LSTM,GRU & BiLSTM.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# BiLSTM, LSTM & GRU"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy\n",
    "import scipy\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import Sequential, layers, callbacks\n",
    "from tensorflow.keras.layers import Dense, LSTM, Dropout, GRU, Bidirectional\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "from IPython.core.pylabtools import figsize\n",
    "\n",
    "from pandas.plotting import register_matplotlib_converters\n",
    "register_matplotlib_converters()\n",
    "import seaborn as sns\n",
    "import pickle"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = pd.read_csv(\"extracted_feature_with_Response.csv\")\n",
    "X = dataset.iloc[:,[1,258]].values\n",
    "y = dataset.iloc[:,[-1]].values\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>CTD.C_A</th>\n",
       "      <th>CTD.C_T</th>\n",
       "      <th>CTD.C_C</th>\n",
       "      <th>CTD.C_G</th>\n",
       "      <th>CTD.T_A_T</th>\n",
       "      <th>CTD.T_A_C</th>\n",
       "      <th>CTD.T_A_G</th>\n",
       "      <th>CTD.T_T_C</th>\n",
       "      <th>CTD.T_T_G</th>\n",
       "      <th>CTD.T_C_G</th>\n",
       "      <th>...</th>\n",
       "      <th>CKSAAP.G0G.2</th>\n",
       "      <th>CKSAAP.G1A.1</th>\n",
       "      <th>CKSAAP.G1T.1</th>\n",
       "      <th>CKSAAP.G1C.1</th>\n",
       "      <th>CKSAAP.G1G.1</th>\n",
       "      <th>CKSAAP.G2A</th>\n",
       "      <th>CKSAAP.G2T</th>\n",
       "      <th>CKSAAP.G2C</th>\n",
       "      <th>CKSAAP.G2G</th>\n",
       "      <th>Response</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>166.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.173911</td>\n",
       "      <td>0.169655</td>\n",
       "      <td>0.326005</td>\n",
       "      <td>0.330429</td>\n",
       "      <td>0.052728</td>\n",
       "      <td>0.112409</td>\n",
       "      <td>0.115872</td>\n",
       "      <td>0.121176</td>\n",
       "      <td>0.106072</td>\n",
       "      <td>0.253488</td>\n",
       "      <td>...</td>\n",
       "      <td>29.198795</td>\n",
       "      <td>17.361446</td>\n",
       "      <td>19.746988</td>\n",
       "      <td>35.885542</td>\n",
       "      <td>30.945783</td>\n",
       "      <td>16.951807</td>\n",
       "      <td>14.813253</td>\n",
       "      <td>34.132530</td>\n",
       "      <td>37.728916</td>\n",
       "      <td>0.825301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.049914</td>\n",
       "      <td>0.043270</td>\n",
       "      <td>0.050741</td>\n",
       "      <td>0.044041</td>\n",
       "      <td>0.031173</td>\n",
       "      <td>0.029391</td>\n",
       "      <td>0.028282</td>\n",
       "      <td>0.026755</td>\n",
       "      <td>0.025699</td>\n",
       "      <td>0.070033</td>\n",
       "      <td>...</td>\n",
       "      <td>13.295805</td>\n",
       "      <td>9.056828</td>\n",
       "      <td>9.055845</td>\n",
       "      <td>17.225215</td>\n",
       "      <td>13.823676</td>\n",
       "      <td>9.254520</td>\n",
       "      <td>7.820413</td>\n",
       "      <td>15.169564</td>\n",
       "      <td>16.390833</td>\n",
       "      <td>0.380858</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.093496</td>\n",
       "      <td>0.094527</td>\n",
       "      <td>0.117409</td>\n",
       "      <td>0.187215</td>\n",
       "      <td>0.008721</td>\n",
       "      <td>0.045249</td>\n",
       "      <td>0.051163</td>\n",
       "      <td>0.052632</td>\n",
       "      <td>0.035000</td>\n",
       "      <td>0.046154</td>\n",
       "      <td>...</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.140474</td>\n",
       "      <td>0.142272</td>\n",
       "      <td>0.308384</td>\n",
       "      <td>0.308831</td>\n",
       "      <td>0.033259</td>\n",
       "      <td>0.090641</td>\n",
       "      <td>0.096441</td>\n",
       "      <td>0.104692</td>\n",
       "      <td>0.087868</td>\n",
       "      <td>0.222408</td>\n",
       "      <td>...</td>\n",
       "      <td>21.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>25.250000</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.168239</td>\n",
       "      <td>0.159995</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.330677</td>\n",
       "      <td>0.046693</td>\n",
       "      <td>0.111993</td>\n",
       "      <td>0.113861</td>\n",
       "      <td>0.117979</td>\n",
       "      <td>0.103933</td>\n",
       "      <td>0.260878</td>\n",
       "      <td>...</td>\n",
       "      <td>27.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>34.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.193515</td>\n",
       "      <td>0.186854</td>\n",
       "      <td>0.353590</td>\n",
       "      <td>0.364342</td>\n",
       "      <td>0.059693</td>\n",
       "      <td>0.129469</td>\n",
       "      <td>0.132042</td>\n",
       "      <td>0.134817</td>\n",
       "      <td>0.119801</td>\n",
       "      <td>0.293355</td>\n",
       "      <td>...</td>\n",
       "      <td>34.000000</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>43.000000</td>\n",
       "      <td>37.750000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>17.750000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>0.380567</td>\n",
       "      <td>0.318008</td>\n",
       "      <td>0.457143</td>\n",
       "      <td>0.427451</td>\n",
       "      <td>0.209459</td>\n",
       "      <td>0.218593</td>\n",
       "      <td>0.236842</td>\n",
       "      <td>0.221918</td>\n",
       "      <td>0.184100</td>\n",
       "      <td>0.396476</td>\n",
       "      <td>...</td>\n",
       "      <td>115.000000</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>177.000000</td>\n",
       "      <td>114.000000</td>\n",
       "      <td>77.000000</td>\n",
       "      <td>53.000000</td>\n",
       "      <td>138.000000</td>\n",
       "      <td>150.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 259 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          CTD.C_A     CTD.C_T     CTD.C_C     CTD.C_G   CTD.T_A_T   CTD.T_A_C  \\\n",
       "count  166.000000  166.000000  166.000000  166.000000  166.000000  166.000000   \n",
       "mean     0.173911    0.169655    0.326005    0.330429    0.052728    0.112409   \n",
       "std      0.049914    0.043270    0.050741    0.044041    0.031173    0.029391   \n",
       "min      0.093496    0.094527    0.117409    0.187215    0.008721    0.045249   \n",
       "25%      0.140474    0.142272    0.308384    0.308831    0.033259    0.090641   \n",
       "50%      0.168239    0.159995    0.333333    0.330677    0.046693    0.111993   \n",
       "75%      0.193515    0.186854    0.353590    0.364342    0.059693    0.129469   \n",
       "max      0.380567    0.318008    0.457143    0.427451    0.209459    0.218593   \n",
       "\n",
       "        CTD.T_A_G   CTD.T_T_C   CTD.T_T_G   CTD.T_C_G  ...  CKSAAP.G0G.2  \\\n",
       "count  166.000000  166.000000  166.000000  166.000000  ...    166.000000   \n",
       "mean     0.115872    0.121176    0.106072    0.253488  ...     29.198795   \n",
       "std      0.028282    0.026755    0.025699    0.070033  ...     13.295805   \n",
       "min      0.051163    0.052632    0.035000    0.046154  ...      7.000000   \n",
       "25%      0.096441    0.104692    0.087868    0.222408  ...     21.000000   \n",
       "50%      0.113861    0.117979    0.103933    0.260878  ...     27.000000   \n",
       "75%      0.132042    0.134817    0.119801    0.293355  ...     34.000000   \n",
       "max      0.236842    0.221918    0.184100    0.396476  ...    115.000000   \n",
       "\n",
       "       CKSAAP.G1A.1  CKSAAP.G1T.1  CKSAAP.G1C.1  CKSAAP.G1G.1  CKSAAP.G2A  \\\n",
       "count    166.000000    166.000000    166.000000    166.000000  166.000000   \n",
       "mean      17.361446     19.746988     35.885542     30.945783   16.951807   \n",
       "std        9.056828      9.055845     17.225215     13.823676    9.254520   \n",
       "min        5.000000      4.000000      6.000000      3.000000    4.000000   \n",
       "25%       11.000000     14.000000     25.250000     22.000000   11.000000   \n",
       "50%       15.000000     18.000000     34.000000     30.000000   15.000000   \n",
       "75%       22.000000     23.000000     43.000000     37.750000   20.000000   \n",
       "max       64.000000     77.000000    177.000000    114.000000   77.000000   \n",
       "\n",
       "       CKSAAP.G2T  CKSAAP.G2C  CKSAAP.G2G    Response  \n",
       "count  166.000000  166.000000  166.000000  166.000000  \n",
       "mean    14.813253   34.132530   37.728916    0.825301  \n",
       "std      7.820413   15.169564   16.390833    0.380858  \n",
       "min      1.000000    3.000000    7.000000    0.000000  \n",
       "25%     10.000000   25.000000   28.000000    1.000000  \n",
       "50%     13.000000   32.000000   35.000000    1.000000  \n",
       "75%     17.750000   41.000000   45.000000    1.000000  \n",
       "max     53.000000  138.000000  150.000000    1.000000  \n",
       "\n",
       "[8 rows x 259 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Split the data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.20, random_state = 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Scaling the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "sc = StandardScaler()\n",
    "X_train_norm = sc.fit_transform(X_train)\n",
    "X_test_norm = sc.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create a 3D input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train.shape:  (4, 30, 2)\n",
      "y_train.shape:  (102, 1)\n",
      "X_test.shape:  (4, 30, 2)\n",
      "y_test.shape:  (102, 1)\n"
     ]
    }
   ],
   "source": [
    "def create_dataset (X, y, time_steps = 1):\n",
    "    Xs, ys = [], []\n",
    "    for i in range(len(X)-time_steps):\n",
    "        v = X[i:i+time_steps, :]\n",
    "        Xs.append(v)\n",
    "        ys.append(y[i+time_steps])\n",
    "    return np.array(Xs), np.array(ys)\n",
    "TIME_STEPS = 30\n",
    "X_test, y_test = create_dataset(X_test_norm, y_test,   \n",
    "                                TIME_STEPS)\n",
    "X_train, y_train = create_dataset(X_train_norm, y_train, \n",
    "                                  TIME_STEPS)\n",
    "print('X_train.shape: ', X_test.shape)\n",
    "print('y_train.shape: ', y_train.shape)\n",
    "print('X_test.shape: ', X_test.shape)\n",
    "print('y_test.shape: ', y_train.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Designing the network architecture"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create BiLSTM, LSTM and GRU model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create BiLSTM model\n",
    "def create_model_bilstm(units):\n",
    "    model = Sequential()\n",
    "    # First layer of BiLSTM\n",
    "    model.add(Bidirectional(LSTM(units = units, return_sequences=True), \n",
    "                            input_shape=(X_train.shape[1], X_train.shape[2])))\n",
    "    # Second layer of BiLSTM\n",
    "    model.add(Bidirectional(LSTM(units = units)))\n",
    "    model.add(Dense(1))\n",
    "    #Compile model\n",
    "    model.compile(loss='mse', optimizer='adam')\n",
    "    return model\n",
    "\n",
    "\n",
    "# Create LSTM or GRU model\n",
    "def create_model(units, m):\n",
    "    model = Sequential()\n",
    "    #First layer of LSTM\n",
    "    model.add(m (units = units, return_sequences = True,\n",
    "                input_shape = [X_train.shape[1], X_train.shape[2]]))\n",
    "    model.add(Dropout(0.2))\n",
    "    #Second Layer of LSTM\n",
    "    model.add(m (units = units))\n",
    "    model.add(Dropout(0.2))\n",
    "    model.add(Dense(units = 1))\n",
    "    #Compile model\n",
    "    model.compile(loss='mse', optimizer='adam')\n",
    "    return model\n",
    "\n",
    "# BiLSTM\n",
    "model_bilstm = create_model_bilstm(64)\n",
    "\n",
    "# GRU\n",
    "model_gru = create_model(64, GRU)\n",
    "\n",
    "# LSTM\n",
    "model_lstm = create_model(64, LSTM)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LSTM model fitting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "3/3 [==============================] - 2s 539ms/step - loss: 0.7776 - val_loss: 1.1114\n",
      "Epoch 2/100\n",
      "3/3 [==============================] - 0s 65ms/step - loss: 0.4798 - val_loss: 1.3476\n",
      "Epoch 3/100\n",
      "3/3 [==============================] - 0s 60ms/step - loss: 0.2265 - val_loss: 1.6495\n",
      "Epoch 4/100\n",
      "3/3 [==============================] - 0s 63ms/step - loss: 0.2367 - val_loss: 1.4210\n",
      "Epoch 5/100\n",
      "3/3 [==============================] - 0s 63ms/step - loss: 0.2009 - val_loss: 1.0477\n",
      "Epoch 6/100\n",
      "3/3 [==============================] - 0s 67ms/step - loss: 0.1572 - val_loss: 0.8005\n",
      "Epoch 7/100\n",
      "3/3 [==============================] - 0s 64ms/step - loss: 0.1615 - val_loss: 0.6509\n",
      "Epoch 8/100\n",
      "3/3 [==============================] - 0s 64ms/step - loss: 0.1552 - val_loss: 0.5508\n",
      "Epoch 9/100\n",
      "3/3 [==============================] - 0s 69ms/step - loss: 0.1426 - val_loss: 0.4777\n",
      "Epoch 10/100\n",
      "3/3 [==============================] - 0s 68ms/step - loss: 0.1414 - val_loss: 0.4245\n",
      "Epoch 11/100\n",
      "3/3 [==============================] - 0s 61ms/step - loss: 0.1432 - val_loss: 0.3884\n",
      "Epoch 12/100\n",
      "3/3 [==============================] - 0s 62ms/step - loss: 0.1375 - val_loss: 0.3665\n",
      "Epoch 13/100\n",
      "3/3 [==============================] - 0s 62ms/step - loss: 0.1351 - val_loss: 0.3535\n",
      "Epoch 14/100\n",
      "3/3 [==============================] - 0s 62ms/step - loss: 0.1355 - val_loss: 0.3431\n",
      "Epoch 15/100\n",
      "3/3 [==============================] - 0s 62ms/step - loss: 0.1332 - val_loss: 0.3321\n",
      "Epoch 16/100\n",
      "3/3 [==============================] - 0s 70ms/step - loss: 0.1307 - val_loss: 0.3212\n",
      "Epoch 17/100\n",
      "3/3 [==============================] - 0s 60ms/step - loss: 0.1306 - val_loss: 0.3124\n",
      "Epoch 18/100\n",
      "3/3 [==============================] - 0s 65ms/step - loss: 0.1301 - val_loss: 0.3059\n",
      "Epoch 19/100\n",
      "3/3 [==============================] - 0s 62ms/step - loss: 0.1288 - val_loss: 0.2992\n",
      "Epoch 20/100\n",
      "3/3 [==============================] - 0s 64ms/step - loss: 0.1280 - val_loss: 0.2897\n",
      "Epoch 21/100\n",
      "3/3 [==============================] - 0s 61ms/step - loss: 0.1274 - val_loss: 0.2777\n",
      "Epoch 22/100\n",
      "3/3 [==============================] - 0s 64ms/step - loss: 0.1265 - val_loss: 0.2662\n",
      "Epoch 23/100\n",
      "3/3 [==============================] - 0s 64ms/step - loss: 0.1261 - val_loss: 0.2575\n",
      "Epoch 24/100\n",
      "3/3 [==============================] - 0s 62ms/step - loss: 0.1255 - val_loss: 0.2512\n",
      "Epoch 25/100\n",
      "3/3 [==============================] - 0s 60ms/step - loss: 0.1248 - val_loss: 0.2460\n",
      "Epoch 26/100\n",
      "3/3 [==============================] - 0s 65ms/step - loss: 0.1243 - val_loss: 0.2415\n",
      "Epoch 27/100\n",
      "3/3 [==============================] - 0s 63ms/step - loss: 0.1238 - val_loss: 0.2382\n",
      "Epoch 28/100\n",
      "3/3 [==============================] - 0s 62ms/step - loss: 0.1233 - val_loss: 0.2360\n",
      "Epoch 29/100\n",
      "3/3 [==============================] - 0s 63ms/step - loss: 0.1228 - val_loss: 0.2341\n",
      "Epoch 30/100\n",
      "3/3 [==============================] - 0s 66ms/step - loss: 0.1224 - val_loss: 0.2325\n",
      "Epoch 31/100\n",
      "3/3 [==============================] - 0s 64ms/step - loss: 0.1220 - val_loss: 0.2322\n",
      "Epoch 32/100\n",
      "3/3 [==============================] - 0s 78ms/step - loss: 0.1217 - val_loss: 0.2331\n",
      "Epoch 33/100\n",
      "3/3 [==============================] - 0s 61ms/step - loss: 0.1214 - val_loss: 0.2341\n",
      "Epoch 34/100\n",
      "3/3 [==============================] - 0s 66ms/step - loss: 0.1210 - val_loss: 0.2352\n",
      "Epoch 35/100\n",
      "3/3 [==============================] - 0s 63ms/step - loss: 0.1207 - val_loss: 0.2371\n",
      "Epoch 36/100\n",
      "3/3 [==============================] - 0s 65ms/step - loss: 0.1204 - val_loss: 0.2394\n",
      "Epoch 37/100\n",
      "3/3 [==============================] - 0s 63ms/step - loss: 0.1201 - val_loss: 0.2407\n",
      "Epoch 38/100\n",
      "3/3 [==============================] - 0s 64ms/step - loss: 0.1197 - val_loss: 0.2407\n",
      "Epoch 39/100\n",
      "3/3 [==============================] - 0s 61ms/step - loss: 0.1194 - val_loss: 0.2406\n",
      "Epoch 40/100\n",
      "3/3 [==============================] - 0s 63ms/step - loss: 0.1191 - val_loss: 0.2405\n",
      "Epoch 41/100\n",
      "3/3 [==============================] - 0s 62ms/step - loss: 0.1188 - val_loss: 0.2396\n",
      "Epoch 1/100\n",
      "3/3 [==============================] - 1s 279ms/step - loss: 0.8167 - val_loss: 1.0407\n",
      "Epoch 2/100\n",
      "3/3 [==============================] - 0s 52ms/step - loss: 0.6520 - val_loss: 1.1240\n",
      "Epoch 3/100\n",
      "3/3 [==============================] - 0s 54ms/step - loss: 0.4580 - val_loss: 1.2245\n",
      "Epoch 4/100\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 0.2165 - val_loss: 1.2994\n",
      "Epoch 5/100\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 0.3023 - val_loss: 1.0463\n",
      "Epoch 6/100\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 0.2165 - val_loss: 0.8158\n",
      "Epoch 7/100\n",
      "3/3 [==============================] - 0s 51ms/step - loss: 0.1710 - val_loss: 0.6448\n",
      "Epoch 8/100\n",
      "3/3 [==============================] - 0s 53ms/step - loss: 0.1765 - val_loss: 0.4991\n",
      "Epoch 9/100\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 0.1609 - val_loss: 0.3634\n",
      "Epoch 10/100\n",
      "3/3 [==============================] - 0s 53ms/step - loss: 0.1463 - val_loss: 0.2562\n",
      "Epoch 11/100\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 0.1626 - val_loss: 0.2025\n",
      "Epoch 12/100\n",
      "3/3 [==============================] - 0s 54ms/step - loss: 0.1532 - val_loss: 0.1835\n",
      "Epoch 13/100\n",
      "3/3 [==============================] - 0s 52ms/step - loss: 0.1313 - val_loss: 0.1744\n",
      "Epoch 14/100\n",
      "3/3 [==============================] - 0s 53ms/step - loss: 0.1479 - val_loss: 0.1683\n",
      "Epoch 15/100\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 0.1378 - val_loss: 0.1650\n",
      "Epoch 16/100\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 0.1457 - val_loss: 0.1649\n",
      "Epoch 17/100\n",
      "3/3 [==============================] - 0s 53ms/step - loss: 0.1451 - val_loss: 0.1649\n",
      "Epoch 18/100\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 0.1521 - val_loss: 0.1641\n",
      "Epoch 19/100\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 0.1361 - val_loss: 0.1640\n",
      "Epoch 20/100\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 0.1355 - val_loss: 0.1636\n",
      "Epoch 21/100\n",
      "3/3 [==============================] - 0s 51ms/step - loss: 0.1338 - val_loss: 0.1629\n",
      "Epoch 22/100\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 0.1435 - val_loss: 0.1623\n",
      "Epoch 23/100\n",
      "3/3 [==============================] - 0s 52ms/step - loss: 0.1400 - val_loss: 0.1616\n",
      "Epoch 24/100\n",
      "3/3 [==============================] - 0s 48ms/step - loss: 0.1444 - val_loss: 0.1611\n",
      "Epoch 25/100\n",
      "3/3 [==============================] - 0s 50ms/step - loss: 0.1288 - val_loss: 0.1623\n",
      "Epoch 26/100\n",
      "3/3 [==============================] - 0s 51ms/step - loss: 0.1430 - val_loss: 0.1649\n",
      "Epoch 27/100\n",
      "3/3 [==============================] - 0s 54ms/step - loss: 0.1447 - val_loss: 0.1630\n",
      "Epoch 28/100\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.1357 - val_loss: 0.1616\n",
      "Epoch 29/100\n",
      "3/3 [==============================] - 0s 53ms/step - loss: 0.1378 - val_loss: 0.1620\n",
      "Epoch 30/100\n",
      "3/3 [==============================] - 0s 51ms/step - loss: 0.1338 - val_loss: 0.1647\n",
      "Epoch 31/100\n",
      "3/3 [==============================] - 0s 52ms/step - loss: 0.1324 - val_loss: 0.1659\n",
      "Epoch 32/100\n",
      "3/3 [==============================] - 0s 51ms/step - loss: 0.1421 - val_loss: 0.1625\n",
      "Epoch 33/100\n",
      "3/3 [==============================] - 0s 55ms/step - loss: 0.1363 - val_loss: 0.1620\n",
      "Epoch 34/100\n",
      "3/3 [==============================] - 0s 49ms/step - loss: 0.1328 - val_loss: 0.1641\n",
      "Epoch 1/100\n",
      "3/3 [==============================] - 1s 314ms/step - loss: 0.8072 - val_loss: 0.9347\n",
      "Epoch 2/100\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 0.6531 - val_loss: 0.8245\n",
      "Epoch 3/100\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 0.4933 - val_loss: 0.6736\n",
      "Epoch 4/100\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.3384 - val_loss: 0.4941\n",
      "Epoch 5/100\n",
      "3/3 [==============================] - 0s 41ms/step - loss: 0.2108 - val_loss: 0.2999\n",
      "Epoch 6/100\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.1616 - val_loss: 0.1635\n",
      "Epoch 7/100\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.2123 - val_loss: 0.1380\n",
      "Epoch 8/100\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 0.1660 - val_loss: 0.1404\n",
      "Epoch 9/100\n",
      "3/3 [==============================] - 0s 42ms/step - loss: 0.1448 - val_loss: 0.1459\n",
      "Epoch 10/100\n",
      "3/3 [==============================] - 0s 46ms/step - loss: 0.1529 - val_loss: 0.1501\n",
      "Epoch 11/100\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.1609 - val_loss: 0.1514\n",
      "Epoch 12/100\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.1528 - val_loss: 0.1514\n",
      "Epoch 13/100\n",
      "3/3 [==============================] - 0s 43ms/step - loss: 0.1578 - val_loss: 0.1511\n",
      "Epoch 14/100\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.1272 - val_loss: 0.1512\n",
      "Epoch 15/100\n",
      "3/3 [==============================] - 0s 44ms/step - loss: 0.1327 - val_loss: 0.1506\n",
      "Epoch 16/100\n",
      "3/3 [==============================] - 0s 45ms/step - loss: 0.1414 - val_loss: 0.1490\n",
      "Epoch 17/100\n",
      "3/3 [==============================] - 0s 47ms/step - loss: 0.1466 - val_loss: 0.1488\n"
     ]
    }
   ],
   "source": [
    "def fit_model(model):\n",
    "    early_stop = keras.callbacks.EarlyStopping(monitor = 'val_loss',\n",
    "                                               patience = 10)\n",
    "    history = model.fit(X_train, y_train, epochs = 100,  \n",
    "                        validation_split = 0.2, batch_size = 32, \n",
    "                        shuffle = False, callbacks = [early_stop])\n",
    "    return history\n",
    "\n",
    "history_bilstm = fit_model(model_bilstm)\n",
    "history_lstm = fit_model(model_lstm)\n",
    "history_gru = fit_model(model_gru)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot train loss and validation loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_loss (history,model_name):\n",
    "    plt.figure(figsize = (10, 6))\n",
    "    plt.plot(history.history['loss'])\n",
    "    plt.plot(history.history['val_loss'])\n",
    "    plt.title('Model Train vs Validation Loss for ' + model_name)\n",
    "    plt.ylabel('Loss')\n",
    "    plt.xlabel('epoch')\n",
    "    plt.legend(['Train loss', 'Validation loss'], loc='upper right')\n",
    "    \n",
    "plot_loss (history_bilstm, 'BiLSTM')\n",
    "plot_loss (history_lstm, 'LSTM')\n",
    "plot_loss (history_gru, 'GRU')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Inverse transform the target variable"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "sc = StandardScaler()\n",
    "y_train = scaler_y.inverse_transform(y_train)\n",
    "y_test = scaler_y.inverse_transform(y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make prediction using LSTM and GRU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def prediction(model):\n",
    "    prediction = model.predict(X_test)\n",
    "    return prediction\n",
    "\n",
    "prediction_bilstm = prediction(model_bilstm)\n",
    "prediction_lstm = prediction(model_lstm)\n",
    "prediction_gru = prediction(model_gru)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot true future vs prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_future(prediction, model_name, y_test):\n",
    "    plt.figure(figsize=(10, 6))\n",
    "    range_future = len(prediction)\n",
    "    plt.plot(np.arange(range_future), np.array(y_test), \n",
    "             label='True Future')     \n",
    "    plt.plot(np.arange(range_future),np.array(prediction),\n",
    "            label='Prediction')\n",
    "    plt.legend(loc='upper left')\n",
    "    plt.title('True Future vs Prediction ' + model_name)\n",
    "    plt.xlabel('Scaffolding')\n",
    "    plt.ylabel('Position')\n",
    "\n",
    "plot_future(prediction_bilstm,'BiLSTM', y_test)\n",
    "plot_future(prediction_lstm,'LSTM', y_test)\n",
    "plot_future(prediction_gru, 'GRU' ,y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculate RMSE and MAE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Bidirectional LSTM:\n",
      "Mean Absolute Error: 0.4104\n",
      "Root Mean Square Error: 0.6003\n",
      "\n",
      "LSTM:\n",
      "Mean Absolute Error: 0.3083\n",
      "Root Mean Square Error: 0.4470\n",
      "\n",
      "GRU:\n",
      "Mean Absolute Error: 0.3818\n",
      "Root Mean Square Error: 0.4287\n",
      "\n"
     ]
    }
   ],
   "source": [
    "def evaluate_prediction(predictions, actual, model_name):\n",
    "    errors =  actual - predictions\n",
    "    mse = np.square(errors).mean()\n",
    "    rmse = np.sqrt(mse)\n",
    "    mae = np.abs(errors).mean()\n",
    "    \n",
    "    print(model_name + ':')\n",
    "    print('Mean Absolute Error: {:.4f}'.format(mae))\n",
    "    print('Root Mean Square Error: {:.4f}'.format(rmse))\n",
    "    print('')\n",
    "\n",
    "evaluate_prediction(prediction_bilstm, y_test, 'Bidirectional LSTM')\n",
    "evaluate_prediction(prediction_lstm, y_test, 'LSTM')\n",
    "evaluate_prediction(prediction_gru, y_test, 'GRU')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Input Curated data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "ename": "FileNotFoundError",
     "evalue": "[Errno 2] No such file or directory: ''",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mFileNotFoundError\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-13-95649b9f9614>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;31m# Import data projection\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mnewinput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread_csv\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m''\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;31m# Order of the variable are important\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mX_new\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnewinput\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mloc\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m258\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/Tensorflow/lib/python3.7/site-packages/pandas/io/parsers.py\u001b[0m in \u001b[0;36mread_csv\u001b[0;34m(filepath_or_buffer, sep, delimiter, header, names, index_col, usecols, squeeze, prefix, mangle_dupe_cols, dtype, engine, converters, true_values, false_values, skipinitialspace, skiprows, skipfooter, nrows, na_values, keep_default_na, na_filter, verbose, skip_blank_lines, parse_dates, infer_datetime_format, keep_date_col, date_parser, dayfirst, cache_dates, iterator, chunksize, compression, thousands, decimal, lineterminator, quotechar, quoting, doublequote, escapechar, comment, encoding, dialect, error_bad_lines, warn_bad_lines, delim_whitespace, low_memory, memory_map, float_precision)\u001b[0m\n\u001b[1;32m    684\u001b[0m     )\n\u001b[1;32m    685\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 686\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0m_read\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfilepath_or_buffer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    687\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    688\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/Tensorflow/lib/python3.7/site-packages/pandas/io/parsers.py\u001b[0m in \u001b[0;36m_read\u001b[0;34m(filepath_or_buffer, kwds)\u001b[0m\n\u001b[1;32m    450\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    451\u001b[0m     \u001b[0;31m# Create the parser.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 452\u001b[0;31m     \u001b[0mparser\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mTextFileReader\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfp_or_buf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    453\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    454\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mchunksize\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0miterator\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/Tensorflow/lib/python3.7/site-packages/pandas/io/parsers.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, f, engine, **kwds)\u001b[0m\n\u001b[1;32m    944\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptions\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"has_index_names\"\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mkwds\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"has_index_names\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    945\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 946\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_make_engine\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    947\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    948\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mclose\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/Tensorflow/lib/python3.7/site-packages/pandas/io/parsers.py\u001b[0m in \u001b[0;36m_make_engine\u001b[0;34m(self, engine)\u001b[0m\n\u001b[1;32m   1176\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_make_engine\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mengine\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"c\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1177\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mengine\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m\"c\"\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1178\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_engine\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mCParserWrapper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptions\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1179\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1180\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mengine\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m\"python\"\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/Tensorflow/lib/python3.7/site-packages/pandas/io/parsers.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, src, **kwds)\u001b[0m\n\u001b[1;32m   2006\u001b[0m         \u001b[0mkwds\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m\"usecols\"\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0musecols\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2007\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2008\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_reader\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mparsers\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTextReader\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msrc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2009\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0munnamed_cols\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_reader\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0munnamed_cols\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2010\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/parsers.pyx\u001b[0m in \u001b[0;36mpandas._libs.parsers.TextReader.__cinit__\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;32mpandas/_libs/parsers.pyx\u001b[0m in \u001b[0;36mpandas._libs.parsers.TextReader._setup_parser_source\u001b[0;34m()\u001b[0m\n",
      "\u001b[0;31mFileNotFoundError\u001b[0m: [Errno 2] No such file or directory: ''"
     ]
    }
   ],
   "source": [
    "# Import data projection\n",
    "newinput = pd.read_csv('')\n",
    "\n",
    "# Order of the variable are important\n",
    "X_new = newinput.loc[:,[1,258]].values \n",
    "X_new"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot histoy and future data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_history_future(y_train, prediction, model_name):\n",
    "    \n",
    "    plt.figure(figsize=(10, 6))\n",
    "    \n",
    "    range_history = len(y_train)\n",
    "    range_future = list(range(range_history, range_history + len(prediction)))\n",
    "    \n",
    "    plt.plot(np.arange(range_history), np.array(y_train), label='History')\n",
    "    plt.plot(range_future, np.array(prediction),label='Prediction')\n",
    "    plt.legend(loc='upper right')\n",
    "    plt.xlabel('Scaffolding')\n",
    "    plt.ylabel('Position')\n",
    "    plt.savefig('path/$.jpg', format = 'jpg', dpi = 1000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## External Data Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'X_new' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-15-ee5a288bb74b>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     14\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0mprediction_actual\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     15\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 16\u001b[0;31m \u001b[0mprediction\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mclassify\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_new\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mTIME_STEPS\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     17\u001b[0m \u001b[0mplot_history_future\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mprediction\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'BiLSTM'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'X_new' is not defined"
     ]
    }
   ],
   "source": [
    "def classify(X_input, time_steps):\n",
    "    # Scale the unseen input with the scaler fitted on the train set\n",
    "    X = input_scaler.transform(X_input)\n",
    "    # Reshape unseen data to a 3D input\n",
    "    Xs = []\n",
    "    for i in range(len(X) - time_steps):\n",
    "        v = X[i:i+time_steps, :]\n",
    "        Xs.append(v)\n",
    "    X_transformed = np.array(Xs)\n",
    "    \n",
    "    # Make prediction for unseen data using LSTM model\n",
    "    prediction = model_lstm.predict(X_transformed)\n",
    "    prediction_actual = scaler_y.inverse_transform(prediction)\n",
    "    return prediction_actual\n",
    "\n",
    "prediction = classify(X_new, TIME_STEPS)\n",
    "plot_history_future(y_train, prediction, 'BiLSTM')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create a Pickle file using serialization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "pickle.dump(X_test, open('model_LSTM.pkl','wb'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading model to compare the results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = pickle.load(open('model_LSTM.pkl','rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Tensorflow",
   "language": "python",
   "name": "tensorflow"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}