Training - CV and Testing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Task 1 Implement a system that can predict the PM2.5 level\n",
    "Using previous data of Five Chinese Cities Data Set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Get Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I already did Data Preprocessing in another Jupyter Notebook File. \n",
    "\n",
    "Please find 'Data Preparation' folder which consists of notebook files to see what steps I have done"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import math\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATASET_DIR_PATH = 'Prep_FiveCitiePMData_MEAN' #'PREP_DATASET'\n",
    "DATASET_FILE_NAME = 'BeijingPM_cleaned.csv' #'PREP_PM_DATASET.csv'\n",
    "DATASET_FILE_PATH = os.path.join(DATASET_DIR_PATH, DATASET_FILE_NAME)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_data_from_csv(data_path):\n",
    "    return pd.read_csv(data_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = load_data_from_csv(DATASET_FILE_PATH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>Unnamed: 0</th>\n",
       "      <th>No</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "      <th>season</th>\n",
       "      <th>DEWP</th>\n",
       "      <th>HUMI</th>\n",
       "      <th>PRES</th>\n",
       "      <th>TEMP</th>\n",
       "      <th>cbwd</th>\n",
       "      <th>Iws</th>\n",
       "      <th>precipitation</th>\n",
       "      <th>Iprec</th>\n",
       "      <th>PM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>24</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>23</td>\n",
       "      <td>4</td>\n",
       "      <td>-17.0</td>\n",
       "      <td>41.0</td>\n",
       "      <td>1020.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>cv</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>129.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>25</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>-16.0</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1020.0</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>1.79</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>148.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>26</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>-15.0</td>\n",
       "      <td>42.0</td>\n",
       "      <td>1020.0</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>2.68</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>159.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>27</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>-11.0</td>\n",
       "      <td>63.5</td>\n",
       "      <td>1021.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>3.57</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>181.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>28</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>-7.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>1022.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>5.36</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>138.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  No  year  month  day  hour  season  DEWP  HUMI    PRES  TEMP  \\\n",
       "0           0  24  2010      1    1    23       4 -17.0  41.0  1020.0  -5.0   \n",
       "1           1  25  2010      1    2     0       4 -16.0  38.0  1020.0  -4.0   \n",
       "2           2  26  2010      1    2     1       4 -15.0  42.0  1020.0  -4.0   \n",
       "3           3  27  2010      1    2     2       4 -11.0  63.5  1021.0  -5.0   \n",
       "4           4  28  2010      1    2     3       4  -7.0  85.0  1022.0  -5.0   \n",
       "\n",
       "  cbwd   Iws  precipitation  Iprec     PM  \n",
       "0   cv  0.89            0.0    0.0  129.0  \n",
       "1   SE  1.79            0.0    0.0  148.0  \n",
       "2   SE  2.68            0.0    0.0  159.0  \n",
       "3   SE  3.57            0.0    0.0  181.0  \n",
       "4   SE  5.36            0.0    0.0  138.0  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will remove Unnamed column. It is appearing on itself...strange"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.drop(['Unnamed: 0'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>No</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "      <th>season</th>\n",
       "      <th>DEWP</th>\n",
       "      <th>HUMI</th>\n",
       "      <th>PRES</th>\n",
       "      <th>TEMP</th>\n",
       "      <th>cbwd</th>\n",
       "      <th>Iws</th>\n",
       "      <th>precipitation</th>\n",
       "      <th>Iprec</th>\n",
       "      <th>PM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>24</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>23</td>\n",
       "      <td>4</td>\n",
       "      <td>-17.0</td>\n",
       "      <td>41.0</td>\n",
       "      <td>1020.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>cv</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>129.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>25</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>-16.0</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1020.0</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>1.79</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>148.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>26</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>-15.0</td>\n",
       "      <td>42.0</td>\n",
       "      <td>1020.0</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>2.68</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>159.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>27</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>-11.0</td>\n",
       "      <td>63.5</td>\n",
       "      <td>1021.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>3.57</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>181.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>28</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>-7.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>1022.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>5.36</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>138.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   No  year  month  day  hour  season  DEWP  HUMI    PRES  TEMP cbwd   Iws  \\\n",
       "0  24  2010      1    1    23       4 -17.0  41.0  1020.0  -5.0   cv  0.89   \n",
       "1  25  2010      1    2     0       4 -16.0  38.0  1020.0  -4.0   SE  1.79   \n",
       "2  26  2010      1    2     1       4 -15.0  42.0  1020.0  -4.0   SE  2.68   \n",
       "3  27  2010      1    2     2       4 -11.0  63.5  1021.0  -5.0   SE  3.57   \n",
       "4  28  2010      1    2     3       4  -7.0  85.0  1022.0  -5.0   SE  5.36   \n",
       "\n",
       "   precipitation  Iprec     PM  \n",
       "0            0.0    0.0  129.0  \n",
       "1            0.0    0.0  148.0  \n",
       "2            0.0    0.0  159.0  \n",
       "3            0.0    0.0  181.0  \n",
       "4            0.0    0.0  138.0  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There you go! 5 top rows in dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 50351 entries, 0 to 50350\n",
      "Data columns (total 15 columns):\n",
      " #   Column         Non-Null Count  Dtype  \n",
      "---  ------         --------------  -----  \n",
      " 0   No             50351 non-null  int64  \n",
      " 1   year           50351 non-null  int64  \n",
      " 2   month          50351 non-null  int64  \n",
      " 3   day            50351 non-null  int64  \n",
      " 4   hour           50351 non-null  int64  \n",
      " 5   season         50351 non-null  int64  \n",
      " 6   DEWP           50351 non-null  float64\n",
      " 7   HUMI           50351 non-null  float64\n",
      " 8   PRES           50351 non-null  float64\n",
      " 9   TEMP           50351 non-null  float64\n",
      " 10  cbwd           50351 non-null  object \n",
      " 11  Iws            50351 non-null  float64\n",
      " 12  precipitation  50351 non-null  float64\n",
      " 13  Iprec          50351 non-null  float64\n",
      " 14  PM             50351 non-null  float64\n",
      "dtypes: float64(8), int64(6), object(1)\n",
      "memory usage: 5.8+ MB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cleaned and non-null and no outliers. Everything was removed in Data Preparation Part\n",
    "\n",
    "And we 436792 instances which is nice"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1080 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.hist(bins=50, figsize=(20,15))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATA_BEFORE_SPLIT = data.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Before we start doing analysis create TEST SET and put aside"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I will go for 20 % of the data for test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train instances: 40280 \n",
      "Test instances:  10071\n"
     ]
    }
   ],
   "source": [
    "train_set, test_set = train_test_split(data, test_size=0.2, random_state=42)\n",
    "print('Train instances:', len(train_set), '\\nTest instances: ', len(test_set))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "OK, now we are safe that we have a test set that hasn't been seen by us"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Let's Analyze our train data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's see a correlations to our PM values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = train_set.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "corr_matrix = data.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PM               1.000000\n",
       "HUMI             0.425022\n",
       "DEWP             0.146742\n",
       "season           0.118982\n",
       "day              0.075205\n",
       "hour             0.000863\n",
       "month           -0.000947\n",
       "precipitation   -0.026398\n",
       "PRES            -0.028674\n",
       "Iprec           -0.044050\n",
       "No              -0.054587\n",
       "year            -0.056297\n",
       "TEMP            -0.110854\n",
       "Iws             -0.242512\n",
       "Name: PM, dtype: float64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_matrix['PM'].sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see TEMP, Iws, HUMI, season, PRES make impact the most. \n",
    "\n",
    "But correlation measures only linear relationship, but what if we have non-linear?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's have a plots of PM with each of these 'TEMP', 'Iws', 'HUMI', 'season', 'PRES'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x22eee3f14e0>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "data.plot(kind='scatter', x='PM', y='TEMP', alpha=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x22eeed22908>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "data.plot(kind='scatter', x='PM', y='Iws', alpha=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x22eeef72a58>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "data.plot(kind='scatter', x='PM', y='HUMI', alpha=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x22eeaf619e8>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "data.plot(kind='scatter', x='PM', y='season', alpha=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x22eebd50ba8>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "data.plot(kind='scatter', x='PM', y='PRES', alpha=0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Well...not so many information we can get from it. There is a correlation, however not strong one"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prepare data for algorithm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before I feed my data to algorithm I need to prepare it first.\n",
    "\n",
    "1. I need to deal with cbwd feature and transfer it to numerical format\n",
    "2. I need to Standartize my data (Subtract mean and divide by variance)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_labels = data['PM'].copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Extracting labels away from all features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "38322    107.25\n",
       "21131     53.00\n",
       "29695     30.75\n",
       "39366     39.75\n",
       "18367     18.00\n",
       "Name: PM, dtype: float64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_labels.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.drop(\"PM\", axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dropping PM attribute"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>No</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "      <th>season</th>\n",
       "      <th>DEWP</th>\n",
       "      <th>HUMI</th>\n",
       "      <th>PRES</th>\n",
       "      <th>TEMP</th>\n",
       "      <th>cbwd</th>\n",
       "      <th>Iws</th>\n",
       "      <th>precipitation</th>\n",
       "      <th>Iprec</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>38322</th>\n",
       "      <td>40216</td>\n",
       "      <td>2014</td>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>15</td>\n",
       "      <td>2</td>\n",
       "      <td>24.0</td>\n",
       "      <td>56.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>NW</td>\n",
       "      <td>8.04</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21131</th>\n",
       "      <td>22735</td>\n",
       "      <td>2012</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>24.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>1011.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>3.57</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29695</th>\n",
       "      <td>31589</td>\n",
       "      <td>2013</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>17.0</td>\n",
       "      <td>77.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>NW</td>\n",
       "      <td>20.12</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39366</th>\n",
       "      <td>41261</td>\n",
       "      <td>2014</td>\n",
       "      <td>9</td>\n",
       "      <td>16</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>10.0</td>\n",
       "      <td>76.0</td>\n",
       "      <td>1018.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>NW</td>\n",
       "      <td>1.79</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18367</th>\n",
       "      <td>19848</td>\n",
       "      <td>2012</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>23</td>\n",
       "      <td>1</td>\n",
       "      <td>-8.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1022.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>29.06</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          No  year  month  day  hour  season  DEWP   HUMI    PRES  TEMP cbwd  \\\n",
       "38322  40216  2014      8    3    15       2  24.0   56.0  1000.0  34.0   NW   \n",
       "21131  22735  2012      8    5     6       2  24.0  100.0  1011.0  24.0   SE   \n",
       "29695  31589  2013      8    9     4       2  17.0   77.0  1000.0  21.0   NW   \n",
       "39366  41261  2014      9   16     4       3  10.0   76.0  1018.0  14.0   NW   \n",
       "18367  19848  2012      4    6    23       1  -8.0   35.0  1022.0   6.0   SE   \n",
       "\n",
       "         Iws  precipitation  Iprec  \n",
       "38322   8.04            0.0    0.0  \n",
       "21131   3.57            0.0    0.0  \n",
       "29695  20.12            0.0    0.0  \n",
       "39366   1.79            0.0    0.0  \n",
       "18367  29.06            0.0    0.0  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_labels(data):\n",
    "    data_labels = data['PM'].copy()\n",
    "    data = data.drop(\"PM\", axis=1)\n",
    "    return data, data_labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Handling text attributes cbwd and City**. \n",
    "\n",
    "We will convert labels to numbers. \n",
    "We will use OneHot encoding to convert categorical data (City) to numberical and convert cbwd to Simple Numbers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import OrdinalEncoder\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.compose import ColumnTransformer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_pipeline = Pipeline([\n",
    "        ('std_scaler', StandardScaler()),\n",
    "    ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_attribs = list(data.columns)\n",
    "num_attribs.remove('cbwd')\n",
    "#num_attribs.remove('City')\n",
    "\n",
    "\n",
    "transform_pipeline = ColumnTransformer([\n",
    "        (\"cbwd\", OrdinalEncoder(), ['cbwd']),\n",
    "        #(\"city\", OneHotEncoder(), ['City']),\n",
    "        (\"num\", num_pipeline, num_attribs),\n",
    "    ])\n",
    "\n",
    "data_prepared = transform_pipeline.fit_transform(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.        ,  0.89828884,  0.85597376, ..., -0.30976393,\n",
       "        -0.07707254, -0.08390197],\n",
       "       [ 2.        , -0.26218505, -0.3225207 , ..., -0.40090613,\n",
       "        -0.07707254, -0.08390197],\n",
       "       [ 1.        ,  0.32558658,  0.26672653, ..., -0.06345571,\n",
       "        -0.07707254, -0.08390197],\n",
       "       ...,\n",
       "       [ 2.        ,  0.88740172,  0.85597376, ...,  0.31028847,\n",
       "        -0.07707254, -0.08390197],\n",
       "       [ 3.        , -1.70831302, -1.50101516, ..., -0.43740379,\n",
       "        -0.07707254, -0.08390197],\n",
       "       [ 0.        , -0.63022276, -0.91176793, ..., -0.19109557,\n",
       "        -0.07707254, -0.08390197]])"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_prepared"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(40280, 14)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_prepared.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(40280,)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_labels.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we have data centered at 0 which is nice for our ML algorithm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prepare Test Set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Prepare Data for Testing. We have to prepare data in the same way as train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_test = test_set.drop('PM', axis=1)\n",
    "data_test_labels = test_set['PM'].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>No</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "      <th>season</th>\n",
       "      <th>DEWP</th>\n",
       "      <th>HUMI</th>\n",
       "      <th>PRES</th>\n",
       "      <th>TEMP</th>\n",
       "      <th>cbwd</th>\n",
       "      <th>Iws</th>\n",
       "      <th>precipitation</th>\n",
       "      <th>Iprec</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>28079</th>\n",
       "      <td>29971</td>\n",
       "      <td>2013</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>18</td>\n",
       "      <td>2</td>\n",
       "      <td>17.0</td>\n",
       "      <td>61.0</td>\n",
       "      <td>1001.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>38.45</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1570</th>\n",
       "      <td>1662</td>\n",
       "      <td>2010</td>\n",
       "      <td>3</td>\n",
       "      <td>11</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>-8.0</td>\n",
       "      <td>73.0</td>\n",
       "      <td>1013.0</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>NW</td>\n",
       "      <td>1.79</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3366</th>\n",
       "      <td>3497</td>\n",
       "      <td>2010</td>\n",
       "      <td>5</td>\n",
       "      <td>26</td>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>10.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1012.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>40.24</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45082</th>\n",
       "      <td>46978</td>\n",
       "      <td>2015</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>4.0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>997.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>NW</td>\n",
       "      <td>4.92</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15000</th>\n",
       "      <td>16393</td>\n",
       "      <td>2011</td>\n",
       "      <td>11</td>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>64.0</td>\n",
       "      <td>1027.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>cv</td>\n",
       "      <td>0.45</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          No  year  month  day  hour  season  DEWP  HUMI    PRES  TEMP cbwd  \\\n",
       "28079  29971  2013      6    2    18       2  17.0  61.0  1001.0  25.0   SE   \n",
       "1570    1662  2010      3   11     5       1  -8.0  73.0  1013.0  -4.0   NW   \n",
       "3366    3497  2010      5   26    16       1  10.0  30.0  1012.0  29.0   SE   \n",
       "45082  46978  2015      5   12     9       1   4.0  37.0   997.0  19.0   NW   \n",
       "15000  16393  2011     11   15     0       3  -4.0  64.0  1027.0   2.0   cv   \n",
       "\n",
       "         Iws  precipitation  Iprec  \n",
       "28079  38.45            0.0    0.0  \n",
       "1570    1.79            0.0    0.0  \n",
       "3366   40.24            0.0    0.0  \n",
       "45082   4.92            0.0    0.0  \n",
       "15000   0.45            0.0    0.0  "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "28079    146.75\n",
       "1570     135.00\n",
       "3366      72.00\n",
       "45082     10.50\n",
       "15000    333.00\n",
       "Name: PM, dtype: float64"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_test_labels.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_test_prepared = transform_pipeline.transform(data_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Test set is prepared as well"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Summary of all steps above"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import math\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import OrdinalEncoder\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.compose import ColumnTransformer\n",
    "\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import r2_score\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "DATASET_DIR_PATH = 'Prep_FiveCitiePMData_MEAN' #'PREP_DATASET'\n",
    "DATASET_FILE_NAME = 'BeijingPM_cleaned.csv' #'PREP_PM_DATASET.csv'\n",
    "\n",
    "def get_data(DATASET_DIR_PATH=DATASET_DIR_PATH, DATASET_FILE_NAME=DATASET_FILE_NAME):\n",
    "    DATASET_FILE_PATH = os.path.join(DATASET_DIR_PATH, DATASET_FILE_NAME)\n",
    "    data = pd.read_csv(DATASET_FILE_PATH)\n",
    "    data = data.drop(['Unnamed: 0'], axis=1)\n",
    "    return data\n",
    "\n",
    "def extract_data_and_labels(data):\n",
    "    data_labels = data['PM'].copy()\n",
    "    data = data.drop(\"PM\", axis=1)\n",
    "    return data, data_labels\n",
    "\n",
    "def prep_data_for_algo(data):\n",
    "    num_attribs = list(data.columns)\n",
    "    num_attribs.remove('cbwd')\n",
    "    #num_attribs.remove('City')\n",
    "    \n",
    "    num_pipeline = Pipeline([\n",
    "            ('std_scaler', StandardScaler()),\n",
    "        ])\n",
    "    \n",
    "    transform_pipeline = ColumnTransformer([\n",
    "            (\"cbwd\", OrdinalEncoder(), ['cbwd']),\n",
    "            #(\"city\", OneHotEncoder(), ['City']),\n",
    "            (\"num\", num_pipeline, num_attribs),\n",
    "        ])\n",
    "    data_prepared = transform_pipeline.fit_transform(data)\n",
    "    return data_prepared\n",
    "\n",
    "data = get_data()\n",
    "train_set, test_set = train_test_split(data, test_size=0.2, random_state=42)\n",
    "\n",
    "data_train, data_train_labels = extract_data_and_labels(train_set)\n",
    "data_test, data_test_labels = extract_data_and_labels(test_set)\n",
    "\n",
    "data_train_prepared = prep_data_for_algo(data_train)\n",
    "data_test_prepared = prep_data_for_algo(data_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally we are ready to start training. I will train different ML algorithms model and try to predict on the same data it was trained in order to check whether ML algorithm learned something  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train and Cross Validate"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So, I have decided to combine training and Cross Validation steps in one section in order to find best model for TEST evaluation\n",
    "\n",
    "as a Scoring technique in Cross Validation I will use MSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import r2_score\n",
    "\n",
    "from sklearn.model_selection import cross_val_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "def display_results(scores):\n",
    "    print('-----------------------------------------\\n')\n",
    "    print('Cross Validation MSE\\n')\n",
    "    print('Scores: ', scores)\n",
    "    print('Average MSE: ', scores.mean())\n",
    "    print('Average RMSE: ', np.sqrt(scores.mean()))\n",
    "    print('\\n-----------------------------------------\\n')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Linear Regression "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "lin_reg = LinearRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-----------------------------------------\n",
      "\n",
      "Cross Validation MSE\n",
      "\n",
      "Scores:  [5721.33182967 6139.72690061 5821.1353172  5573.9136687  5918.58042736]\n",
      "Average MSE:  5834.937628707687\n",
      "Average RMSE:  76.3867634391436\n",
      "\n",
      "-----------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "lin_reg_scores = cross_val_score(lin_reg, data_train_prepared, data_train_labels, \n",
    "                             scoring=\"neg_mean_squared_error\", cv=5)\n",
    "display_results(-lin_reg_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=False)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lin_reg.fit(data_train_prepared, data_train_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decision Tree regressor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Imperically, I tried to find nice parameters and tune the model according to the Cross Validation scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "tree_reg = DecisionTreeRegressor(max_features=14, max_depth=25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-----------------------------------------\n",
      "\n",
      "Cross Validation MSE\n",
      "\n",
      "Scores:  [2619.45984959 2704.15505412 3105.69633789 2676.52412337 2605.62344418]\n",
      "Average MSE:  2742.2917618303372\n",
      "Average RMSE:  52.366895667304334\n",
      "\n",
      "-----------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "tree_reg_scores = cross_val_score(tree_reg, data_train_prepared, data_train_labels, \n",
    "                             scoring=\"neg_mean_squared_error\", cv=5)\n",
    "display_results(-tree_reg_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeRegressor(criterion='mse', max_depth=25, max_features=14,\n",
       "                      max_leaf_nodes=None, min_impurity_decrease=0.0,\n",
       "                      min_impurity_split=None, min_samples_leaf=1,\n",
       "                      min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
       "                      presort=False, random_state=None, splitter='best')"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree_reg.fit(data_train_prepared, data_train_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest Regressor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Imperically, I tried to find nice parameters and tune the model according to the Cross Validation scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "random_forest_reg = RandomForestRegressor(n_estimators=100, max_features=14, max_depth=25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-----------------------------------------\n",
      "\n",
      "Cross Validation MSE\n",
      "\n",
      "Scores:  [1291.32650721 1343.59955415 1312.45870165 1216.00464394 1226.44012658]\n",
      "Average MSE:  1277.9659067080456\n",
      "Average RMSE:  35.748649019341215\n",
      "\n",
      "-----------------------------------------\n",
      "\n"
     ]
    }
   ],
   "source": [
    "random_forest_reg_scores = cross_val_score(random_forest_reg, data_train_prepared, data_train_labels, \n",
    "                             scoring=\"neg_mean_squared_error\", cv=5)\n",
    "display_results(-random_forest_reg_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=25,\n",
       "                      max_features=14, max_leaf_nodes=None,\n",
       "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                      min_samples_leaf=1, min_samples_split=2,\n",
       "                      min_weight_fraction_leaf=0.0, n_estimators=100,\n",
       "                      n_jobs=None, oob_score=False, random_state=None,\n",
       "                      verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_forest_reg.fit(data_train_prepared, data_train_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TEST SET EVALUATION "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are going to use trained models for testing one by one"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below, evaluation metrics will be used for evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import r2_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict_and_evaluate(model, data_test, data_test_labels, display=True):\n",
    "    model_predictions = model.predict(data_test)\n",
    "    model_mse = mean_squared_error(data_test_labels, model_predictions)\n",
    "    model_rmse = np.sqrt(model_mse)\n",
    "    model_r2 = r2_score(data_test_labels, model_predictions)\n",
    "    if display:\n",
    "        print('-> TEST SET EVALUATION', '\\nMSE: ', model_mse, '\\nRMSE: ', \n",
    "          model_rmse, '\\nR2 score: ', model_r2)\n",
    "    results = [model_mse, model_rmse, model_r2]\n",
    "    return model_predictions, results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_actual_vs_predicted(actual, predicted):\n",
    "    fig,ax = plt.subplots()\n",
    "    ax.scatter(actual, predicted)\n",
    "    ax.plot([actual.min(), actual.max()], [actual.min(), actual.max()], 'k--', lw=4)\n",
    "    ax.set_xlabel('Actual')\n",
    "    ax.set_ylabel('Predicted')\n",
    "    fig.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's start with Linear Regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Linear Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> TEST SET EVALUATION \n",
      "MSE:  5720.582531302023 \n",
      "RMSE:  75.63453266400225 \n",
      "R2 score:  0.287122750925105\n"
     ]
    }
   ],
   "source": [
    "lin_reg_predictions, lin_reg_results = predict_and_evaluate(lin_reg, data_test_prepared, data_test_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Anaconda3\\envs\\pm25\\lib\\site-packages\\ipykernel_launcher.py:7: UserWarning: Matplotlib is currently using module://ipykernel.pylab.backend_inline, which is a non-GUI backend, so cannot show the figure.\n",
      "  import sys\n"
     ]
    },
    {
     "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_actual_vs_predicted(data_test_labels, lin_reg_predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> TEST SET EVALUATION \n",
      "MSE:  4912.486706107801 \n",
      "RMSE:  70.08913400883051 \n",
      "R2 score:  0.3878245808001559\n"
     ]
    }
   ],
   "source": [
    "tree_predictions, tree_results = predict_and_evaluate(tree_reg, data_test_prepared, data_test_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Anaconda3\\envs\\pm25\\lib\\site-packages\\ipykernel_launcher.py:7: UserWarning: Matplotlib is currently using module://ipykernel.pylab.backend_inline, which is a non-GUI backend, so cannot show the figure.\n",
      "  import sys\n"
     ]
    },
    {
     "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_actual_vs_predicted(data_test_labels, tree_predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-> TEST SET EVALUATION \n",
      "MSE:  2364.7701052004236 \n",
      "RMSE:  48.628901953472315 \n",
      "R2 score:  0.7053113388251149\n"
     ]
    }
   ],
   "source": [
    "rnd_clf_predictions, rnd_clf_results = predict_and_evaluate(random_forest_reg, data_test_prepared, data_test_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Anaconda3\\envs\\pm25\\lib\\site-packages\\ipykernel_launcher.py:7: UserWarning: Matplotlib is currently using module://ipykernel.pylab.backend_inline, which is a non-GUI backend, so cannot show the figure.\n",
      "  import sys\n"
     ]
    },
    {
     "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_actual_vs_predicted(data_test_labels, rnd_clf_predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_results = [lin_reg_results, tree_results, rnd_clf_results]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# RESULTS SUMMARY"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_final_result(title, results, metrics_index):\n",
    "    plt.style.use('ggplot')\n",
    "    x = ['Linear Regression', 'Decision Tree', 'Random Forest']\n",
    "    vals = []\n",
    "    for model in results:\n",
    "        vals.append(model[metrics_index])\n",
    "\n",
    "    x_pos = [i for i, _ in enumerate(x)]\n",
    "    plt.bar(x_pos, vals, color=['red', 'green', 'blue'])\n",
    "    plt.xlabel(\"Models\")\n",
    "    plt.ylabel(\"Value\")\n",
    "    plt.title(title)\n",
    "    plt.xticks(x_pos, x)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Mean Square Error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_final_result(title='Mean Square Error', results=test_results, metrics_index=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The mean squared error tells you how close a regression line is to a set of points. It does this by taking the distances from the points to the regression line (these distances are the “errors”) and squaring them"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Conclusion**: **BEST MODEL - RANDOM FOREST** average error in square is MSE:  474.6362469150816"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Root Mean Square Error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_final_result(title='Root Mean Square Error', results=test_results, metrics_index=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Root Mean Square Error (RMSE) is the standard deviation of the residuals (prediction errors). Residuals are a measure of how far from the regression line data points are; RMSE is a measure of how spread out these residuals are. In other words, it tells you how concentrated the data is around the line of best fit."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Conclusion**: **BEST MODEL - RANDOM FOREST** predicting us PM2.5 valu with error +-20 units. RMSE:  21.786148051344036"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## R2 score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_final_result(title='R2 score', results=test_results, metrics_index=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "R-squared is a statistical measure of how close the data are to the fitted regression line. It is also known as the coefficient of determination, or the coefficient of multiple determination for multiple regression.\n",
    "\n",
    "The definition of R-squared is fairly straight-forward; it is the percentage of the response variable variation that is explained by a linear model. Or:\n",
    "\n",
    "R-squared = Explained variation / Total variation\n",
    "\n",
    "R-squared is always between 0 and 100%:\n",
    "\n",
    "* 0% indicates that the model explains none of the variability of the response data around its mean.\n",
    "* 100% indicates that the model explains all the variability of the response data around its mean."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Conclusion**: **BEST MODEL - RANDOM FOREST** explains 90% the variability of the response data around its mean. R2 score:  0.8969523276582472"
   ]
  }
 ],
 "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}