Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\", message=\"numpy.dtype size changed\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing Training Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "training_set = pd.read_csv('Google_Stock_Price_Train.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Date</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1/3/2012</td>\n",
       "      <td>325.25</td>\n",
       "      <td>332.83</td>\n",
       "      <td>324.97</td>\n",
       "      <td>663.59</td>\n",
       "      <td>7,380,500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1/4/2012</td>\n",
       "      <td>331.27</td>\n",
       "      <td>333.87</td>\n",
       "      <td>329.08</td>\n",
       "      <td>666.45</td>\n",
       "      <td>5,749,400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1/5/2012</td>\n",
       "      <td>329.83</td>\n",
       "      <td>330.75</td>\n",
       "      <td>326.89</td>\n",
       "      <td>657.21</td>\n",
       "      <td>6,590,300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1/6/2012</td>\n",
       "      <td>328.34</td>\n",
       "      <td>328.77</td>\n",
       "      <td>323.68</td>\n",
       "      <td>648.24</td>\n",
       "      <td>5,405,900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1/9/2012</td>\n",
       "      <td>322.04</td>\n",
       "      <td>322.29</td>\n",
       "      <td>309.46</td>\n",
       "      <td>620.76</td>\n",
       "      <td>11,688,800</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Date    Open    High     Low   Close      Volume\n",
       "0  1/3/2012  325.25  332.83  324.97  663.59   7,380,500\n",
       "1  1/4/2012  331.27  333.87  329.08  666.45   5,749,400\n",
       "2  1/5/2012  329.83  330.75  326.89  657.21   6,590,300\n",
       "3  1/6/2012  328.34  328.77  323.68  648.24   5,405,900\n",
       "4  1/9/2012  322.04  322.29  309.46  620.76  11,688,800"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "training_set.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "training_set = training_set.iloc[:,1:2].values #Getting open stock prices"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "sc = MinMaxScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "training_set = sc.fit_transform(training_set)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Getting Inputs & Outputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1258, 1)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "training_set.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = training_set[0:1257]  #Stock Price at time T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train = training_set[1:1258]  #Prediction for Stock Price at time T+1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reshaping X_train to include timestamp alongwith no of observations & no of features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1257, 1)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = np.reshape(X_train,(1257,1,1)) #Second Dimension is Timestamp = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1257, 1, 1)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Building RNN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense,LSTM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "regressor = Sequential()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Adding the Input Layer & the LSTM Layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "regressor.add(LSTM(units=4,activation='sigmoid',input_shape=(None,1))) #None referes to the timestamp, shows that model can accept any timestamp can also be specified as 1 since our model uses 1 timestamp"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Adding the Output Layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "regressor.add(Dense(units=1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Compiling RNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "regressor.compile(optimizer='adam',loss='mean_squared_error')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Fitting RNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "1257/1257 [==============================] - 1s 693us/step - loss: 0.8800\n",
      "Epoch 2/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.7226\n",
      "Epoch 3/200\n",
      "1257/1257 [==============================] - 0s 93us/step - loss: 0.5939\n",
      "Epoch 4/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 0.4873\n",
      "Epoch 5/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.3997\n",
      "Epoch 6/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.3271\n",
      "Epoch 7/200\n",
      "1257/1257 [==============================] - 0s 92us/step - loss: 0.2672\n",
      "Epoch 8/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.2191\n",
      "Epoch 9/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.1805\n",
      "Epoch 10/200\n",
      "1257/1257 [==============================] - 0s 88us/step - loss: 0.1500\n",
      "Epoch 11/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.1261\n",
      "Epoch 12/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.1077\n",
      "Epoch 13/200\n",
      "1257/1257 [==============================] - 0s 94us/step - loss: 0.0941\n",
      "Epoch 14/200\n",
      "1257/1257 [==============================] - 0s 108us/step - loss: 0.0839\n",
      "Epoch 15/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.0767\n",
      "Epoch 16/200\n",
      "1257/1257 [==============================] - 0s 96us/step - loss: 0.0715\n",
      "Epoch 17/200\n",
      "1257/1257 [==============================] - 0s 95us/step - loss: 0.0678\n",
      "Epoch 18/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 0.0652\n",
      "Epoch 19/200\n",
      "1257/1257 [==============================] - 0s 99us/step - loss: 0.0633\n",
      "Epoch 20/200\n",
      "1257/1257 [==============================] - 0s 106us/step - loss: 0.0620\n",
      "Epoch 21/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 0.0609\n",
      "Epoch 22/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0600\n",
      "Epoch 23/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 0.0592\n",
      "Epoch 24/200\n",
      "1257/1257 [==============================] - 0s 88us/step - loss: 0.0585\n",
      "Epoch 25/200\n",
      "1257/1257 [==============================] - 0s 101us/step - loss: 0.0578\n",
      "Epoch 26/200\n",
      "1257/1257 [==============================] - 0s 101us/step - loss: 0.0571\n",
      "Epoch 27/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0564\n",
      "Epoch 28/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0556\n",
      "Epoch 29/200\n",
      "1257/1257 [==============================] - 0s 93us/step - loss: 0.0549\n",
      "Epoch 30/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.0541\n",
      "Epoch 31/200\n",
      "1257/1257 [==============================] - 0s 92us/step - loss: 0.0533\n",
      "Epoch 32/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.0525\n",
      "Epoch 33/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0516\n",
      "Epoch 34/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0508\n",
      "Epoch 35/200\n",
      "1257/1257 [==============================] - 0s 101us/step - loss: 0.0499\n",
      "Epoch 36/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 0.0489\n",
      "Epoch 37/200\n",
      "1257/1257 [==============================] - 0s 96us/step - loss: 0.0480\n",
      "Epoch 38/200\n",
      "1257/1257 [==============================] - 0s 95us/step - loss: 0.0470\n",
      "Epoch 39/200\n",
      "1257/1257 [==============================] - 0s 96us/step - loss: 0.0460\n",
      "Epoch 40/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0449\n",
      "Epoch 41/200\n",
      "1257/1257 [==============================] - 0s 88us/step - loss: 0.0439\n",
      "Epoch 42/200\n",
      "1257/1257 [==============================] - 0s 88us/step - loss: 0.0428\n",
      "Epoch 43/200\n",
      "1257/1257 [==============================] - 0s 85us/step - loss: 0.0418\n",
      "Epoch 44/200\n",
      "1257/1257 [==============================] - 0s 98us/step - loss: 0.0407\n",
      "Epoch 45/200\n",
      "1257/1257 [==============================] - 0s 98us/step - loss: 0.0396\n",
      "Epoch 46/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 0.0385\n",
      "Epoch 47/200\n",
      "1257/1257 [==============================] - 0s 97us/step - loss: 0.0374\n",
      "Epoch 48/200\n",
      "1257/1257 [==============================] - 0s 92us/step - loss: 0.0363\n",
      "Epoch 49/200\n",
      "1257/1257 [==============================] - 0s 93us/step - loss: 0.0351\n",
      "Epoch 50/200\n",
      "1257/1257 [==============================] - 0s 92us/step - loss: 0.0340\n",
      "Epoch 51/200\n",
      "1257/1257 [==============================] - 0s 95us/step - loss: 0.0329\n",
      "Epoch 52/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 0.0318\n",
      "Epoch 53/200\n",
      "1257/1257 [==============================] - 0s 93us/step - loss: 0.0307\n",
      "Epoch 54/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0296\n",
      "Epoch 55/200\n",
      "1257/1257 [==============================] - 0s 86us/step - loss: 0.0284\n",
      "Epoch 56/200\n",
      "1257/1257 [==============================] - 0s 94us/step - loss: 0.0273\n",
      "Epoch 57/200\n",
      "1257/1257 [==============================] - 0s 88us/step - loss: 0.0262\n",
      "Epoch 58/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0251\n",
      "Epoch 59/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0240\n",
      "Epoch 60/200\n",
      "1257/1257 [==============================] - 0s 84us/step - loss: 0.0229\n",
      "Epoch 61/200\n",
      "1257/1257 [==============================] - 0s 88us/step - loss: 0.0218\n",
      "Epoch 62/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.0207\n",
      "Epoch 63/200\n",
      "1257/1257 [==============================] - 0s 86us/step - loss: 0.0196\n",
      "Epoch 64/200\n",
      "1257/1257 [==============================] - 0s 84us/step - loss: 0.0185\n",
      "Epoch 65/200\n",
      "1257/1257 [==============================] - 0s 86us/step - loss: 0.0175\n",
      "Epoch 66/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.0164\n",
      "Epoch 67/200\n",
      "1257/1257 [==============================] - 0s 85us/step - loss: 0.0154\n",
      "Epoch 68/200\n",
      "1257/1257 [==============================] - 0s 93us/step - loss: 0.0144\n",
      "Epoch 69/200\n",
      "1257/1257 [==============================] - 0s 94us/step - loss: 0.0135\n",
      "Epoch 70/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0125\n",
      "Epoch 71/200\n",
      "1257/1257 [==============================] - 0s 83us/step - loss: 0.0116\n",
      "Epoch 72/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 0.0108\n",
      "Epoch 73/200\n",
      "1257/1257 [==============================] - 0s 101us/step - loss: 0.0099\n",
      "Epoch 74/200\n",
      "1257/1257 [==============================] - 0s 102us/step - loss: 0.0091\n",
      "Epoch 75/200\n",
      "1257/1257 [==============================] - 0s 85us/step - loss: 0.0084\n",
      "Epoch 76/200\n",
      "1257/1257 [==============================] - 0s 85us/step - loss: 0.0077\n",
      "Epoch 77/200\n",
      "1257/1257 [==============================] - 0s 86us/step - loss: 0.0070\n",
      "Epoch 78/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.0063\n",
      "Epoch 79/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 0.0057\n",
      "Epoch 80/200\n",
      "1257/1257 [==============================] - 0s 94us/step - loss: 0.0052\n",
      "Epoch 81/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 0.0047\n",
      "Epoch 82/200\n",
      "1257/1257 [==============================] - 0s 92us/step - loss: 0.0042\n",
      "Epoch 83/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.0037\n",
      "Epoch 84/200\n",
      "1257/1257 [==============================] - 0s 84us/step - loss: 0.0033\n",
      "Epoch 85/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.0030\n",
      "Epoch 86/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 0.0026\n",
      "Epoch 87/200\n",
      "1257/1257 [==============================] - 0s 110us/step - loss: 0.0023\n",
      "Epoch 88/200\n",
      "1257/1257 [==============================] - 0s 113us/step - loss: 0.0020\n",
      "Epoch 89/200\n",
      "1257/1257 [==============================] - 0s 115us/step - loss: 0.0018\n",
      "Epoch 90/200\n",
      "1257/1257 [==============================] - 0s 113us/step - loss: 0.0016\n",
      "Epoch 91/200\n",
      "1257/1257 [==============================] - 0s 113us/step - loss: 0.0014\n",
      "Epoch 92/200\n",
      "1257/1257 [==============================] - 0s 108us/step - loss: 0.0012\n",
      "Epoch 93/200\n",
      "1257/1257 [==============================] - 0s 110us/step - loss: 0.0011\n",
      "Epoch 94/200\n",
      "1257/1257 [==============================] - 0s 112us/step - loss: 9.5242e-04\n",
      "Epoch 95/200\n",
      "1257/1257 [==============================] - 0s 108us/step - loss: 8.4161e-04\n",
      "Epoch 96/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1257/1257 [==============================] - 0s 110us/step - loss: 7.4641e-04\n",
      "Epoch 97/200\n",
      "1257/1257 [==============================] - 0s 86us/step - loss: 6.6493e-04\n",
      "Epoch 98/200\n",
      "1257/1257 [==============================] - 0s 101us/step - loss: 5.9565e-04\n",
      "Epoch 99/200\n",
      "1257/1257 [==============================] - 0s 107us/step - loss: 5.3812e-04\n",
      "Epoch 100/200\n",
      "1257/1257 [==============================] - 0s 106us/step - loss: 4.8814e-04\n",
      "Epoch 101/200\n",
      "1257/1257 [==============================] - 0s 103us/step - loss: 4.4730e-04\n",
      "Epoch 102/200\n",
      "1257/1257 [==============================] - 0s 102us/step - loss: 4.1309e-04\n",
      "Epoch 103/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 3.8503e-04\n",
      "Epoch 104/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 3.6288e-04\n",
      "Epoch 105/200\n",
      "1257/1257 [==============================] - 0s 88us/step - loss: 3.4387e-04\n",
      "Epoch 106/200\n",
      "1257/1257 [==============================] - 0s 98us/step - loss: 3.2807e-04\n",
      "Epoch 107/200\n",
      "1257/1257 [==============================] - 0s 107us/step - loss: 3.1597e-04\n",
      "Epoch 108/200\n",
      "1257/1257 [==============================] - 0s 99us/step - loss: 3.0608e-04\n",
      "Epoch 109/200\n",
      "1257/1257 [==============================] - 0s 110us/step - loss: 2.9812e-04\n",
      "Epoch 110/200\n",
      "1257/1257 [==============================] - 0s 120us/step - loss: 2.9157e-04\n",
      "Epoch 111/200\n",
      "1257/1257 [==============================] - 0s 93us/step - loss: 2.8677e-04\n",
      "Epoch 112/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 2.8250e-04\n",
      "Epoch 113/200\n",
      "1257/1257 [==============================] - 0s 90us/step - loss: 2.7940e-04\n",
      "Epoch 114/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 2.7684e-04\n",
      "Epoch 115/200\n",
      "1257/1257 [==============================] - 0s 102us/step - loss: 2.7463e-04\n",
      "Epoch 116/200\n",
      "1257/1257 [==============================] - 0s 115us/step - loss: 2.7330e-04\n",
      "Epoch 117/200\n",
      "1257/1257 [==============================] - 0s 114us/step - loss: 2.7186e-04\n",
      "Epoch 118/200\n",
      "1257/1257 [==============================] - 0s 120us/step - loss: 2.7099e-04\n",
      "Epoch 119/200\n",
      "1257/1257 [==============================] - 0s 115us/step - loss: 2.7122e-04\n",
      "Epoch 120/200\n",
      "1257/1257 [==============================] - 0s 110us/step - loss: 2.6843e-04\n",
      "Epoch 121/200\n",
      "1257/1257 [==============================] - 0s 97us/step - loss: 2.6813e-04\n",
      "Epoch 122/200\n",
      "1257/1257 [==============================] - 0s 85us/step - loss: 2.6777e-04\n",
      "Epoch 123/200\n",
      "1257/1257 [==============================] - 0s 89us/step - loss: 2.6673e-04\n",
      "Epoch 124/200\n",
      "1257/1257 [==============================] - 0s 92us/step - loss: 2.6590e-04\n",
      "Epoch 125/200\n",
      "1257/1257 [==============================] - 0s 94us/step - loss: 2.6504e-04\n",
      "Epoch 126/200\n",
      "1257/1257 [==============================] - 0s 105us/step - loss: 2.6590e-04\n",
      "Epoch 127/200\n",
      "1257/1257 [==============================] - 0s 112us/step - loss: 2.6456e-04\n",
      "Epoch 128/200\n",
      "1257/1257 [==============================] - 0s 123us/step - loss: 2.6356e-04\n",
      "Epoch 129/200\n",
      "1257/1257 [==============================] - 0s 123us/step - loss: 2.6337e-04\n",
      "Epoch 130/200\n",
      "1257/1257 [==============================] - 0s 113us/step - loss: 2.6221e-04\n",
      "Epoch 131/200\n",
      "1257/1257 [==============================] - 0s 105us/step - loss: 2.6175e-04\n",
      "Epoch 132/200\n",
      "1257/1257 [==============================] - 0s 113us/step - loss: 2.6108e-04\n",
      "Epoch 133/200\n",
      "1257/1257 [==============================] - 0s 116us/step - loss: 2.6099e-04\n",
      "Epoch 134/200\n",
      "1257/1257 [==============================] - 0s 102us/step - loss: 2.6051e-04\n",
      "Epoch 135/200\n",
      "1257/1257 [==============================] - 0s 100us/step - loss: 2.5982e-04\n",
      "Epoch 136/200\n",
      "1257/1257 [==============================] - 0s 106us/step - loss: 2.5980e-04\n",
      "Epoch 137/200\n",
      "1257/1257 [==============================] - 0s 113us/step - loss: 2.5928e-04\n",
      "Epoch 138/200\n",
      "1257/1257 [==============================] - 0s 111us/step - loss: 2.5906e-04\n",
      "Epoch 139/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 2.5930e-04\n",
      "Epoch 140/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 2.5878e-04\n",
      "Epoch 141/200\n",
      "1257/1257 [==============================] - 0s 93us/step - loss: 2.5932e-04\n",
      "Epoch 142/200\n",
      "1257/1257 [==============================] - 0s 97us/step - loss: 2.5818e-04\n",
      "Epoch 143/200\n",
      "1257/1257 [==============================] - 0s 108us/step - loss: 2.5768e-04\n",
      "Epoch 144/200\n",
      "1257/1257 [==============================] - 0s 114us/step - loss: 2.5774e-04\n",
      "Epoch 145/200\n",
      "1257/1257 [==============================] - 0s 115us/step - loss: 2.5874e-04\n",
      "Epoch 146/200\n",
      "1257/1257 [==============================] - 0s 110us/step - loss: 2.5824e-04\n",
      "Epoch 147/200\n",
      "1257/1257 [==============================] - 0s 121us/step - loss: 2.5706e-04\n",
      "Epoch 148/200\n",
      "1257/1257 [==============================] - 0s 101us/step - loss: 2.5641e-04\n",
      "Epoch 149/200\n",
      "1257/1257 [==============================] - 0s 101us/step - loss: 2.5611e-04\n",
      "Epoch 150/200\n",
      "1257/1257 [==============================] - 0s 108us/step - loss: 2.5611e-04\n",
      "Epoch 151/200\n",
      "1257/1257 [==============================] - 0s 106us/step - loss: 2.5606e-04\n",
      "Epoch 152/200\n",
      "1257/1257 [==============================] - 0s 108us/step - loss: 2.5540e-04\n",
      "Epoch 153/200\n",
      "1257/1257 [==============================] - 0s 113us/step - loss: 2.5507e-04\n",
      "Epoch 154/200\n",
      "1257/1257 [==============================] - 0s 113us/step - loss: 2.5480e-04\n",
      "Epoch 155/200\n",
      "1257/1257 [==============================] - 0s 110us/step - loss: 2.5491e-04\n",
      "Epoch 156/200\n",
      "1257/1257 [==============================] - 0s 112us/step - loss: 2.5521e-04\n",
      "Epoch 157/200\n",
      "1257/1257 [==============================] - 0s 109us/step - loss: 2.5418e-04\n",
      "Epoch 158/200\n",
      "1257/1257 [==============================] - 0s 115us/step - loss: 2.5458e-04\n",
      "Epoch 159/200\n",
      "1257/1257 [==============================] - 0s 102us/step - loss: 2.5393e-04\n",
      "Epoch 160/200\n",
      "1257/1257 [==============================] - 0s 99us/step - loss: 2.5351e-04\n",
      "Epoch 161/200\n",
      "1257/1257 [==============================] - 0s 107us/step - loss: 2.5373e-04\n",
      "Epoch 162/200\n",
      "1257/1257 [==============================] - 0s 112us/step - loss: 2.5335e-04\n",
      "Epoch 163/200\n",
      "1257/1257 [==============================] - 0s 125us/step - loss: 2.5354e-04\n",
      "Epoch 164/200\n",
      "1257/1257 [==============================] - 0s 106us/step - loss: 2.5343e-04\n",
      "Epoch 165/200\n",
      "1257/1257 [==============================] - 0s 103us/step - loss: 2.5295e-04\n",
      "Epoch 166/200\n",
      "1257/1257 [==============================] - 0s 100us/step - loss: 2.5301e-04\n",
      "Epoch 167/200\n",
      "1257/1257 [==============================] - 0s 93us/step - loss: 2.5370e-04\n",
      "Epoch 168/200\n",
      "1257/1257 [==============================] - 0s 96us/step - loss: 2.5303e-04\n",
      "Epoch 169/200\n",
      "1257/1257 [==============================] - 0s 114us/step - loss: 2.5374e-04\n",
      "Epoch 170/200\n",
      "1257/1257 [==============================] - 0s 108us/step - loss: 2.5237e-04\n",
      "Epoch 171/200\n",
      "1257/1257 [==============================] - 0s 105us/step - loss: 2.5186e-04\n",
      "Epoch 172/200\n",
      "1257/1257 [==============================] - 0s 106us/step - loss: 2.5328e-04\n",
      "Epoch 173/200\n",
      "1257/1257 [==============================] - 0s 108us/step - loss: 2.5307e-04\n",
      "Epoch 174/200\n",
      "1257/1257 [==============================] - 0s 110us/step - loss: 2.5195e-04\n",
      "Epoch 175/200\n",
      "1257/1257 [==============================] - 0s 102us/step - loss: 2.5236e-04\n",
      "Epoch 176/200\n",
      "1257/1257 [==============================] - 0s 106us/step - loss: 2.5261e-04\n",
      "Epoch 177/200\n",
      "1257/1257 [==============================] - 0s 101us/step - loss: 2.5254e-04\n",
      "Epoch 178/200\n",
      "1257/1257 [==============================] - 0s 102us/step - loss: 2.5234e-04\n",
      "Epoch 179/200\n",
      "1257/1257 [==============================] - 0s 112us/step - loss: 2.5250e-04\n",
      "Epoch 180/200\n",
      "1257/1257 [==============================] - 0s 104us/step - loss: 2.5282e-04\n",
      "Epoch 181/200\n",
      "1257/1257 [==============================] - 0s 99us/step - loss: 2.5157e-04\n",
      "Epoch 182/200\n",
      "1257/1257 [==============================] - 0s 106us/step - loss: 2.5336e-04\n",
      "Epoch 183/200\n",
      "1257/1257 [==============================] - 0s 99us/step - loss: 2.5207e-04\n",
      "Epoch 184/200\n",
      "1257/1257 [==============================] - 0s 104us/step - loss: 2.5329e-04\n",
      "Epoch 185/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1257/1257 [==============================] - 0s 115us/step - loss: 2.5338e-04\n",
      "Epoch 186/200\n",
      "1257/1257 [==============================] - 0s 106us/step - loss: 2.5285e-04\n",
      "Epoch 187/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 2.5112e-04\n",
      "Epoch 188/200\n",
      "1257/1257 [==============================] - 0s 88us/step - loss: 2.5257e-04\n",
      "Epoch 189/200\n",
      "1257/1257 [==============================] - 0s 85us/step - loss: 2.5180e-04\n",
      "Epoch 190/200\n",
      "1257/1257 [==============================] - 0s 88us/step - loss: 2.5052e-04\n",
      "Epoch 191/200\n",
      "1257/1257 [==============================] - 0s 91us/step - loss: 2.5183e-04\n",
      "Epoch 192/200\n",
      "1257/1257 [==============================] - 0s 110us/step - loss: 2.5031e-04\n",
      "Epoch 193/200\n",
      "1257/1257 [==============================] - 0s 100us/step - loss: 2.5173e-04\n",
      "Epoch 194/200\n",
      "1257/1257 [==============================] - 0s 112us/step - loss: 2.5141e-04\n",
      "Epoch 195/200\n",
      "1257/1257 [==============================] - 0s 120us/step - loss: 2.5164e-04\n",
      "Epoch 196/200\n",
      "1257/1257 [==============================] - 0s 107us/step - loss: 2.5216e-04\n",
      "Epoch 197/200\n",
      "1257/1257 [==============================] - 0s 112us/step - loss: 2.5379e-04\n",
      "Epoch 198/200\n",
      "1257/1257 [==============================] - 0s 109us/step - loss: 2.5125e-04\n",
      "Epoch 199/200\n",
      "1257/1257 [==============================] - 0s 117us/step - loss: 2.5276e-04\n",
      "Epoch 200/200\n",
      "1257/1257 [==============================] - 0s 122us/step - loss: 2.5253e-04\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x11142d630>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regressor.fit(X_train,y_train,batch_size=32,epochs=200)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Making Predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing Test Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_set = pd.read_csv('Google_Stock_Price_Test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Date</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1/3/2017</td>\n",
       "      <td>778.81</td>\n",
       "      <td>789.63</td>\n",
       "      <td>775.80</td>\n",
       "      <td>786.14</td>\n",
       "      <td>1,657,300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1/4/2017</td>\n",
       "      <td>788.36</td>\n",
       "      <td>791.34</td>\n",
       "      <td>783.16</td>\n",
       "      <td>786.90</td>\n",
       "      <td>1,073,000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1/5/2017</td>\n",
       "      <td>786.08</td>\n",
       "      <td>794.48</td>\n",
       "      <td>785.02</td>\n",
       "      <td>794.02</td>\n",
       "      <td>1,335,200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1/6/2017</td>\n",
       "      <td>795.26</td>\n",
       "      <td>807.90</td>\n",
       "      <td>792.20</td>\n",
       "      <td>806.15</td>\n",
       "      <td>1,640,200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1/9/2017</td>\n",
       "      <td>806.40</td>\n",
       "      <td>809.97</td>\n",
       "      <td>802.83</td>\n",
       "      <td>806.65</td>\n",
       "      <td>1,272,400</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Date    Open    High     Low   Close     Volume\n",
       "0  1/3/2017  778.81  789.63  775.80  786.14  1,657,300\n",
       "1  1/4/2017  788.36  791.34  783.16  786.90  1,073,000\n",
       "2  1/5/2017  786.08  794.48  785.02  794.02  1,335,200\n",
       "3  1/6/2017  795.26  807.90  792.20  806.15  1,640,200\n",
       "4  1/9/2017  806.40  809.97  802.83  806.65  1,272,400"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_set.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "real_stock_price = test_set.iloc[:,1:2].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs = real_stock_price"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Scaling Test Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs = sc.transform(inputs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(20, 1)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inputs.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Reshaping Test Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs = np.reshape(inputs,(20,1,1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "predicted_stock_price = regressor.predict(inputs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.92572939],\n",
       "       [ 0.94204855],\n",
       "       [ 0.93815601],\n",
       "       [ 0.95381546],\n",
       "       [ 0.96747911],\n",
       "       [ 0.96874464],\n",
       "       [ 0.96626461],\n",
       "       [ 0.96812069],\n",
       "       [ 0.96841538],\n",
       "       [ 0.96806872],\n",
       "       [ 0.96696746],\n",
       "       [ 0.96636879],\n",
       "       [ 0.96792138],\n",
       "       [ 0.96821606],\n",
       "       [ 0.98119867],\n",
       "       [ 0.98747003],\n",
       "       [ 0.99445474],\n",
       "       [ 0.99181497],\n",
       "       [ 0.9746232 ],\n",
       "       [ 0.95654106]], dtype=float32)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predicted_stock_price"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "predicted_stock_price = sc.inverse_transform(predicted_stock_price) #Bringing to the same scale as original data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Visualising Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x111d7a160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(real_stock_price,color='red',label='Real Google Stock Price')\n",
    "plt.plot(predicted_stock_price,color='blue',label='Predicted Google Stock Price')\n",
    "plt.title('Google Stock Price Prediction')\n",
    "plt.xlabel('Time')\n",
    "plt.ylabel('Google Stock Price')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Evaluating the RNN "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "from sklearn.metrics import mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "rmse = math.sqrt(mean_squared_error(real_stock_price,predicted_stock_price))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "11.032800780219805"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rmse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "807.52600000000007"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.average(real_stock_price)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "error_percentage = (rmse/807)*100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.3671376431499138"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "error_percentage"
   ]
  },
  {
   "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}