# colors = np.array(['#660000', '#ff4d4d', '#ffb3b3', '#99e699', '#29a329'])\n",
    "# stars_labels = np.array(range(N_star_categories)) + 1;\n",
    "# #star_category_dist_fig = plt.figure(figsize=(12,8))\n",
    "# bar_plot_indices = np.arange(N_star_categories) \n",
    "# star_category_absolute_frequencies = df_upsampled.stars.value_counts(ascending=True);\n",
    "# star_category_relative_frequencies = np.array(star_category_absolute_frequencies)/float(sum(star_category_absolute_frequencies))\n",
    "# sns.set()\n",
    "# sns.set(style=\"white\")\n",
    "# fig2 = plt.figure(figsize=(12, 8))\n",
    "# ax = sns.barplot(x=bar_plot_indices + 1, y=star_category_relative_frequencies)\n",
    "# ax.set_xlabel('Star Category');\n",
    "# ax.set_ylabel('Relative Frequency');\n",
    "# ax.set_title('Star Category Distribution for {0} Reviews'.format(len(df_upsampled)));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 537,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({5.0: 3021, 4.0: 1225, 1.0: 731, 3.0: 407, 2.0: 201, 4.5: 15, 3.5: 1})\n"
     ]
    }
   ],
   "source": [
    "stars = df['stars']\n",
    "print(Counter(stars))\n",
    "balanced_y = stars\n",
    "balanced_x = df['text']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 538,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_upsampled\n",
    "# df_upsampled.reset_index()\n",
    "\n",
    "balanced_y=df['stars'].round(0).astype(int)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data bais"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 539,
   "metadata": {},
   "outputs": [],
   "source": [
    "# bow_transformer = CountVectorizer(analyzer=text_process).fit(X_train)\n",
    "# df_upsampled = df_upsampled.loc[:,~df_upsampled.columns.duplicated()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 540,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_upsampled\n",
    "# df_upsampled.reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 541,
   "metadata": {},
   "outputs": [],
   "source": [
    "df2=df[['sentiment','text','stars','oneword']]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 542,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sentiment</th>\n",
       "      <th>text</th>\n",
       "      <th>stars</th>\n",
       "      <th>oneword</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>small north indian</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>cheap low low also</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>compact north indian</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>5-8min bus</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>hidden much</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.0</td>\n",
       "      <td>much poor menu next</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great saturday huge</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.0</td>\n",
       "      <td>sumana gr8 seat</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great hyp</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>5.0</td>\n",
       "      <td>dinner good u</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great neha cant</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1.0</td>\n",
       "      <td></td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>5.0</td>\n",
       "      <td>im sure famous</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>5.0</td>\n",
       "      <td>perfect also</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1.0</td>\n",
       "      <td>due</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>4.0</td>\n",
       "      <td>.we sprite fresh</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.0</td>\n",
       "      <td>salt</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1.0</td>\n",
       "      <td>final long well</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1.0</td>\n",
       "      <td>.the</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good red</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>5.0</td>\n",
       "      <td>im sure famous</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>5.0</td>\n",
       "      <td>perfect also</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1.0</td>\n",
       "      <td>due</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>4.0</td>\n",
       "      <td>.we sprite fresh</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1.0</td>\n",
       "      <td>salt</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1.0</td>\n",
       "      <td>final long well</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1.0</td>\n",
       "      <td>.the</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good red</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5571</th>\n",
       "      <td>5.0</td>\n",
       "      <td>red good</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5572</th>\n",
       "      <td>1.0</td>\n",
       "      <td>pathetic/poor sour even pure</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5573</th>\n",
       "      <td>1.0</td>\n",
       "      <td>ive chicken also good</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5574</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great nice agreat</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5575</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5576</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good higher</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5577</th>\n",
       "      <td>1.0</td>\n",
       "      <td></td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5578</th>\n",
       "      <td>1.0</td>\n",
       "      <td>next</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5579</th>\n",
       "      <td>1.0</td>\n",
       "      <td>cafe</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5580</th>\n",
       "      <td>1.0</td>\n",
       "      <td>similar chicken</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5581</th>\n",
       "      <td>5.0</td>\n",
       "      <td>pleasant plonk</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5582</th>\n",
       "      <td>5.0</td>\n",
       "      <td>much nice</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5583</th>\n",
       "      <td>1.0</td>\n",
       "      <td>less n poor small never</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5584</th>\n",
       "      <td>1.0</td>\n",
       "      <td>quiet late</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5585</th>\n",
       "      <td>1.0</td>\n",
       "      <td>less send</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5586</th>\n",
       "      <td>1.0</td>\n",
       "      <td>decent north south indian small</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5587</th>\n",
       "      <td>1.0</td>\n",
       "      <td>babycorn local</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5588</th>\n",
       "      <td>4.0</td>\n",
       "      <td>great less prior</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5589</th>\n",
       "      <td>1.0</td>\n",
       "      <td>popular sad</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5590</th>\n",
       "      <td>1.0</td>\n",
       "      <td>regular</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5591</th>\n",
       "      <td>1.0</td>\n",
       "      <td>chicken sandwich</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5592</th>\n",
       "      <td>1.0</td>\n",
       "      <td>come garma wohi</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5593</th>\n",
       "      <td>1.0</td>\n",
       "      <td>ever eaten bad poor</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5594</th>\n",
       "      <td>5.0</td>\n",
       "      <td>apart crust still good</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5595</th>\n",
       "      <td>1.0</td>\n",
       "      <td>worst ill never first bad cheap</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5596</th>\n",
       "      <td>1.0</td>\n",
       "      <td>moti</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5597</th>\n",
       "      <td>5.0</td>\n",
       "      <td>decent perfect</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5598</th>\n",
       "      <td>5.0</td>\n",
       "      <td>nice interior</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5599</th>\n",
       "      <td>1.0</td>\n",
       "      <td>last lot space..</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5600</th>\n",
       "      <td>1.0</td>\n",
       "      <td>ever</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Blah!</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5601 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      sentiment                             text  stars  oneword\n",
       "0           1.0               small north indian    5.0  Insane!\n",
       "1           2.0               cheap low low also    1.0   Avoid!\n",
       "2           1.0             compact north indian    5.0  Insane!\n",
       "3           1.0                       5-8min bus    5.0  Insane!\n",
       "4           1.0                      hidden much    5.0  Insane!\n",
       "5           1.0              much poor menu next    1.0   Avoid!\n",
       "6           5.0              great saturday huge    5.0  Insane!\n",
       "7           1.0                  sumana gr8 seat    5.0  Insane!\n",
       "8           5.0                        great hyp    3.0  Average\n",
       "9           5.0                    dinner good u    3.0  Average\n",
       "10          5.0                  great neha cant    5.0  Insane!\n",
       "11          1.0                                     5.0  Insane!\n",
       "12          5.0                   im sure famous    5.0  Insane!\n",
       "13          5.0                     perfect also    5.0  Insane!\n",
       "14          1.0                              due    3.0  Average\n",
       "15          4.0                 .we sprite fresh    5.0  Insane!\n",
       "16          1.0                             salt    5.0  Insane!\n",
       "17          1.0                  final long well    5.0  Insane!\n",
       "18          1.0                             .the    3.0  Average\n",
       "19          5.0                         good red    5.0  Insane!\n",
       "20          5.0                            great    5.0  Insane!\n",
       "21          5.0                   im sure famous    5.0  Insane!\n",
       "22          5.0                     perfect also    5.0  Insane!\n",
       "23          1.0                              due    3.0  Average\n",
       "24          4.0                 .we sprite fresh    5.0  Insane!\n",
       "25          1.0                             salt    5.0  Insane!\n",
       "26          1.0                  final long well    5.0  Insane!\n",
       "27          1.0                             .the    3.0  Average\n",
       "28          5.0                         good red    5.0  Insane!\n",
       "29          5.0                            great    5.0  Insane!\n",
       "...         ...                              ...    ...      ...\n",
       "5571        5.0                         red good    4.0   Great!\n",
       "5572        1.0     pathetic/poor sour even pure    1.0   Avoid!\n",
       "5573        1.0            ive chicken also good    4.0   Great!\n",
       "5574        5.0                great nice agreat    5.0  Insane!\n",
       "5575        5.0                             good    5.0  Insane!\n",
       "5576        5.0                      good higher    5.0  Insane!\n",
       "5577        1.0                                     5.0  Insane!\n",
       "5578        1.0                             next    5.0  Insane!\n",
       "5579        1.0                             cafe    5.0  Insane!\n",
       "5580        1.0                  similar chicken    3.0  Average\n",
       "5581        5.0                   pleasant plonk    4.0   Great!\n",
       "5582        5.0                        much nice    4.0   Great!\n",
       "5583        1.0          less n poor small never    1.0   Avoid!\n",
       "5584        1.0                       quiet late    4.0   Great!\n",
       "5585        1.0                        less send    1.0   Avoid!\n",
       "5586        1.0  decent north south indian small    4.0   Great!\n",
       "5587        1.0                   babycorn local    4.0   Great!\n",
       "5588        4.0                 great less prior    5.0  Insane!\n",
       "5589        1.0                      popular sad    1.0   Avoid!\n",
       "5590        1.0                          regular    3.0  Average\n",
       "5591        1.0                 chicken sandwich    5.0  Insane!\n",
       "5592        1.0                  come garma wohi    5.0  Insane!\n",
       "5593        1.0              ever eaten bad poor    1.0   Avoid!\n",
       "5594        5.0           apart crust still good    4.0   Great!\n",
       "5595        1.0  worst ill never first bad cheap    1.0   Avoid!\n",
       "5596        1.0                             moti    1.0   Avoid!\n",
       "5597        5.0                   decent perfect    5.0  Insane!\n",
       "5598        5.0                    nice interior    4.0   Great!\n",
       "5599        1.0                 last lot space..    4.0   Great!\n",
       "5600        1.0                             ever    2.0    Blah!\n",
       "\n",
       "[5601 rows x 4 columns]"
      ]
     },
     "execution_count": 542,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 543,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df2.groupby('stars', as_index=False)['sentiment'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 544,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3481"
      ]
     },
     "execution_count": 544,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "positive"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 545,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2120"
      ]
     },
     "execution_count": 545,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "negative"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### TF-IDF ###\n",
    "### This vectorizer breaks text into single words and bi-grams and then calculates the TF-IDF representation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 546,
   "metadata": {},
   "outputs": [],
   "source": [
    "# vectorizer = TfidfVectorizer(ngram_range=(1,2))\n",
    "# vectors = vectorizer.fit_transform(balanced_x)\n",
    "# balanced_x.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 547,
   "metadata": {},
   "outputs": [],
   "source": [
    "# print(vectors)\n",
    "# print(vectors.shape[0])\n",
    "# print(vectors.shape[1])\n",
    "df2=df2.drop(['stars'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 548,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(df2,balanced_y, test_size=0.20, random_state=42)\n",
    "X_train1=copy.deepcopy(X_train)\n",
    "y_train1=copy.deepcopy(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 549,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Text Data\n",
      "898                nice also good\n",
      "428                          isnt\n",
      "3678                  ghevar good\n",
      "1505                        great\n",
      "1839    good great .. shugam well\n",
      "Name: text, dtype: object\n",
      "\n",
      "Numeric Data\n",
      "      sentiment\n",
      "898         5.0\n",
      "428         1.0\n",
      "3678        5.0\n",
      "1505        5.0\n",
      "1839        5.0\n"
     ]
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.pipeline import FeatureUnion\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.base import BaseEstimator, TransformerMixin\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.preprocessing import FunctionTransformer\n",
    "\n",
    "from sklearn.preprocessing import FunctionTransformer\n",
    "get_text_data = FunctionTransformer(lambda x: x['text'], validate=False)\n",
    "get_text1_data=FunctionTransformer(lambda x: x['oneword'], validate=False)\n",
    "get_numeric_data = FunctionTransformer(lambda x: x[['sentiment']], validate=False)\n",
    "\n",
    "just_text_data = get_text_data.fit_transform(X_train)\n",
    "just_numeric_data = get_numeric_data.fit_transform(X_train)\n",
    "just_text1_data=get_text1_data.fit_transform(X_train)\n",
    "# Print head to check results\n",
    "print('Text Data')\n",
    "print(just_text_data.head())\n",
    "print('\\nNumeric Data')\n",
    "print(just_numeric_data.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 550,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.neural_network import MLPClassifier \n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "#Predict Output\n",
    "\n",
    "from sklearn import svm\n",
    "# from sklearn import LogisticRegression\n",
    "process_and_join_features = FeatureUnion(\n",
    "            transformer_list = [\n",
    "                ('numeric_features', Pipeline([\n",
    "                    ('selector', get_numeric_data),\n",
    "#                     ('imputer', Imputer())\n",
    "                ])),\n",
    "                ('text_features', Pipeline([\n",
    "                    ('selector', get_text_data),\n",
    "                    ('vectorizer', CountVectorizer())\n",
    "                ]))\n",
    "        ,\n",
    "                ('text1_features', Pipeline([\n",
    "                    ('selector', get_text1_data),\n",
    "                    ('vectorizer', CountVectorizer())\n",
    "#                     ('imputer', Imputer())\n",
    "                ]))\n",
    "             ]\n",
    "        )\n",
    "\n",
    "pl = Pipeline([\n",
    "        ('union', process_and_join_features),\n",
    "        ('clf', OneVsRestClassifier(LogisticRegression()))\n",
    "    ])\n",
    "\n",
    "pl1 = Pipeline([\n",
    "        ('union', process_and_join_features),\n",
    "        ('clf',MLPClassifier(hidden_layer_sizes=(5,5), max_iter=50, alpha=0.0001,\n",
    "                     solver='sgd', verbose=10,  random_state=21,tol=0.000000001))\n",
    "    ])\n",
    "pl2 = Pipeline([\n",
    "        ('union', process_and_join_features),\n",
    "        ('clf', MultinomialNB())\n",
    "    ])\n",
    "pl3 = Pipeline([\n",
    "        ('union', process_and_join_features),\n",
    "        ('clf', GradientBoostingClassifier(learning_rate=0.1,max_depth=5,max_features=0.5,random_state=999999))\n",
    "    ])\n",
    "pl4 = Pipeline([\n",
    "        ('union', process_and_join_features),\n",
    "        ('clf',LogisticRegression(random_state=0, solver='lbfgs',\n",
    "                         multi_class='multinomial'))\n",
    "    ])\n",
    "pl5 = Pipeline([\n",
    "        ('union', process_and_join_features),\n",
    "        ('clf',svm.SVC(gamma='scale', decision_function_shape='ovo'))\n",
    "    ])\n",
    "pl6 = Pipeline([\n",
    "        ('union', process_and_join_features),\n",
    "        ('clf',KNeighborsClassifier(n_neighbors=5))\n",
    "    ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 551,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sentiment</th>\n",
       "      <th>text</th>\n",
       "      <th>oneword</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>898</th>\n",
       "      <td>5.0</td>\n",
       "      <td>nice also good</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>428</th>\n",
       "      <td>1.0</td>\n",
       "      <td>isnt</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3678</th>\n",
       "      <td>5.0</td>\n",
       "      <td>ghevar good</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1505</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1839</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good great .. shugam well</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2093</th>\n",
       "      <td>1.0</td>\n",
       "      <td></td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4260</th>\n",
       "      <td>5.0</td>\n",
       "      <td>yanki nice much</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1738</th>\n",
       "      <td>4.0</td>\n",
       "      <td>secret terrain great</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4438</th>\n",
       "      <td>1.0</td>\n",
       "      <td>roti</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1219</th>\n",
       "      <td>2.0</td>\n",
       "      <td>read right extra</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2973</th>\n",
       "      <td>4.0</td>\n",
       "      <td>live better high</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3914</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good late perfect fine</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2776</th>\n",
       "      <td>1.0</td>\n",
       "      <td>okiesh 40-45</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3408</th>\n",
       "      <td>5.0</td>\n",
       "      <td>nice courteous</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1302</th>\n",
       "      <td>1.0</td>\n",
       "      <td>indian bad bad well us.they</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2119</th>\n",
       "      <td>5.0</td>\n",
       "      <td>sure</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1374</th>\n",
       "      <td>4.0</td>\n",
       "      <td>ive first chembur n</td>\n",
       "      <td>Blah!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1888</th>\n",
       "      <td>1.0</td>\n",
       "      <td>never great</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5108</th>\n",
       "      <td>1.0</td>\n",
       "      <td>raita bad</td>\n",
       "      <td>Blah!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2607</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good good good</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4709</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good baingan masala</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2627</th>\n",
       "      <td>1.0</td>\n",
       "      <td></td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3621</th>\n",
       "      <td>5.0</td>\n",
       "      <td>cool even mukesh</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1514</th>\n",
       "      <td>1.0</td>\n",
       "      <td>high all-day casual</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2451</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great great also</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1126</th>\n",
       "      <td>1.0</td>\n",
       "      <td>small good small</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1876</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5200</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great fast.keep</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3690</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good neat</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2047</th>\n",
       "      <td>5.0</td>\n",
       "      <td>splendid stress</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2558</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good ..high</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1082</th>\n",
       "      <td>1.0</td>\n",
       "      <td></td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>474</th>\n",
       "      <td>5.0</td>\n",
       "      <td>nice well well sugandhi</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2904</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great nakul rajesh quick</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4843</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great better usual</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4117</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good good good</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3385</th>\n",
       "      <td>1.0</td>\n",
       "      <td>lemon</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4555</th>\n",
       "      <td>5.0</td>\n",
       "      <td>nice nice</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1184</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great fast</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5051</th>\n",
       "      <td>1.0</td>\n",
       "      <td>chicken worth sure</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5311</th>\n",
       "      <td>1.0</td>\n",
       "      <td>total tandoor</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2433</th>\n",
       "      <td>1.0</td>\n",
       "      <td>ive bao saturday</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2391</th>\n",
       "      <td>5.0</td>\n",
       "      <td>place.. soup nice</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>769</th>\n",
       "      <td>1.0</td>\n",
       "      <td>lunch gate</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1685</th>\n",
       "      <td>5.0</td>\n",
       "      <td>dipen good good</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>5.0</td>\n",
       "      <td>friends.roof top</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2919</th>\n",
       "      <td>1.0</td>\n",
       "      <td></td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3171</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great spacious spacious</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3444</th>\n",
       "      <td>5.0</td>\n",
       "      <td>pleasant courteous well good</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5578</th>\n",
       "      <td>1.0</td>\n",
       "      <td>next</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4426</th>\n",
       "      <td>5.0</td>\n",
       "      <td>chaat good nice great also good</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5334</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>466</th>\n",
       "      <td>4.0</td>\n",
       "      <td>standard high second best</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3092</th>\n",
       "      <td>1.0</td>\n",
       "      <td>long last sooo</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3772</th>\n",
       "      <td>5.0</td>\n",
       "      <td>youth also good also</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5191</th>\n",
       "      <td>2.0</td>\n",
       "      <td>also store normal</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5226</th>\n",
       "      <td>1.0</td>\n",
       "      <td>well tbh sick ok sick v</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5390</th>\n",
       "      <td>4.0</td>\n",
       "      <td>first</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>860</th>\n",
       "      <td>2.0</td>\n",
       "      <td>dahisar-malad live found</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4480 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      sentiment                             text  oneword\n",
       "898         5.0                   nice also good  Insane!\n",
       "428         1.0                             isnt  Insane!\n",
       "3678        5.0                      ghevar good   Great!\n",
       "1505        5.0                            great  Insane!\n",
       "1839        5.0        good great .. shugam well  Insane!\n",
       "2093        1.0                                   Insane!\n",
       "4260        5.0                  yanki nice much  Insane!\n",
       "1738        4.0             secret terrain great   Great!\n",
       "4438        1.0                             roti   Avoid!\n",
       "1219        2.0                 read right extra  Insane!\n",
       "2973        4.0                 live better high   Great!\n",
       "3914        5.0           good late perfect fine   Great!\n",
       "2776        1.0                     okiesh 40-45   Avoid!\n",
       "3408        5.0                   nice courteous  Insane!\n",
       "1302        1.0      indian bad bad well us.they   Avoid!\n",
       "2119        5.0                             sure  Insane!\n",
       "1374        4.0              ive first chembur n    Blah!\n",
       "1888        1.0                      never great   Great!\n",
       "5108        1.0                        raita bad    Blah!\n",
       "2607        5.0                   good good good  Insane!\n",
       "4709        5.0              good baingan masala  Insane!\n",
       "2627        1.0                                   Insane!\n",
       "3621        5.0                 cool even mukesh  Insane!\n",
       "1514        1.0              high all-day casual  Insane!\n",
       "2451        5.0                 great great also  Insane!\n",
       "1126        1.0                 small good small   Great!\n",
       "1876        5.0                            great  Insane!\n",
       "5200        5.0                  great fast.keep  Insane!\n",
       "3690        5.0                            great   Great!\n",
       "70          5.0                        good neat   Great!\n",
       "...         ...                              ...      ...\n",
       "2047        5.0                  splendid stress  Insane!\n",
       "2558        5.0                      good ..high  Insane!\n",
       "1082        1.0                                   Insane!\n",
       "474         5.0          nice well well sugandhi  Insane!\n",
       "2904        5.0         great nakul rajesh quick  Insane!\n",
       "4843        5.0               great better usual   Great!\n",
       "4117        5.0                   good good good  Insane!\n",
       "3385        1.0                            lemon  Insane!\n",
       "4555        5.0                        nice nice  Insane!\n",
       "1184        5.0                       great fast  Insane!\n",
       "5051        1.0               chicken worth sure  Insane!\n",
       "5311        1.0                    total tandoor  Insane!\n",
       "2433        1.0                 ive bao saturday  Insane!\n",
       "2391        5.0                place.. soup nice  Insane!\n",
       "769         1.0                       lunch gate   Great!\n",
       "1685        5.0                  dipen good good  Insane!\n",
       "130         5.0                 friends.roof top   Great!\n",
       "2919        1.0                                    Great!\n",
       "3171        5.0          great spacious spacious  Insane!\n",
       "3444        5.0     pleasant courteous well good   Great!\n",
       "5578        1.0                             next  Insane!\n",
       "4426        5.0  chaat good nice great also good  Insane!\n",
       "5334        5.0                             good  Insane!\n",
       "466         4.0        standard high second best  Insane!\n",
       "3092        1.0                   long last sooo   Great!\n",
       "3772        5.0             youth also good also  Insane!\n",
       "5191        2.0                also store normal  Average\n",
       "5226        1.0          well tbh sick ok sick v  Average\n",
       "5390        4.0                            first   Avoid!\n",
       "860         2.0         dahisar-malad live found   Great!\n",
       "\n",
       "[4480 rows x 3 columns]"
      ]
     },
     "execution_count": 551,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 552,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "898     5\n",
       "428     5\n",
       "3678    4\n",
       "1505    5\n",
       "1839    5\n",
       "2093    5\n",
       "4260    5\n",
       "1738    4\n",
       "4438    1\n",
       "1219    5\n",
       "2973    4\n",
       "3914    4\n",
       "2776    1\n",
       "3408    5\n",
       "1302    1\n",
       "2119    5\n",
       "1374    2\n",
       "1888    4\n",
       "5108    2\n",
       "2607    5\n",
       "4709    5\n",
       "2627    5\n",
       "3621    5\n",
       "1514    5\n",
       "2451    5\n",
       "1126    4\n",
       "1876    5\n",
       "5200    5\n",
       "3690    4\n",
       "70      4\n",
       "       ..\n",
       "2047    5\n",
       "2558    5\n",
       "1082    5\n",
       "474     5\n",
       "2904    5\n",
       "4843    4\n",
       "4117    5\n",
       "3385    5\n",
       "4555    5\n",
       "1184    5\n",
       "5051    5\n",
       "5311    5\n",
       "2433    5\n",
       "2391    5\n",
       "769     4\n",
       "1685    5\n",
       "130     4\n",
       "2919    4\n",
       "3171    5\n",
       "3444    4\n",
       "5578    5\n",
       "4426    5\n",
       "5334    5\n",
       "466     5\n",
       "3092    4\n",
       "3772    5\n",
       "5191    3\n",
       "5226    3\n",
       "5390    1\n",
       "860     4\n",
       "Name: stars, Length: 4480, dtype: int64"
      ]
     },
     "execution_count": 552,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 553,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.series.Series'>\n"
     ]
    }
   ],
   "source": [
    "print(type(y_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 554,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/dhawal/.local/lib/python3.6/site-packages/sklearn/linear_model/logistic.py:433: FutureWarning: Default solver will be changed to 'lbfgs' in 0.22. Specify a solver to silence this warning.\n",
      "  FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('union', FeatureUnion(n_jobs=None,\n",
       "       transformer_list=[('numeric_features', Pipeline(memory=None,\n",
       "     steps=[('selector', FunctionTransformer(accept_sparse=False, check_inverse=True,\n",
       "          func=<function <lambda> at 0x7f14cfe11378>, inv_kw_args=None,\n",
       "          inverse_func=None, kw...te=None, solver='warn',\n",
       "          tol=0.0001, verbose=0, warm_start=False),\n",
       "          n_jobs=None))])"
      ]
     },
     "execution_count": 554,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl.fit(X_train, y_train)\n",
    "\n",
    "# Compute and print accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 555,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>sentiment</th>\n",
       "      <th>text</th>\n",
       "      <th>oneword</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4082</th>\n",
       "      <td>4.0</td>\n",
       "      <td>nice present</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2104</th>\n",
       "      <td>1.0</td>\n",
       "      <td>pls</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2815</th>\n",
       "      <td>1.0</td>\n",
       "      <td>hub ive various</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5268</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good much best</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3628</th>\n",
       "      <td>5.0</td>\n",
       "      <td>superb</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4113</th>\n",
       "      <td>1.0</td>\n",
       "      <td>bengan</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1702</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good good great</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3630</th>\n",
       "      <td>5.0</td>\n",
       "      <td>best great good</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2480</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good good good special golden</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5438</th>\n",
       "      <td>1.0</td>\n",
       "      <td>give dena</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1115</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good away</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2702</th>\n",
       "      <td>1.0</td>\n",
       "      <td>bad normal nice rooftop much</td>\n",
       "      <td>Blah!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4713</th>\n",
       "      <td>5.0</td>\n",
       "      <td>huge pizza</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1813</th>\n",
       "      <td>4.0</td>\n",
       "      <td>good open super</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2942</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good ambience-</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3891</th>\n",
       "      <td>1.0</td>\n",
       "      <td></td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1835</th>\n",
       "      <td>1.0</td>\n",
       "      <td></td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>1.0</td>\n",
       "      <td>mr. shubhankar</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3023</th>\n",
       "      <td>5.0</td>\n",
       "      <td>umantra great nice good nice good</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2288</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great decent non</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2653</th>\n",
       "      <td>1.0</td>\n",
       "      <td>well yet cant lesser</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>296</th>\n",
       "      <td>1.0</td>\n",
       "      <td>also also ac</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3617</th>\n",
       "      <td>1.0</td>\n",
       "      <td>outlets..</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3545</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good nice good better</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4981</th>\n",
       "      <td>1.0</td>\n",
       "      <td>umdaa aur</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4863</th>\n",
       "      <td>5.0</td>\n",
       "      <td>special</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good non-vegetarian special</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1044</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great huge</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1897</th>\n",
       "      <td>5.0</td>\n",
       "      <td>@ nice nice</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2373</th>\n",
       "      <td>5.0</td>\n",
       "      <td>susanta well nice</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4648</th>\n",
       "      <td>5.0</td>\n",
       "      <td>want perfect also</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>322</th>\n",
       "      <td>5.0</td>\n",
       "      <td>well visit special</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4046</th>\n",
       "      <td>4.0</td>\n",
       "      <td>ever instead od straight</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2080</th>\n",
       "      <td>5.0</td>\n",
       "      <td>ambience.. great karthik.. wish</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2770</th>\n",
       "      <td>5.0</td>\n",
       "      <td>nice good courteous</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334</th>\n",
       "      <td>1.0</td>\n",
       "      <td>roll</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2847</th>\n",
       "      <td>1.0</td>\n",
       "      <td>last lo</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1844</th>\n",
       "      <td>5.0</td>\n",
       "      <td>nd good</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>833</th>\n",
       "      <td>2.0</td>\n",
       "      <td>loud</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2231</th>\n",
       "      <td>1.0</td>\n",
       "      <td>chicken</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2921</th>\n",
       "      <td>5.0</td>\n",
       "      <td>nice good south indian</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4823</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good .i ur sandwich ever</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4035</th>\n",
       "      <td>5.0</td>\n",
       "      <td>rather full</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1732</th>\n",
       "      <td>1.0</td>\n",
       "      <td>waiter give</td>\n",
       "      <td>Avoid!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>731</th>\n",
       "      <td>1.0</td>\n",
       "      <td>ive express id ever</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291</th>\n",
       "      <td>5.0</td>\n",
       "      <td>nice</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>5.0</td>\n",
       "      <td>courteous higher</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4121</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1504</th>\n",
       "      <td>5.0</td>\n",
       "      <td>great extra great</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2586</th>\n",
       "      <td>1.0</td>\n",
       "      <td>flechazo rajesh gonwxtra</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>533</th>\n",
       "      <td>4.0</td>\n",
       "      <td>good french regular</td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4776</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good much much rest good</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1158</th>\n",
       "      <td>1.0</td>\n",
       "      <td>non excess</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5151</th>\n",
       "      <td>5.0</td>\n",
       "      <td>good nice sushi good</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5098</th>\n",
       "      <td>1.0</td>\n",
       "      <td></td>\n",
       "      <td>Average</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>657</th>\n",
       "      <td>1.0</td>\n",
       "      <td>pleased.wish</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3781</th>\n",
       "      <td>5.0</td>\n",
       "      <td>manchurian good nice give</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5335</th>\n",
       "      <td>1.0</td>\n",
       "      <td>u</td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2900</th>\n",
       "      <td>5.0</td>\n",
       "      <td>nice nice nice good</td>\n",
       "      <td>Insane!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5535</th>\n",
       "      <td>1.0</td>\n",
       "      <td></td>\n",
       "      <td>Great!</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1121 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      sentiment                               text  oneword\n",
       "4082        4.0                       nice present  Insane!\n",
       "2104        1.0                                pls  Insane!\n",
       "2815        1.0                    hub ive various   Great!\n",
       "5268        5.0                     good much best  Insane!\n",
       "3628        5.0                             superb   Great!\n",
       "4113        1.0                             bengan   Avoid!\n",
       "1702        5.0                    good good great   Great!\n",
       "3630        5.0                    best great good  Insane!\n",
       "2480        5.0      good good good special golden  Insane!\n",
       "5438        1.0                          give dena   Avoid!\n",
       "1115        5.0                          good away  Insane!\n",
       "2702        1.0       bad normal nice rooftop much    Blah!\n",
       "4713        5.0                         huge pizza   Avoid!\n",
       "1813        4.0                    good open super  Insane!\n",
       "2942        5.0                     good ambience-   Great!\n",
       "3891        1.0                                     Average\n",
       "1835        1.0                                     Average\n",
       "101         1.0                     mr. shubhankar  Insane!\n",
       "3023        5.0  umantra great nice good nice good  Insane!\n",
       "2288        5.0                   great decent non   Great!\n",
       "2653        1.0               well yet cant lesser   Great!\n",
       "296         1.0                       also also ac  Average\n",
       "3617        1.0                          outlets..  Insane!\n",
       "3545        5.0              good nice good better  Average\n",
       "4981        1.0                          umdaa aur  Insane!\n",
       "4863        5.0                            special  Insane!\n",
       "79          5.0        good non-vegetarian special   Great!\n",
       "1044        5.0                         great huge  Insane!\n",
       "1897        5.0                        @ nice nice  Insane!\n",
       "2373        5.0                  susanta well nice  Insane!\n",
       "...         ...                                ...      ...\n",
       "4648        5.0                  want perfect also  Insane!\n",
       "322         5.0                 well visit special   Great!\n",
       "4046        4.0           ever instead od straight   Avoid!\n",
       "2080        5.0    ambience.. great karthik.. wish  Insane!\n",
       "2770        5.0                nice good courteous   Great!\n",
       "334         1.0                               roll  Insane!\n",
       "2847        1.0                            last lo   Great!\n",
       "1844        5.0                            nd good  Insane!\n",
       "833         2.0                               loud  Insane!\n",
       "2231        1.0                            chicken  Insane!\n",
       "2921        5.0             nice good south indian   Great!\n",
       "4823        5.0           good .i ur sandwich ever  Insane!\n",
       "4035        5.0                        rather full  Insane!\n",
       "1732        1.0                        waiter give   Avoid!\n",
       "731         1.0                ive express id ever  Insane!\n",
       "291         5.0                               nice  Insane!\n",
       "184         5.0                   courteous higher   Great!\n",
       "4121        5.0                               good  Average\n",
       "1504        5.0                  great extra great  Insane!\n",
       "2586        1.0           flechazo rajesh gonwxtra  Insane!\n",
       "533         4.0                good french regular  Average\n",
       "4776        5.0           good much much rest good   Great!\n",
       "1158        1.0                         non excess  Insane!\n",
       "5151        5.0               good nice sushi good   Great!\n",
       "5098        1.0                                     Average\n",
       "657         1.0                       pleased.wish  Insane!\n",
       "3781        5.0          manchurian good nice give   Great!\n",
       "5335        1.0                                  u   Great!\n",
       "2900        5.0                nice nice nice good  Insane!\n",
       "5535        1.0                                      Great!\n",
       "\n",
       "[1121 rows x 3 columns]"
      ]
     },
     "execution_count": 555,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 556,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_test1 = pl.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 557,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[5 5 4 ... 4 5 4]\n",
      "1121\n"
     ]
    }
   ],
   "source": [
    "print(y_test1)\n",
    "print(len(y_test1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 558,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1121\n"
     ]
    }
   ],
   "source": [
    "y_test\n",
    "print(len(y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 559,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9991079393398751\n"
     ]
    }
   ],
   "source": [
    "y_test\n",
    "print(accuracy_score(y_test,y_test1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 560,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration 1, loss = 1.76155126\n",
      "Iteration 2, loss = 1.68872188\n",
      "Iteration 3, loss = 1.61803167\n",
      "Iteration 4, loss = 1.55666866\n",
      "Iteration 5, loss = 1.50418605\n",
      "Iteration 6, loss = 1.46021025\n",
      "Iteration 7, loss = 1.42357585\n",
      "Iteration 8, loss = 1.39293749\n",
      "Iteration 9, loss = 1.36801429\n",
      "Iteration 10, loss = 1.34778202\n",
      "Iteration 11, loss = 1.33140010\n",
      "Iteration 12, loss = 1.31785521\n",
      "Iteration 13, loss = 1.30698619\n",
      "Iteration 14, loss = 1.29794437\n",
      "Iteration 15, loss = 1.29043755\n",
      "Iteration 16, loss = 1.28409710\n",
      "Iteration 17, loss = 1.27877511\n",
      "Iteration 18, loss = 1.27412992\n",
      "Iteration 19, loss = 1.27010940\n",
      "Iteration 20, loss = 1.26674830\n",
      "Iteration 21, loss = 1.26384170\n",
      "Iteration 22, loss = 1.26123185\n",
      "Iteration 23, loss = 1.25894560\n",
      "Iteration 24, loss = 1.25690087\n",
      "Iteration 25, loss = 1.25515468\n",
      "Iteration 26, loss = 1.25355349\n",
      "Iteration 27, loss = 1.25221327\n",
      "Iteration 28, loss = 1.25098141\n",
      "Iteration 29, loss = 1.24993310\n",
      "Iteration 30, loss = 1.24883522\n",
      "Iteration 31, loss = 1.24777454\n",
      "Iteration 32, loss = 1.24679840\n",
      "Iteration 33, loss = 1.24594194\n",
      "Iteration 34, loss = 1.24507710\n",
      "Iteration 35, loss = 1.24424476\n",
      "Iteration 36, loss = 1.24337715\n",
      "Iteration 37, loss = 1.24254411\n",
      "Iteration 38, loss = 1.24166628\n",
      "Iteration 39, loss = 1.24074806\n",
      "Iteration 40, loss = 1.23979389\n",
      "Iteration 41, loss = 1.23880233\n",
      "Iteration 42, loss = 1.23772921\n",
      "Iteration 43, loss = 1.23655099\n",
      "Iteration 44, loss = 1.23526595\n",
      "Iteration 45, loss = 1.23387533\n",
      "Iteration 46, loss = 1.23227640\n",
      "Iteration 47, loss = 1.23049111\n",
      "Iteration 48, loss = 1.22851682\n",
      "Iteration 49, loss = 1.22624001\n",
      "Iteration 50, loss = 1.22367797\n",
      "0.5272078501338091\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/dhawal/.local/lib/python3.6/site-packages/sklearn/neural_network/multilayer_perceptron.py:562: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (50) reached and the optimization hasn't converged yet.\n",
      "  % self.max_iter, ConvergenceWarning)\n"
     ]
    }
   ],
   "source": [
    "pl1.fit(X_train, y_train)\n",
    "y_test2 = pl1.predict(X_test)\n",
    "print(accuracy_score(y_test,y_test2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 561,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "     steps=[('union', FeatureUnion(n_jobs=None,\n",
       "       transformer_list=[('numeric_features', Pipeline(memory=None,\n",
       "     steps=[('selector', FunctionTransformer(accept_sparse=False, check_inverse=True,\n",
       "          func=<function <lambda> at 0x7f14cfe11378>, inv_kw_args=None,\n",
       "          inverse_func=None, kw...    transformer_weights=None)), ('clf', MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True))])"
      ]
     },
     "execution_count": 561,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl2.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 562,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9875111507582516\n"
     ]
    }
   ],
   "source": [
    "y_test3 = pl2.predict(X_test)\n",
    "print(accuracy_score(y_test,y_test3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 563,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9991079393398751\n"
     ]
    }
   ],
   "source": [
    "pl3.fit(X_train, y_train)\n",
    "y_test4 = pl3.predict(X_test)\n",
    "print(accuracy_score(y_test,y_test4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 564,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9991079393398751\n"
     ]
    }
   ],
   "source": [
    "pl4.fit(X_train, y_train)\n",
    "y_test5= pl4.predict(X_test)\n",
    "print(accuracy_score(y_test,y_test5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 565,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9991079393398751\n"
     ]
    }
   ],
   "source": [
    "pl5.fit(X_train, y_train)\n",
    "y_test6= pl5.predict(X_test)\n",
    "print(accuracy_score(y_test,y_test6))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 566,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9910793933987511\n"
     ]
    }
   ],
   "source": [
    "pl6.fit(X_train, y_train)\n",
    "y_test7= pl6.predict(X_test)\n",
    "print(accuracy_score(y_test,y_test7))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}