Predict_with_BRNN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# univariate bidirectional lstm example\n",
    "from numpy import array\n",
    "from keras.models import Sequential\n",
    "from keras.layers import LSTM\n",
    "from keras.layers import Dense\n",
    "from keras.layers import Bidirectional\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import StandardScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60\n",
      "10\n"
     ]
    }
   ],
   "source": [
    "# split a univariate sequence\n",
    "def split_sequence(sequence, n_steps):\n",
    "    X, y = list(), list()\n",
    "    for i in range(len(sequence)):\n",
    "        # find the end of this pattern\n",
    "        end_ix = i + n_steps\n",
    "        # check if we are beyond the sequence\n",
    "        if end_ix > len(sequence)-1:\n",
    "            break\n",
    "        # gather input and output parts of the pattern\n",
    "        seq_x, seq_y = sequence[i:end_ix], sequence[end_ix]\n",
    "        X.append(seq_x)\n",
    "        y.append(seq_y)\n",
    "    return array(X), array(y)\n",
    "\n",
    "#Read the dataset\n",
    "df = pd.read_csv(\"train.csv\")\n",
    "\n",
    "df = df.drop(\"Province_State\",axis=1)\n",
    "df = df[df[\"Country_Region\"]==\"India\"]\n",
    "\n",
    "confirmed_data = df[\"ConfirmedCases\"].values\n",
    "fatalities_data = df[\"Fatalities\"].values\n",
    "dates = df[\"Date\"]\n",
    "train_date = dates[:60]\n",
    "test_date = dates[-10:]\n",
    "print(len(train_date))\n",
    "print(len(test_date))\n",
    "\n",
    "#normalising the dataset\n",
    "scaler = StandardScaler()\n",
    "scaled_confirmed = scaler.fit_transform(confirmed_data.reshape(-1, 1))\n",
    "scaled_fatalities = scaler.fit_transform(fatalities_data.reshape(-1, 1))\n",
    "\n",
    "train_confirmed = scaled_confirmed[:60]\n",
    "test_confirmed = scaled_confirmed[60:]\n",
    "\n",
    "\n",
    "train_fatalities = scaled_fatalities[:60]\n",
    "test_fatalities = scaled_fatalities[60:]\n",
    "\n",
    "inp_size = 7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x1c8c5f24808>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Model to predict Confirmed cases\n",
    "\n",
    "X,y = split_sequence(train_confirmed,inp_size)\n",
    "\n",
    "# reshape from [samples, timesteps] into [samples, timesteps, features]\n",
    "n_features = 1\n",
    "X = X.reshape((X.shape[0], X.shape[1], n_features))\n",
    "# print(X)\n",
    "\n",
    "# define model\n",
    "model = Sequential()\n",
    "model.add(Bidirectional(LSTM(50, activation='relu'), input_shape=(inp_size, n_features)))\n",
    "model.add(Dense(1))\n",
    "model.compile(optimizer='adam', loss='mse')\n",
    "\n",
    "# fit model\n",
    "model.fit(X, y, epochs=400, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "confirmed_test_x,confirmed_test_y = split_sequence(test_confirmed,inp_size)\n",
    "# print(confirmed_test_x,confirmed_test_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.34991008, 0.41782272, 0.47939843, 0.55790204, 0.6434252, 0.7686823, 0.93325406, 1.1433891, 1.4695555, 1.876453]\n",
      "10\n"
     ]
    }
   ],
   "source": [
    "\n",
    "pred_conf_cases = []\n",
    "for i in confirmed_test_x:\n",
    "    x_input = array(i)\n",
    "    x_input = x_input.reshape((1, inp_size, n_features))\n",
    "    yhat = model.predict(x_input, verbose=0)\n",
    "    pred_conf_cases.append(yhat[0][0])\n",
    "print(pred_conf_cases)\n",
    "print(len(pred_conf_cases))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (18,9))\n",
    "plt.plot(df['Date'],scaled_confirmed,label=\"Training Data\")\n",
    "plt.plot(train_date,train_confirmed,label=\"Testing Data\")\n",
    "plt.plot(test_date,pred_conf_cases,label=\"Predicted Data\")\n",
    "plt.xticks(rotation=90)\n",
    "plt.xlabel('Date',fontsize=18)\n",
    "plt.ylabel('confirmed cases',fontsize=18)\n",
    "plt.title(\"graph for Confirmed Cases\")\n",
    "plt.legend()\n",
    "plt.show()\n",
    "# train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x1c8c8b1ce88>"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Model to predict Fatalities\n",
    "X2,y2 = split_sequence(train_fatalities,inp_size)\n",
    "# reshape from [samples, timesteps] into [samples, timesteps, features]\n",
    "n_features = 1\n",
    "X2 = X2.reshape((X2.shape[0], X2.shape[1], n_features))\n",
    "# print(X2)\n",
    "# define model\n",
    "model2 = Sequential()\n",
    "model2.add(Bidirectional(LSTM(50, activation='relu'), input_shape=(inp_size, n_features)))\n",
    "model2.add(Dense(1))\n",
    "model2.compile(optimizer='adam', loss='mse')\n",
    "# fit model\n",
    "model2.fit(X2, y2, epochs=400, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "Fatalities_test_x,Fatalities_test_y = split_sequence(test_fatalities,inp_size)\n",
    "# print(Fatalities_test_x,Fatalities_test_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.046254262, 0.075326666, 0.096385926, 0.11979187, 0.14568666, 0.16429186, 0.19092241, 0.23568699, 0.29810685, 0.40468368]\n"
     ]
    }
   ],
   "source": [
    "#predicted2 = test[inp_size:]\n",
    "#predicted_Date2 = df[\"Date\"]\n",
    "pred_fata_cases = []\n",
    "for i in Fatalities_test_x:\n",
    "    x_input = array(i)\n",
    "    x_input = x_input.reshape((1, inp_size, n_features))\n",
    "    yhat = model2.predict(x_input, verbose=0)\n",
    "    pred_fata_cases.append(yhat[0][0])\n",
    "print(pred_fata_cases)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.figure(figsize = (18,9))\n",
    "plt.plot(df['Date'],scaled_fatalities,label=\"Training Data\")\n",
    "plt.plot(train_date,train_fatalities,label=\"Testing Data\")\n",
    "plt.plot(test_date,pred_fata_cases,label=\"Predicted Data\")\n",
    "plt.xticks(rotation=90)\n",
    "plt.xlabel('Date',fontsize=18)\n",
    "plt.ylabel('Fatalities',fontsize=18)\n",
    "plt.title(\"graph for fatalities\")\n",
    "plt.legend()\n",
    "plt.show()\n",
    "# train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_7\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "bidirectional_7 (Bidirection (None, 100)               20800     \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              (None, 1)                 101       \n",
      "=================================================================\n",
      "Total params: 20,901\n",
      "Trainable params: 20,901\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}