notebook/21_keras_time_series_prediction_lstm.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using matplotlib backend: MacOSX\n",
      "Populating the interactive namespace from numpy and matplotlib\n",
      "CPython 2.7.12\n",
      "IPython 5.1.0\n",
      "\n",
      "scipy 0.18.1\n",
      "sklearn 0.18.1\n",
      "numpy 1.12.0\n",
      "\n",
      "compiler   : GCC 4.2.1 (Based on Apple Inc. build 5658) (LLVM build 2336.11.00)\n",
      "system     : Darwin\n",
      "release    : 16.4.0\n",
      "machine    : x86_64\n",
      "processor  : i386\n",
      "CPU cores  : 8\n",
      "interpreter: 64bit\n"
     ]
    }
   ],
   "source": [
    "%pylab\n",
    "%matplotlib inline\n",
    "%load_ext watermark\n",
    "%watermark -v -m -p scipy,sklearn,numpy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas\n",
    "import math\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import LSTM\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "# fix random seed for reproducibility\n",
    "numpy.random.seed(7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "before normalization :  [[ 112.]\n",
      " [ 118.]\n",
      " [ 132.]\n",
      " [ 129.]\n",
      " [ 121.]]\n",
      "after normalization :  [[ 0.01544401]\n",
      " [ 0.02702703]\n",
      " [ 0.05405405]\n",
      " [ 0.04826255]\n",
      " [ 0.03281853]]\n"
     ]
    }
   ],
   "source": [
    "# load the dataset\n",
    "dataframe = pandas.read_csv('international-airline-passengers.csv', usecols=[1], engine='python', skipfooter=3)\n",
    "dataset = dataframe.values\n",
    "dataset = dataset.astype('float32')\n",
    "\n",
    "print \"before normalization : \", dataset[:5]\n",
    "\n",
    "# normalize the dataset\n",
    "scaler = MinMaxScaler(feature_range=(0, 1))\n",
    "dataset = scaler.fit_transform(dataset)\n",
    "\n",
    "print \"after normalization : \", dataset[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(96, 48)\n"
     ]
    }
   ],
   "source": [
    "# split into train and test sets\n",
    "train_size = int(len(dataset) * 0.67)\n",
    "test_size = len(dataset) - train_size\n",
    "train, test = dataset[0:train_size,:], dataset[train_size:len(dataset),:]\n",
    "print(len(train), len(test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# convert an array of values into a dataset matrix\n",
    "def create_dataset(dataset, look_back=1):\n",
    "    dataX, dataY = [], []\n",
    "    for i in range(len(dataset)-look_back-1):\n",
    "        a = dataset[i:(i+look_back), 0]\n",
    "        dataX.append(a)\n",
    "        dataY.append(dataset[i + look_back, 0])\n",
    "    return numpy.array(dataX), numpy.array(dataY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " (94, 1) (46, 1)\n",
      "(94, 1, 1) (46, 1, 1)\n"
     ]
    }
   ],
   "source": [
    "# reshape into X=t and Y=t+1\n",
    "look_back = 1\n",
    "trainX, trainY = create_dataset(train, look_back)\n",
    "testX, testY = create_dataset(test, look_back)\n",
    "\n",
    "print trainX.shape, testX.shape\n",
    "\n",
    "# reshape input to be [samples, time steps, features]\n",
    "trainX = numpy.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))\n",
    "testX = numpy.reshape(testX, (testX.shape[0], 1, testX.shape[1]))\n",
    "\n",
    "print trainX.shape, testX.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "1s - loss: 0.0321\n",
      "Epoch 2/100\n",
      "0s - loss: 0.0147\n",
      "Epoch 3/100\n",
      "0s - loss: 0.0116\n",
      "Epoch 4/100\n",
      "0s - loss: 0.0106\n",
      "Epoch 5/100\n",
      "0s - loss: 0.0097\n",
      "Epoch 6/100\n",
      "0s - loss: 0.0089\n",
      "Epoch 7/100\n",
      "0s - loss: 0.0082\n",
      "Epoch 8/100\n",
      "0s - loss: 0.0074\n",
      "Epoch 9/100\n",
      "0s - loss: 0.0065\n",
      "Epoch 10/100\n",
      "0s - loss: 0.0058\n",
      "Epoch 11/100\n",
      "0s - loss: 0.0052\n",
      "Epoch 12/100\n",
      "0s - loss: 0.0045\n",
      "Epoch 13/100\n",
      "0s - loss: 0.0040\n",
      "Epoch 14/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 15/100\n",
      "0s - loss: 0.0031\n",
      "Epoch 16/100\n",
      "0s - loss: 0.0028\n",
      "Epoch 17/100\n",
      "0s - loss: 0.0026\n",
      "Epoch 18/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 19/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 20/100\n",
      "0s - loss: 0.0022\n",
      "Epoch 21/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 22/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 23/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 24/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 25/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 26/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 27/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 28/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 29/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 30/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 31/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 32/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 33/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 34/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 35/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 36/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 37/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 38/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 39/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 40/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 41/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 42/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 43/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 44/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 45/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 46/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 47/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 48/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 49/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 50/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 51/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 52/100\n",
      "0s - loss: 0.0019\n",
      "Epoch 53/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 54/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 55/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 56/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 57/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 58/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 59/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 60/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 61/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 62/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 63/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 64/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 65/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 66/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 67/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 68/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 69/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 70/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 71/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 72/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 73/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 74/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 75/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 76/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 77/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 78/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 79/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 80/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 81/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 82/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 83/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 84/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 85/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 86/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 87/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 88/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 89/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 90/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 91/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 92/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 93/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 94/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 95/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 96/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 97/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 98/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 99/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 100/100\n",
      "0s - loss: 0.0020\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x11fc3bc50>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create and fit the LSTM network\n",
    "model = Sequential()\n",
    "model.add(LSTM(4, input_dim=look_back))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "model.fit(trainX, trainY, nb_epoch=100, batch_size=1, verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Score: 23.57 RMSE\n",
      "Test Score: 51.02 RMSE\n"
     ]
    }
   ],
   "source": [
    "# make predictions\n",
    "trainPredict = model.predict(trainX)\n",
    "testPredict = model.predict(testX)\n",
    "\n",
    "# invert predictions\n",
    "trainPredict = scaler.inverse_transform(trainPredict)\n",
    "trainY = scaler.inverse_transform([trainY])\n",
    "testPredict = scaler.inverse_transform(testPredict)\n",
    "testY = scaler.inverse_transform([testY])\n",
    "\n",
    "# calculate root mean squared error\n",
    "trainScore = math.sqrt(mean_squared_error(trainY[0], trainPredict[:,0]))\n",
    "print('Train Score: %.2f RMSE' % (trainScore))\n",
    "testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:,0]))\n",
    "print('Test Score: %.2f RMSE' % (testScore))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x120731610>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# shift train predictions for plotting\n",
    "trainPredictPlot = numpy.empty_like(dataset)\n",
    "trainPredictPlot[:, :] = numpy.nan\n",
    "trainPredictPlot[look_back:len(trainPredict)+look_back, :] = trainPredict\n",
    "\n",
    "# shift test predictions for plotting\n",
    "testPredictPlot = numpy.empty_like(dataset)\n",
    "testPredictPlot[:, :] = numpy.nan\n",
    "testPredictPlot[len(trainPredict)+(look_back*2)+1:len(dataset)-1, :] = testPredict\n",
    "\n",
    "# plot baseline and predictions\n",
    "plt.plot(scaler.inverse_transform(dataset))\n",
    "plt.plot(trainPredictPlot)\n",
    "plt.plot(testPredictPlot)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(92, 3) (44, 3)\n",
      "(92, 1, 3) (44, 1, 3)\n"
     ]
    }
   ],
   "source": [
    "# reshape into X=t and Y=t+1\n",
    "look_back = 3\n",
    "trainX, trainY = create_dataset(train, look_back)\n",
    "testX, testY = create_dataset(test, look_back)\n",
    "\n",
    "print trainX.shape, testX.shape\n",
    "\n",
    "# reshape input to be [samples, time steps, features]\n",
    "trainX = numpy.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))\n",
    "testX = numpy.reshape(testX, (testX.shape[0], 1, testX.shape[1]))\n",
    "\n",
    "print trainX.shape, testX.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "1s - loss: 0.0175\n",
      "Epoch 2/100\n",
      "0s - loss: 0.0084\n",
      "Epoch 3/100\n",
      "0s - loss: 0.0078\n",
      "Epoch 4/100\n",
      "0s - loss: 0.0071\n",
      "Epoch 5/100\n",
      "0s - loss: 0.0067\n",
      "Epoch 6/100\n",
      "0s - loss: 0.0064\n",
      "Epoch 7/100\n",
      "0s - loss: 0.0060\n",
      "Epoch 8/100\n",
      "0s - loss: 0.0058\n",
      "Epoch 9/100\n",
      "0s - loss: 0.0056\n",
      "Epoch 10/100\n",
      "0s - loss: 0.0054\n",
      "Epoch 11/100\n",
      "0s - loss: 0.0053\n",
      "Epoch 12/100\n",
      "0s - loss: 0.0051\n",
      "Epoch 13/100\n",
      "0s - loss: 0.0051\n",
      "Epoch 14/100\n",
      "0s - loss: 0.0050\n",
      "Epoch 15/100\n",
      "0s - loss: 0.0048\n",
      "Epoch 16/100\n",
      "0s - loss: 0.0046\n",
      "Epoch 17/100\n",
      "0s - loss: 0.0050\n",
      "Epoch 18/100\n",
      "0s - loss: 0.0047\n",
      "Epoch 19/100\n",
      "0s - loss: 0.0044\n",
      "Epoch 20/100\n",
      "0s - loss: 0.0046\n",
      "Epoch 21/100\n",
      "0s - loss: 0.0043\n",
      "Epoch 22/100\n",
      "0s - loss: 0.0043\n",
      "Epoch 23/100\n",
      "0s - loss: 0.0044\n",
      "Epoch 24/100\n",
      "0s - loss: 0.0043\n",
      "Epoch 25/100\n",
      "0s - loss: 0.0041\n",
      "Epoch 26/100\n",
      "0s - loss: 0.0040\n",
      "Epoch 27/100\n",
      "0s - loss: 0.0039\n",
      "Epoch 28/100\n",
      "0s - loss: 0.0039\n",
      "Epoch 29/100\n",
      "0s - loss: 0.0039\n",
      "Epoch 30/100\n",
      "0s - loss: 0.0038\n",
      "Epoch 31/100\n",
      "0s - loss: 0.0038\n",
      "Epoch 32/100\n",
      "0s - loss: 0.0037\n",
      "Epoch 33/100\n",
      "0s - loss: 0.0037\n",
      "Epoch 34/100\n",
      "0s - loss: 0.0036\n",
      "Epoch 35/100\n",
      "0s - loss: 0.0037\n",
      "Epoch 36/100\n",
      "0s - loss: 0.0035\n",
      "Epoch 37/100\n",
      "0s - loss: 0.0035\n",
      "Epoch 38/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 39/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 40/100\n",
      "0s - loss: 0.0033\n",
      "Epoch 41/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 42/100\n",
      "0s - loss: 0.0033\n",
      "Epoch 43/100\n",
      "0s - loss: 0.0032\n",
      "Epoch 44/100\n",
      "0s - loss: 0.0033\n",
      "Epoch 45/100\n",
      "0s - loss: 0.0031\n",
      "Epoch 46/100\n",
      "0s - loss: 0.0032\n",
      "Epoch 47/100\n",
      "0s - loss: 0.0032\n",
      "Epoch 48/100\n",
      "0s - loss: 0.0031\n",
      "Epoch 49/100\n",
      "0s - loss: 0.0030\n",
      "Epoch 50/100\n",
      "0s - loss: 0.0030\n",
      "Epoch 51/100\n",
      "0s - loss: 0.0030\n",
      "Epoch 52/100\n",
      "0s - loss: 0.0030\n",
      "Epoch 53/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 54/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 55/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 56/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 57/100\n",
      "0s - loss: 0.0028\n",
      "Epoch 58/100\n",
      "0s - loss: 0.0028\n",
      "Epoch 59/100\n",
      "0s - loss: 0.0027\n",
      "Epoch 60/100\n",
      "0s - loss: 0.0028\n",
      "Epoch 61/100\n",
      "0s - loss: 0.0028\n",
      "Epoch 62/100\n",
      "0s - loss: 0.0026\n",
      "Epoch 63/100\n",
      "0s - loss: 0.0027\n",
      "Epoch 64/100\n",
      "0s - loss: 0.0026\n",
      "Epoch 65/100\n",
      "0s - loss: 0.0025\n",
      "Epoch 66/100\n",
      "0s - loss: 0.0026\n",
      "Epoch 67/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 68/100\n",
      "0s - loss: 0.0025\n",
      "Epoch 69/100\n",
      "0s - loss: 0.0025\n",
      "Epoch 70/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 71/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 72/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 73/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 74/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 75/100\n",
      "0s - loss: 0.0022\n",
      "Epoch 76/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 77/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 78/100\n",
      "0s - loss: 0.0022\n",
      "Epoch 79/100\n",
      "0s - loss: 0.0022\n",
      "Epoch 80/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 81/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 82/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 83/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 84/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 85/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 86/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 87/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 88/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 89/100\n",
      "0s - loss: 0.0019\n",
      "Epoch 90/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 91/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 92/100\n",
      "0s - loss: 0.0019\n",
      "Epoch 93/100\n",
      "0s - loss: 0.0019\n",
      "Epoch 94/100\n",
      "0s - loss: 0.0019\n",
      "Epoch 95/100\n",
      "0s - loss: 0.0019\n",
      "Epoch 96/100\n",
      "0s - loss: 0.0019\n",
      "Epoch 97/100\n",
      "0s - loss: 0.0019\n",
      "Epoch 98/100\n",
      "0s - loss: 0.0018\n",
      "Epoch 99/100\n",
      "0s - loss: 0.0019\n",
      "Epoch 100/100\n",
      "0s - loss: 0.0018\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x12041ec10>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create and fit the LSTM network\n",
    "model = Sequential()\n",
    "model.add(LSTM(4, input_dim=look_back))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "model.fit(trainX, trainY, nb_epoch=100, batch_size=1, verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Score: 21.72 RMSE\n",
      "Test Score: 55.14 RMSE\n"
     ]
    }
   ],
   "source": [
    "# make predictions\n",
    "trainPredict = model.predict(trainX)\n",
    "testPredict = model.predict(testX)\n",
    "\n",
    "# invert predictions\n",
    "trainPredict = scaler.inverse_transform(trainPredict)\n",
    "trainY = scaler.inverse_transform([trainY])\n",
    "testPredict = scaler.inverse_transform(testPredict)\n",
    "testY = scaler.inverse_transform([testY])\n",
    "\n",
    "# calculate root mean squared error\n",
    "trainScore = math.sqrt(mean_squared_error(trainY[0], trainPredict[:,0]))\n",
    "print('Train Score: %.2f RMSE' % (trainScore))\n",
    "testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:,0]))\n",
    "print('Test Score: %.2f RMSE' % (testScore))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x122b2ac50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# shift train predictions for plotting\n",
    "trainPredictPlot = numpy.empty_like(dataset)\n",
    "trainPredictPlot[:, :] = numpy.nan\n",
    "trainPredictPlot[look_back:len(trainPredict)+look_back, :] = trainPredict\n",
    "\n",
    "# shift test predictions for plotting\n",
    "testPredictPlot = numpy.empty_like(dataset)\n",
    "testPredictPlot[:, :] = numpy.nan\n",
    "testPredictPlot[len(trainPredict)+(look_back*2)+1:len(dataset)-1, :] = testPredict\n",
    "\n",
    "# plot baseline and predictions\n",
    "plt.plot(scaler.inverse_transform(dataset))\n",
    "plt.plot(trainPredictPlot)\n",
    "plt.plot(testPredictPlot)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(92, 3) (44, 3)\n",
      "(92, 3, 1) (44, 3, 1)\n"
     ]
    }
   ],
   "source": [
    "# reshape into X=t and Y=t+1\n",
    "look_back = 3\n",
    "trainX, trainY = create_dataset(train, look_back)\n",
    "testX, testY = create_dataset(test, look_back)\n",
    "\n",
    "print trainX.shape, testX.shape\n",
    "\n",
    "# reshape input to be [samples, time steps, features]\n",
    "trainX = numpy.reshape(trainX, (trainX.shape[0], trainX.shape[1], 1))\n",
    "testX = numpy.reshape(testX, (testX.shape[0], testX.shape[1], 1))\n",
    "\n",
    "print trainX.shape, testX.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "2s - loss: 0.0170\n",
      "Epoch 2/100\n",
      "0s - loss: 0.0078\n",
      "Epoch 3/100\n",
      "0s - loss: 0.0058\n",
      "Epoch 4/100\n",
      "0s - loss: 0.0048\n",
      "Epoch 5/100\n",
      "0s - loss: 0.0041\n",
      "Epoch 6/100\n",
      "0s - loss: 0.0039\n",
      "Epoch 7/100\n",
      "0s - loss: 0.0037\n",
      "Epoch 8/100\n",
      "0s - loss: 0.0036\n",
      "Epoch 9/100\n",
      "0s - loss: 0.0036\n",
      "Epoch 10/100\n",
      "0s - loss: 0.0035\n",
      "Epoch 11/100\n",
      "0s - loss: 0.0035\n",
      "Epoch 12/100\n",
      "0s - loss: 0.0036\n",
      "Epoch 13/100\n",
      "0s - loss: 0.0036\n",
      "Epoch 14/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 15/100\n",
      "0s - loss: 0.0035\n",
      "Epoch 16/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 17/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 18/100\n",
      "0s - loss: 0.0035\n",
      "Epoch 19/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 20/100\n",
      "0s - loss: 0.0033\n",
      "Epoch 21/100\n",
      "0s - loss: 0.0036\n",
      "Epoch 22/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 23/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 24/100\n",
      "0s - loss: 0.0033\n",
      "Epoch 25/100\n",
      "0s - loss: 0.0034\n",
      "Epoch 26/100\n",
      "0s - loss: 0.0033\n",
      "Epoch 27/100\n",
      "0s - loss: 0.0033\n",
      "Epoch 28/100\n",
      "0s - loss: 0.0033\n",
      "Epoch 29/100\n",
      "0s - loss: 0.0033\n",
      "Epoch 30/100\n",
      "0s - loss: 0.0032\n",
      "Epoch 31/100\n",
      "0s - loss: 0.0033\n",
      "Epoch 32/100\n",
      "0s - loss: 0.0032\n",
      "Epoch 33/100\n",
      "0s - loss: 0.0032\n",
      "Epoch 34/100\n",
      "0s - loss: 0.0032\n",
      "Epoch 35/100\n",
      "0s - loss: 0.0032\n",
      "Epoch 36/100\n",
      "0s - loss: 0.0031\n",
      "Epoch 37/100\n",
      "0s - loss: 0.0031\n",
      "Epoch 38/100\n",
      "0s - loss: 0.0031\n",
      "Epoch 39/100\n",
      "0s - loss: 0.0032\n",
      "Epoch 40/100\n",
      "0s - loss: 0.0030\n",
      "Epoch 41/100\n",
      "0s - loss: 0.0031\n",
      "Epoch 42/100\n",
      "0s - loss: 0.0030\n",
      "Epoch 43/100\n",
      "0s - loss: 0.0031\n",
      "Epoch 44/100\n",
      "0s - loss: 0.0030\n",
      "Epoch 45/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 46/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 47/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 48/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 49/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 50/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 51/100\n",
      "0s - loss: 0.0029\n",
      "Epoch 52/100\n",
      "0s - loss: 0.0030\n",
      "Epoch 53/100\n",
      "1s - loss: 0.0029\n",
      "Epoch 54/100\n",
      "1s - loss: 0.0028\n",
      "Epoch 55/100\n",
      "0s - loss: 0.0030\n",
      "Epoch 56/100\n",
      "0s - loss: 0.0027\n",
      "Epoch 57/100\n",
      "0s - loss: 0.0030\n",
      "Epoch 58/100\n",
      "0s - loss: 0.0027\n",
      "Epoch 59/100\n",
      "0s - loss: 0.0027\n",
      "Epoch 60/100\n",
      "0s - loss: 0.0028\n",
      "Epoch 61/100\n",
      "0s - loss: 0.0028\n",
      "Epoch 62/100\n",
      "1s - loss: 0.0026\n",
      "Epoch 63/100\n",
      "0s - loss: 0.0026\n",
      "Epoch 64/100\n",
      "0s - loss: 0.0027\n",
      "Epoch 65/100\n",
      "0s - loss: 0.0025\n",
      "Epoch 66/100\n",
      "0s - loss: 0.0026\n",
      "Epoch 67/100\n",
      "0s - loss: 0.0027\n",
      "Epoch 68/100\n",
      "0s - loss: 0.0026\n",
      "Epoch 69/100\n",
      "0s - loss: 0.0025\n",
      "Epoch 70/100\n",
      "0s - loss: 0.0026\n",
      "Epoch 71/100\n",
      "0s - loss: 0.0025\n",
      "Epoch 72/100\n",
      "0s - loss: 0.0025\n",
      "Epoch 73/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 74/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 75/100\n",
      "0s - loss: 0.0026\n",
      "Epoch 76/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 77/100\n",
      "0s - loss: 0.0025\n",
      "Epoch 78/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 79/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 80/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 81/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 82/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 83/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 84/100\n",
      "0s - loss: 0.0022\n",
      "Epoch 85/100\n",
      "0s - loss: 0.0024\n",
      "Epoch 86/100\n",
      "0s - loss: 0.0022\n",
      "Epoch 87/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 88/100\n",
      "0s - loss: 0.0022\n",
      "Epoch 89/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 90/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 91/100\n",
      "0s - loss: 0.0023\n",
      "Epoch 92/100\n",
      "0s - loss: 0.0022\n",
      "Epoch 93/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 94/100\n",
      "0s - loss: 0.0022\n",
      "Epoch 95/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 96/100\n",
      "0s - loss: 0.0022\n",
      "Epoch 97/100\n",
      "0s - loss: 0.0021\n",
      "Epoch 98/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 99/100\n",
      "0s - loss: 0.0020\n",
      "Epoch 100/100\n",
      "0s - loss: 0.0020\n",
      "Train Score: 22.93 RMSE\n",
      "Test Score: 60.27 RMSE\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIP4GycyyTMBc8wkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x124ef7990>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# create and fit the LSTM network\n",
    "model = Sequential()\n",
    "model.add(LSTM(4, input_dim=1))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "model.fit(trainX, trainY, nb_epoch=100, batch_size=1, verbose=2)\n",
    "\n",
    "# make predictions\n",
    "trainPredict = model.predict(trainX)\n",
    "testPredict = model.predict(testX)\n",
    "\n",
    "# invert predictions\n",
    "trainPredict = scaler.inverse_transform(trainPredict)\n",
    "trainY = scaler.inverse_transform([trainY])\n",
    "testPredict = scaler.inverse_transform(testPredict)\n",
    "testY = scaler.inverse_transform([testY])\n",
    "\n",
    "# calculate root mean squared error\n",
    "trainScore = math.sqrt(mean_squared_error(trainY[0], trainPredict[:,0]))\n",
    "print('Train Score: %.2f RMSE' % (trainScore))\n",
    "testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:,0]))\n",
    "print('Test Score: %.2f RMSE' % (testScore))\n",
    "\n",
    "# shift train predictions for plotting\n",
    "trainPredictPlot = numpy.empty_like(dataset)\n",
    "trainPredictPlot[:, :] = numpy.nan\n",
    "trainPredictPlot[look_back:len(trainPredict)+look_back, :] = trainPredict\n",
    "\n",
    "# shift test predictions for plotting\n",
    "testPredictPlot = numpy.empty_like(dataset)\n",
    "testPredictPlot[:, :] = numpy.nan\n",
    "testPredictPlot[len(trainPredict)+(look_back*2)+1:len(dataset)-1, :] = testPredict\n",
    "\n",
    "# plot baseline and predictions\n",
    "plt.plot(scaler.inverse_transform(dataset))\n",
    "plt.plot(trainPredictPlot)\n",
    "plt.plot(testPredictPlot)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "2s - loss: 0.0071\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0098\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0066\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0055\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0049\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0045\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0043\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0042\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0041\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0040\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0039\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0039\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0038\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0038\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0037\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0037\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0036\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0036\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0036\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0035\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0035\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0035\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0034\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0034\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0034\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0034\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0033\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0033\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0033\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0033\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0032\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0032\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0032\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0032\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0031\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0031\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0031\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0031\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0031\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0030\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0030\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0030\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0030\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0029\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0029\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0029\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0029\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0029\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0028\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0028\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0028\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0028\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0027\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0027\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0027\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0027\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0026\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0026\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0026\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0026\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0025\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0025\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0025\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0025\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0024\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0024\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0024\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0024\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0024\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0023\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0023\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0023\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0023\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0023\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0023\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0022\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0022\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0022\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0022\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0022\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0022\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0021\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0021\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0021\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0021\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0021\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0021\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0019\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0019\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0019\n",
      "Epoch 1/1\n",
      "0s - loss: 0.0019\n",
      "Train Score: 22.61 RMSE\n",
      "Test Score: 53.22 RMSE\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1260d8cd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# reshape into X=t and Y=t+1\n",
    "look_back = 3\n",
    "trainX, trainY = create_dataset(train, look_back)\n",
    "testX, testY = create_dataset(test, look_back)\n",
    "\n",
    "# reshape input to be [samples, time steps, features]\n",
    "trainX = numpy.reshape(trainX, (trainX.shape[0], trainX.shape[1], 1))\n",
    "testX = numpy.reshape(testX, (testX.shape[0], testX.shape[1], 1))\n",
    "\n",
    "# create and fit the LSTM network\n",
    "batch_size = 1\n",
    "model = Sequential()\n",
    "model.add(LSTM(4, batch_input_shape=(batch_size, look_back, 1), stateful=True))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "\n",
    "for i in range(100):\n",
    "    model.fit(trainX, trainY, nb_epoch=1, batch_size=batch_size, verbose=2, shuffle=False)\n",
    "    model.reset_states()\n",
    "\n",
    "# make predictions\n",
    "trainPredict = model.predict(trainX, batch_size=batch_size)\n",
    "model.reset_states()\n",
    "testPredict = model.predict(testX, batch_size=batch_size)\n",
    "\n",
    "# invert predictions\n",
    "trainPredict = scaler.inverse_transform(trainPredict)\n",
    "trainY = scaler.inverse_transform([trainY])\n",
    "testPredict = scaler.inverse_transform(testPredict)\n",
    "testY = scaler.inverse_transform([testY])\n",
    "\n",
    "# calculate root mean squared error\n",
    "trainScore = math.sqrt(mean_squared_error(trainY[0], trainPredict[:,0]))\n",
    "print('Train Score: %.2f RMSE' % (trainScore))\n",
    "testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:,0]))\n",
    "print('Test Score: %.2f RMSE' % (testScore))\n",
    "\n",
    "# shift train predictions for plotting\n",
    "trainPredictPlot = numpy.empty_like(dataset)\n",
    "trainPredictPlot[:, :] = numpy.nan\n",
    "trainPredictPlot[look_back:len(trainPredict)+look_back, :] = trainPredict\n",
    "\n",
    "# shift test predictions for plotting\n",
    "testPredictPlot = numpy.empty_like(dataset)\n",
    "testPredictPlot[:, :] = numpy.nan\n",
    "testPredictPlot[len(trainPredict)+(look_back*2)+1:len(dataset)-1, :] = testPredict\n",
    "\n",
    "# plot baseline and predictions\n",
    "plt.plot(scaler.inverse_transform(dataset))\n",
    "plt.plot(trainPredictPlot)\n",
    "plt.plot(testPredictPlot)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "3s - loss: 0.0117\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0199\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1s - loss: 0.0016\n",
      "Train Score: 21.36 RMSE\n",
      "Test Score: 71.92 RMSE\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4PvkzgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x12b417550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# reshape into X=t and Y=t+1\n",
    "look_back = 3\n",
    "trainX, trainY = create_dataset(train, look_back)\n",
    "testX, testY = create_dataset(test, look_back)\n",
    "\n",
    "# reshape input to be [samples, time steps, features]\n",
    "trainX = numpy.reshape(trainX, (trainX.shape[0], trainX.shape[1], 1))\n",
    "testX = numpy.reshape(testX, (testX.shape[0], testX.shape[1], 1))\n",
    "\n",
    "# create and fit the LSTM network\n",
    "batch_size = 1\n",
    "model = Sequential()\n",
    "model.add(LSTM(4, batch_input_shape=(batch_size, look_back, 1), stateful=True, return_sequences=True))\n",
    "model.add(LSTM(4, batch_input_shape=(batch_size, look_back, 1), stateful=True))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "\n",
    "for i in range(100):\n",
    "    model.fit(trainX, trainY, nb_epoch=1, batch_size=batch_size, verbose=2, shuffle=False)\n",
    "    model.reset_states()\n",
    "\n",
    "# make predictions\n",
    "trainPredict = model.predict(trainX, batch_size=batch_size)\n",
    "model.reset_states()\n",
    "testPredict = model.predict(testX, batch_size=batch_size)\n",
    "\n",
    "# invert predictions\n",
    "trainPredict = scaler.inverse_transform(trainPredict)\n",
    "trainY = scaler.inverse_transform([trainY])\n",
    "testPredict = scaler.inverse_transform(testPredict)\n",
    "testY = scaler.inverse_transform([testY])\n",
    "\n",
    "# calculate root mean squared error\n",
    "trainScore = math.sqrt(mean_squared_error(trainY[0], trainPredict[:,0]))\n",
    "print('Train Score: %.2f RMSE' % (trainScore))\n",
    "testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:,0]))\n",
    "print('Test Score: %.2f RMSE' % (testScore))\n",
    "\n",
    "# shift train predictions for plotting\n",
    "trainPredictPlot = numpy.empty_like(dataset)\n",
    "trainPredictPlot[:, :] = numpy.nan\n",
    "trainPredictPlot[look_back:len(trainPredict)+look_back, :] = trainPredict\n",
    "\n",
    "# shift test predictions for plotting\n",
    "testPredictPlot = numpy.empty_like(dataset)\n",
    "testPredictPlot[:, :] = numpy.nan\n",
    "testPredictPlot[len(trainPredict)+(look_back*2)+1:len(dataset)-1, :] = testPredict\n",
    "\n",
    "# plot baseline and predictions\n",
    "plt.plot(scaler.inverse_transform(dataset))\n",
    "plt.plot(trainPredictPlot)\n",
    "plt.plot(testPredictPlot)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}