notebook/20_keras_time_series_prediction.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": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0XccxykhXPQdx3FKCBd9x3GcEuL/A6xFLGDd40wIAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1116cfb90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "dataset = pandas.read_csv('international-airline-passengers.csv', usecols=[1], engine='python', skipfooter=3)\n",
    "plt.plot(dataset)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import numpy\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "\n",
    "# fix random seed for reproducibility\n",
    "numpy.random.seed(7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "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')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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": 6,
   "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": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 112.]\n",
      " [ 118.]\n",
      " [ 132.]\n",
      " [ 129.]\n",
      " [ 121.]]\n",
      "[ 118.  132.  129.  121.  135.]\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[:5]\n",
    "print trainY[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "0s - loss: 46211.7889\n",
      "Epoch 2/200\n",
      "0s - loss: 32824.0573\n",
      "Epoch 3/200\n",
      "0s - loss: 22463.3847\n",
      "Epoch 4/200\n",
      "0s - loss: 14574.2490\n",
      "Epoch 5/200\n",
      "0s - loss: 9015.9301\n",
      "Epoch 6/200\n",
      "0s - loss: 5305.5743\n",
      "Epoch 7/200\n",
      "0s - loss: 3040.9822\n",
      "Epoch 8/200\n",
      "0s - loss: 1736.5572\n",
      "Epoch 9/200\n",
      "0s - loss: 1064.1941\n",
      "Epoch 10/200\n",
      "0s - loss: 751.0096\n",
      "Epoch 11/200\n",
      "0s - loss: 613.6888\n",
      "Epoch 12/200\n",
      "0s - loss: 564.4179\n",
      "Epoch 13/200\n",
      "0s - loss: 543.9144\n",
      "Epoch 14/200\n",
      "0s - loss: 539.3391\n",
      "Epoch 15/200\n",
      "0s - loss: 538.5246\n",
      "Epoch 16/200\n",
      "0s - loss: 537.5370\n",
      "Epoch 17/200\n",
      "0s - loss: 546.8038\n",
      "Epoch 18/200\n",
      "0s - loss: 548.7347\n",
      "Epoch 19/200\n",
      "0s - loss: 539.2023\n",
      "Epoch 20/200\n",
      "0s - loss: 539.8034\n",
      "Epoch 21/200\n",
      "0s - loss: 538.0942\n",
      "Epoch 22/200\n",
      "0s - loss: 537.0314\n",
      "Epoch 23/200\n",
      "0s - loss: 538.4210\n",
      "Epoch 24/200\n",
      "0s - loss: 536.7596\n",
      "Epoch 25/200\n",
      "0s - loss: 539.7120\n",
      "Epoch 26/200\n",
      "0s - loss: 540.2790\n",
      "Epoch 27/200\n",
      "0s - loss: 539.8297\n",
      "Epoch 28/200\n",
      "0s - loss: 538.8166\n",
      "Epoch 29/200\n",
      "0s - loss: 539.6627\n",
      "Epoch 30/200\n",
      "0s - loss: 541.0410\n",
      "Epoch 31/200\n",
      "0s - loss: 537.3542\n",
      "Epoch 32/200\n",
      "0s - loss: 540.8072\n",
      "Epoch 33/200\n",
      "0s - loss: 537.9920\n",
      "Epoch 34/200\n",
      "0s - loss: 541.2672\n",
      "Epoch 35/200\n",
      "0s - loss: 538.8388\n",
      "Epoch 36/200\n",
      "0s - loss: 540.4268\n",
      "Epoch 37/200\n",
      "0s - loss: 539.7244\n",
      "Epoch 38/200\n",
      "0s - loss: 543.6918\n",
      "Epoch 39/200\n",
      "0s - loss: 539.9526\n",
      "Epoch 40/200\n",
      "0s - loss: 537.4356\n",
      "Epoch 41/200\n",
      "0s - loss: 538.2532\n",
      "Epoch 42/200\n",
      "0s - loss: 539.5876\n",
      "Epoch 43/200\n",
      "0s - loss: 545.5434\n",
      "Epoch 44/200\n",
      "0s - loss: 540.3093\n",
      "Epoch 45/200\n",
      "0s - loss: 539.1076\n",
      "Epoch 46/200\n",
      "0s - loss: 544.1970\n",
      "Epoch 47/200\n",
      "0s - loss: 543.6950\n",
      "Epoch 48/200\n",
      "0s - loss: 543.1104\n",
      "Epoch 49/200\n",
      "0s - loss: 536.5909\n",
      "Epoch 50/200\n",
      "0s - loss: 542.2569\n",
      "Epoch 51/200\n",
      "0s - loss: 539.9582\n",
      "Epoch 52/200\n",
      "0s - loss: 536.2569\n",
      "Epoch 53/200\n",
      "0s - loss: 538.2601\n",
      "Epoch 54/200\n",
      "0s - loss: 545.8824\n",
      "Epoch 55/200\n",
      "0s - loss: 538.7492\n",
      "Epoch 56/200\n",
      "0s - loss: 543.5097\n",
      "Epoch 57/200\n",
      "0s - loss: 552.9656\n",
      "Epoch 58/200\n",
      "0s - loss: 547.6196\n",
      "Epoch 59/200\n",
      "0s - loss: 542.5405\n",
      "Epoch 60/200\n",
      "0s - loss: 543.7982\n",
      "Epoch 61/200\n",
      "0s - loss: 542.9588\n",
      "Epoch 62/200\n",
      "0s - loss: 547.6986\n",
      "Epoch 63/200\n",
      "0s - loss: 539.4204\n",
      "Epoch 64/200\n",
      "0s - loss: 542.9874\n",
      "Epoch 65/200\n",
      "0s - loss: 540.6029\n",
      "Epoch 66/200\n",
      "0s - loss: 541.2689\n",
      "Epoch 67/200\n",
      "0s - loss: 551.8631\n",
      "Epoch 68/200\n",
      "0s - loss: 565.4273\n",
      "Epoch 69/200\n",
      "0s - loss: 540.0199\n",
      "Epoch 70/200\n",
      "0s - loss: 539.0941\n",
      "Epoch 71/200\n",
      "0s - loss: 546.5400\n",
      "Epoch 72/200\n",
      "0s - loss: 543.9756\n",
      "Epoch 73/200\n",
      "0s - loss: 541.5063\n",
      "Epoch 74/200\n",
      "0s - loss: 542.0506\n",
      "Epoch 75/200\n",
      "0s - loss: 544.3379\n",
      "Epoch 76/200\n",
      "0s - loss: 538.3940\n",
      "Epoch 77/200\n",
      "0s - loss: 543.0666\n",
      "Epoch 78/200\n",
      "0s - loss: 536.6899\n",
      "Epoch 79/200\n",
      "0s - loss: 544.3609\n",
      "Epoch 80/200\n",
      "0s - loss: 543.8001\n",
      "Epoch 81/200\n",
      "0s - loss: 539.4327\n",
      "Epoch 82/200\n",
      "0s - loss: 538.1242\n",
      "Epoch 83/200\n",
      "0s - loss: 554.9938\n",
      "Epoch 84/200\n",
      "0s - loss: 549.5590\n",
      "Epoch 85/200\n",
      "0s - loss: 540.3159\n",
      "Epoch 86/200\n",
      "0s - loss: 540.8831\n",
      "Epoch 87/200\n",
      "0s - loss: 539.2654\n",
      "Epoch 88/200\n",
      "0s - loss: 541.7687\n",
      "Epoch 89/200\n",
      "0s - loss: 543.3090\n",
      "Epoch 90/200\n",
      "0s - loss: 553.6016\n",
      "Epoch 91/200\n",
      "0s - loss: 545.0680\n",
      "Epoch 92/200\n",
      "0s - loss: 544.7325\n",
      "Epoch 93/200\n",
      "0s - loss: 539.3530\n",
      "Epoch 94/200\n",
      "0s - loss: 538.8370\n",
      "Epoch 95/200\n",
      "0s - loss: 547.2192\n",
      "Epoch 96/200\n",
      "0s - loss: 540.5375\n",
      "Epoch 97/200\n",
      "0s - loss: 542.9520\n",
      "Epoch 98/200\n",
      "0s - loss: 539.0761\n",
      "Epoch 99/200\n",
      "0s - loss: 545.9798\n",
      "Epoch 100/200\n",
      "0s - loss: 540.5965\n",
      "Epoch 101/200\n",
      "0s - loss: 540.4004\n",
      "Epoch 102/200\n",
      "0s - loss: 538.0091\n",
      "Epoch 103/200\n",
      "0s - loss: 544.8114\n",
      "Epoch 104/200\n",
      "0s - loss: 536.8192\n",
      "Epoch 105/200\n",
      "0s - loss: 546.0621\n",
      "Epoch 106/200\n",
      "0s - loss: 544.7274\n",
      "Epoch 107/200\n",
      "0s - loss: 540.7268\n",
      "Epoch 108/200\n",
      "0s - loss: 545.3241\n",
      "Epoch 109/200\n",
      "0s - loss: 538.6303\n",
      "Epoch 110/200\n",
      "0s - loss: 545.2612\n",
      "Epoch 111/200\n",
      "0s - loss: 540.7357\n",
      "Epoch 112/200\n",
      "0s - loss: 544.0329\n",
      "Epoch 113/200\n",
      "0s - loss: 541.7410\n",
      "Epoch 114/200\n",
      "0s - loss: 535.5473\n",
      "Epoch 115/200\n",
      "0s - loss: 549.4917\n",
      "Epoch 116/200\n",
      "0s - loss: 542.3723\n",
      "Epoch 117/200\n",
      "0s - loss: 542.5242\n",
      "Epoch 118/200\n",
      "0s - loss: 541.9145\n",
      "Epoch 119/200\n",
      "0s - loss: 542.2058\n",
      "Epoch 120/200\n",
      "0s - loss: 541.8279\n",
      "Epoch 121/200\n",
      "0s - loss: 543.7890\n",
      "Epoch 122/200\n",
      "0s - loss: 543.0165\n",
      "Epoch 123/200\n",
      "0s - loss: 540.2106\n",
      "Epoch 124/200\n",
      "0s - loss: 539.0697\n",
      "Epoch 125/200\n",
      "0s - loss: 553.9405\n",
      "Epoch 126/200\n",
      "0s - loss: 536.1454\n",
      "Epoch 127/200\n",
      "0s - loss: 538.5323\n",
      "Epoch 128/200\n",
      "0s - loss: 547.4024\n",
      "Epoch 129/200\n",
      "0s - loss: 542.4658\n",
      "Epoch 130/200\n",
      "0s - loss: 545.7219\n",
      "Epoch 131/200\n",
      "0s - loss: 541.4138\n",
      "Epoch 132/200\n",
      "0s - loss: 547.5129\n",
      "Epoch 133/200\n",
      "0s - loss: 549.0994\n",
      "Epoch 134/200\n",
      "0s - loss: 540.7679\n",
      "Epoch 135/200\n",
      "0s - loss: 539.1448\n",
      "Epoch 136/200\n",
      "0s - loss: 553.9308\n",
      "Epoch 137/200\n",
      "0s - loss: 542.9812\n",
      "Epoch 138/200\n",
      "0s - loss: 549.1420\n",
      "Epoch 139/200\n",
      "0s - loss: 544.1188\n",
      "Epoch 140/200\n",
      "0s - loss: 565.0828\n",
      "Epoch 141/200\n",
      "0s - loss: 537.3349\n",
      "Epoch 142/200\n",
      "0s - loss: 541.4444\n",
      "Epoch 143/200\n",
      "0s - loss: 537.6784\n",
      "Epoch 144/200\n",
      "0s - loss: 537.8722\n",
      "Epoch 145/200\n",
      "0s - loss: 544.9403\n",
      "Epoch 146/200\n",
      "0s - loss: 553.9888\n",
      "Epoch 147/200\n",
      "0s - loss: 550.8749\n",
      "Epoch 148/200\n",
      "0s - loss: 549.3067\n",
      "Epoch 149/200\n",
      "0s - loss: 540.0851\n",
      "Epoch 150/200\n",
      "0s - loss: 543.1973\n",
      "Epoch 151/200\n",
      "0s - loss: 541.4359\n",
      "Epoch 152/200\n",
      "0s - loss: 546.4467\n",
      "Epoch 153/200\n",
      "0s - loss: 542.3676\n",
      "Epoch 154/200\n",
      "0s - loss: 537.7796\n",
      "Epoch 155/200\n",
      "0s - loss: 545.7723\n",
      "Epoch 156/200\n",
      "0s - loss: 536.4816\n",
      "Epoch 157/200\n",
      "0s - loss: 546.3242\n",
      "Epoch 158/200\n",
      "0s - loss: 548.4483\n",
      "Epoch 159/200\n",
      "0s - loss: 538.9017\n",
      "Epoch 160/200\n",
      "0s - loss: 539.6211\n",
      "Epoch 161/200\n",
      "0s - loss: 539.4636\n",
      "Epoch 162/200\n",
      "0s - loss: 567.5564\n",
      "Epoch 163/200\n",
      "0s - loss: 557.6617\n",
      "Epoch 164/200\n",
      "0s - loss: 538.4400\n",
      "Epoch 165/200\n",
      "0s - loss: 538.4565\n",
      "Epoch 166/200\n",
      "0s - loss: 545.1274\n",
      "Epoch 167/200\n",
      "0s - loss: 537.7568\n",
      "Epoch 168/200\n",
      "0s - loss: 544.0281\n",
      "Epoch 169/200\n",
      "0s - loss: 539.3020\n",
      "Epoch 170/200\n",
      "0s - loss: 559.3423\n",
      "Epoch 171/200\n",
      "0s - loss: 544.6370\n",
      "Epoch 172/200\n",
      "0s - loss: 544.5213\n",
      "Epoch 173/200\n",
      "0s - loss: 554.0639\n",
      "Epoch 174/200\n",
      "0s - loss: 547.5519\n",
      "Epoch 175/200\n",
      "0s - loss: 538.8427\n",
      "Epoch 176/200\n",
      "0s - loss: 545.9267\n",
      "Epoch 177/200\n",
      "0s - loss: 534.5817\n",
      "Epoch 178/200\n",
      "0s - loss: 539.2070\n",
      "Epoch 179/200\n",
      "0s - loss: 541.1818\n",
      "Epoch 180/200\n",
      "0s - loss: 543.9873\n",
      "Epoch 181/200\n",
      "0s - loss: 558.9915\n",
      "Epoch 182/200\n",
      "0s - loss: 547.7058\n",
      "Epoch 183/200\n",
      "0s - loss: 536.4981\n",
      "Epoch 184/200\n",
      "0s - loss: 535.0668\n",
      "Epoch 185/200\n",
      "0s - loss: 567.0480\n",
      "Epoch 186/200\n",
      "0s - loss: 535.5100\n",
      "Epoch 187/200\n",
      "0s - loss: 539.7379\n",
      "Epoch 188/200\n",
      "0s - loss: 541.1387\n",
      "Epoch 189/200\n",
      "0s - loss: 547.7782\n",
      "Epoch 190/200\n",
      "0s - loss: 541.6068\n",
      "Epoch 191/200\n",
      "0s - loss: 539.6431\n",
      "Epoch 192/200\n",
      "0s - loss: 541.7338\n",
      "Epoch 193/200\n",
      "0s - loss: 553.4232\n",
      "Epoch 194/200\n",
      "0s - loss: 544.3559\n",
      "Epoch 195/200\n",
      "0s - loss: 536.7003\n",
      "Epoch 196/200\n",
      "0s - loss: 555.4203\n",
      "Epoch 197/200\n",
      "0s - loss: 552.2813\n",
      "Epoch 198/200\n",
      "0s - loss: 541.2206\n",
      "Epoch 199/200\n",
      "0s - loss: 542.3271\n",
      "Epoch 200/200\n",
      "0s - loss: 534.2085\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x11e8bfa90>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create and fit Multilayer Perceptron model\n",
    "model = Sequential()\n",
    "model.add(Dense(8, input_dim=look_back, activation='relu'))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "model.fit(trainX, trainY, nb_epoch=200, batch_size=2, verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Score: 532.59 MSE (23.08 RMSE)\n",
      "Test Score: 2358.07 MSE (48.56 RMSE)\n"
     ]
    }
   ],
   "source": [
    "# estimate model performance\n",
    "trainScore = model.evaluate(trainX, trainY, verbose=0)\n",
    "print('Train Score: %.2f MSE (%.2f RMSE)' % (trainScore, math.sqrt(trainScore)))\n",
    "testScore = model.evaluate(testX, testY, verbose=0)\n",
    "print('Test Score: %.2f MSE (%.2f RMSE)' % (testScore, math.sqrt(testScore)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg4RQT8IgqBDRNAPgiDoEH8Cr2Vg1Fhi57wAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113fbb3d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# generate predictions for training\n",
    "trainPredict = model.predict(trainX)\n",
    "testPredict = model.predict(testX)\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(dataset)\n",
    "plt.plot(trainPredictPlot)\n",
    "plt.plot(testPredictPlot)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 112.  118.  132.]\n",
      " [ 118.  132.  129.]\n",
      " [ 132.  129.  121.]\n",
      " [ 129.  121.  135.]\n",
      " [ 121.  135.  148.]]\n",
      "[ 129.  121.  135.  148.  148.]\n"
     ]
    }
   ],
   "source": [
    "# reshape dataset\n",
    "look_back = 3\n",
    "trainX, trainY = create_dataset(train, look_back)\n",
    "testX, testY = create_dataset(test, look_back)\n",
    "\n",
    "print trainX[:5]\n",
    "print trainY[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      "0s - loss: 83671.8294\n",
      "Epoch 2/200\n",
      "0s - loss: 40638.3085\n",
      "Epoch 3/200\n",
      "0s - loss: 17466.7385\n",
      "Epoch 4/200\n",
      "0s - loss: 6076.9101\n",
      "Epoch 5/200\n",
      "0s - loss: 2034.2136\n",
      "Epoch 6/200\n",
      "0s - loss: 932.3104\n",
      "Epoch 7/200\n",
      "0s - loss: 748.6754\n",
      "Epoch 8/200\n",
      "0s - loss: 734.4747\n",
      "Epoch 9/200\n",
      "0s - loss: 727.3987\n",
      "Epoch 10/200\n",
      "0s - loss: 726.3489\n",
      "Epoch 11/200\n",
      "0s - loss: 722.0118\n",
      "Epoch 12/200\n",
      "0s - loss: 719.5558\n",
      "Epoch 13/200\n",
      "0s - loss: 714.6521\n",
      "Epoch 14/200\n",
      "0s - loss: 714.7315\n",
      "Epoch 15/200\n",
      "0s - loss: 711.5931\n",
      "Epoch 16/200\n",
      "0s - loss: 712.9244\n",
      "Epoch 17/200\n",
      "0s - loss: 709.0207\n",
      "Epoch 18/200\n",
      "0s - loss: 723.3592\n",
      "Epoch 19/200\n",
      "0s - loss: 704.4948\n",
      "Epoch 20/200\n",
      "0s - loss: 722.6836\n",
      "Epoch 21/200\n",
      "0s - loss: 701.0576\n",
      "Epoch 22/200\n",
      "0s - loss: 709.4541\n",
      "Epoch 23/200\n",
      "0s - loss: 690.1373\n",
      "Epoch 24/200\n",
      "0s - loss: 692.1519\n",
      "Epoch 25/200\n",
      "0s - loss: 680.6153\n",
      "Epoch 26/200\n",
      "0s - loss: 694.8279\n",
      "Epoch 27/200\n",
      "0s - loss: 678.1819\n",
      "Epoch 28/200\n",
      "0s - loss: 713.8649\n",
      "Epoch 29/200\n",
      "0s - loss: 667.2840\n",
      "Epoch 30/200\n",
      "0s - loss: 673.4817\n",
      "Epoch 31/200\n",
      "0s - loss: 663.2846\n",
      "Epoch 32/200\n",
      "0s - loss: 665.4083\n",
      "Epoch 33/200\n",
      "0s - loss: 656.9186\n",
      "Epoch 34/200\n",
      "0s - loss: 649.5907\n",
      "Epoch 35/200\n",
      "0s - loss: 661.0228\n",
      "Epoch 36/200\n",
      "0s - loss: 647.2040\n",
      "Epoch 37/200\n",
      "0s - loss: 664.3735\n",
      "Epoch 38/200\n",
      "0s - loss: 644.9487\n",
      "Epoch 39/200\n",
      "0s - loss: 642.1060\n",
      "Epoch 40/200\n",
      "0s - loss: 633.1854\n",
      "Epoch 41/200\n",
      "0s - loss: 631.7398\n",
      "Epoch 42/200\n",
      "0s - loss: 648.3790\n",
      "Epoch 43/200\n",
      "0s - loss: 624.4023\n",
      "Epoch 44/200\n",
      "0s - loss: 624.6325\n",
      "Epoch 45/200\n",
      "0s - loss: 629.7916\n",
      "Epoch 46/200\n",
      "0s - loss: 613.1485\n",
      "Epoch 47/200\n",
      "0s - loss: 604.0728\n",
      "Epoch 48/200\n",
      "0s - loss: 607.1932\n",
      "Epoch 49/200\n",
      "0s - loss: 612.8795\n",
      "Epoch 50/200\n",
      "0s - loss: 599.3480\n",
      "Epoch 51/200\n",
      "0s - loss: 601.9850\n",
      "Epoch 52/200\n",
      "0s - loss: 615.3567\n",
      "Epoch 53/200\n",
      "0s - loss: 613.0596\n",
      "Epoch 54/200\n",
      "0s - loss: 583.3278\n",
      "Epoch 55/200\n",
      "0s - loss: 589.8806\n",
      "Epoch 56/200\n",
      "0s - loss: 643.9668\n",
      "Epoch 57/200\n",
      "0s - loss: 589.8897\n",
      "Epoch 58/200\n",
      "0s - loss: 597.7172\n",
      "Epoch 59/200\n",
      "0s - loss: 567.1977\n",
      "Epoch 60/200\n",
      "0s - loss: 579.9892\n",
      "Epoch 61/200\n",
      "0s - loss: 569.3119\n",
      "Epoch 62/200\n",
      "0s - loss: 576.3969\n",
      "Epoch 63/200\n",
      "0s - loss: 581.8441\n",
      "Epoch 64/200\n",
      "0s - loss: 583.1991\n",
      "Epoch 65/200\n",
      "0s - loss: 559.3836\n",
      "Epoch 66/200\n",
      "0s - loss: 584.8118\n",
      "Epoch 67/200\n",
      "0s - loss: 560.9598\n",
      "Epoch 68/200\n",
      "0s - loss: 559.0342\n",
      "Epoch 69/200\n",
      "0s - loss: 560.8672\n",
      "Epoch 70/200\n",
      "0s - loss: 570.1347\n",
      "Epoch 71/200\n",
      "0s - loss: 548.5865\n",
      "Epoch 72/200\n",
      "0s - loss: 561.4219\n",
      "Epoch 73/200\n",
      "0s - loss: 549.3091\n",
      "Epoch 74/200\n",
      "0s - loss: 573.4158\n",
      "Epoch 75/200\n",
      "0s - loss: 541.3438\n",
      "Epoch 76/200\n",
      "0s - loss: 533.3077\n",
      "Epoch 77/200\n",
      "0s - loss: 551.7571\n",
      "Epoch 78/200\n",
      "0s - loss: 545.5883\n",
      "Epoch 79/200\n",
      "0s - loss: 531.6289\n",
      "Epoch 80/200\n",
      "0s - loss: 539.7587\n",
      "Epoch 81/200\n",
      "0s - loss: 513.1625\n",
      "Epoch 82/200\n",
      "0s - loss: 558.3036\n",
      "Epoch 83/200\n",
      "0s - loss: 545.1788\n",
      "Epoch 84/200\n",
      "0s - loss: 525.3847\n",
      "Epoch 85/200\n",
      "0s - loss: 533.3447\n",
      "Epoch 86/200\n",
      "0s - loss: 534.4642\n",
      "Epoch 87/200\n",
      "0s - loss: 520.5509\n",
      "Epoch 88/200\n",
      "0s - loss: 508.7012\n",
      "Epoch 89/200\n",
      "0s - loss: 519.3589\n",
      "Epoch 90/200\n",
      "0s - loss: 522.2292\n",
      "Epoch 91/200\n",
      "0s - loss: 524.6929\n",
      "Epoch 92/200\n",
      "0s - loss: 516.6113\n",
      "Epoch 93/200\n",
      "0s - loss: 515.8494\n",
      "Epoch 94/200\n",
      "0s - loss: 530.9749\n",
      "Epoch 95/200\n",
      "0s - loss: 521.6811\n",
      "Epoch 96/200\n",
      "0s - loss: 505.9416\n",
      "Epoch 97/200\n",
      "0s - loss: 519.8670\n",
      "Epoch 98/200\n",
      "0s - loss: 565.5602\n",
      "Epoch 99/200\n",
      "0s - loss: 506.7910\n",
      "Epoch 100/200\n",
      "0s - loss: 506.3179\n",
      "Epoch 101/200\n",
      "0s - loss: 519.9980\n",
      "Epoch 102/200\n",
      "0s - loss: 504.5565\n",
      "Epoch 103/200\n",
      "0s - loss: 510.9430\n",
      "Epoch 104/200\n",
      "0s - loss: 539.7244\n",
      "Epoch 105/200\n",
      "0s - loss: 535.6204\n",
      "Epoch 106/200\n",
      "0s - loss: 498.2837\n",
      "Epoch 107/200\n",
      "0s - loss: 524.3867\n",
      "Epoch 108/200\n",
      "0s - loss: 524.3399\n",
      "Epoch 109/200\n",
      "0s - loss: 516.0490\n",
      "Epoch 110/200\n",
      "0s - loss: 501.0972\n",
      "Epoch 111/200\n",
      "0s - loss: 554.1910\n",
      "Epoch 112/200\n",
      "0s - loss: 498.6965\n",
      "Epoch 113/200\n",
      "0s - loss: 530.7698\n",
      "Epoch 114/200\n",
      "0s - loss: 500.5335\n",
      "Epoch 115/200\n",
      "0s - loss: 518.2787\n",
      "Epoch 116/200\n",
      "0s - loss: 512.0570\n",
      "Epoch 117/200\n",
      "0s - loss: 502.9501\n",
      "Epoch 118/200\n",
      "0s - loss: 513.0629\n",
      "Epoch 119/200\n",
      "0s - loss: 503.7768\n",
      "Epoch 120/200\n",
      "0s - loss: 542.8642\n",
      "Epoch 121/200\n",
      "0s - loss: 506.4581\n",
      "Epoch 122/200\n",
      "0s - loss: 503.4713\n",
      "Epoch 123/200\n",
      "0s - loss: 523.9442\n",
      "Epoch 124/200\n",
      "0s - loss: 501.2942\n",
      "Epoch 125/200\n",
      "0s - loss: 504.4891\n",
      "Epoch 126/200\n",
      "0s - loss: 519.6716\n",
      "Epoch 127/200\n",
      "0s - loss: 510.6814\n",
      "Epoch 128/200\n",
      "0s - loss: 480.3752\n",
      "Epoch 129/200\n",
      "0s - loss: 529.9507\n",
      "Epoch 130/200\n",
      "0s - loss: 502.9895\n",
      "Epoch 131/200\n",
      "0s - loss: 518.9047\n",
      "Epoch 132/200\n",
      "0s - loss: 503.7314\n",
      "Epoch 133/200\n",
      "0s - loss: 485.6632\n",
      "Epoch 134/200\n",
      "0s - loss: 512.2738\n",
      "Epoch 135/200\n",
      "0s - loss: 501.4503\n",
      "Epoch 136/200\n",
      "0s - loss: 495.6301\n",
      "Epoch 137/200\n",
      "0s - loss: 538.7277\n",
      "Epoch 138/200\n",
      "0s - loss: 488.5048\n",
      "Epoch 139/200\n",
      "0s - loss: 507.3818\n",
      "Epoch 140/200\n",
      "0s - loss: 498.2006\n",
      "Epoch 141/200\n",
      "0s - loss: 495.1735\n",
      "Epoch 142/200\n",
      "0s - loss: 500.8449\n",
      "Epoch 143/200\n",
      "0s - loss: 507.8072\n",
      "Epoch 144/200\n",
      "0s - loss: 496.9582\n",
      "Epoch 145/200\n",
      "0s - loss: 487.8992\n",
      "Epoch 146/200\n",
      "0s - loss: 497.8694\n",
      "Epoch 147/200\n",
      "0s - loss: 546.3169\n",
      "Epoch 148/200\n",
      "0s - loss: 513.7862\n",
      "Epoch 149/200\n",
      "0s - loss: 518.4526\n",
      "Epoch 150/200\n",
      "0s - loss: 481.3360\n",
      "Epoch 151/200\n",
      "0s - loss: 505.1137\n",
      "Epoch 152/200\n",
      "0s - loss: 506.2680\n",
      "Epoch 153/200\n",
      "0s - loss: 504.4509\n",
      "Epoch 154/200\n",
      "0s - loss: 496.1040\n",
      "Epoch 155/200\n",
      "0s - loss: 500.4510\n",
      "Epoch 156/200\n",
      "0s - loss: 494.7459\n",
      "Epoch 157/200\n",
      "0s - loss: 539.7890\n",
      "Epoch 158/200\n",
      "0s - loss: 508.7687\n",
      "Epoch 159/200\n",
      "0s - loss: 514.3376\n",
      "Epoch 160/200\n",
      "0s - loss: 505.8222\n",
      "Epoch 161/200\n",
      "0s - loss: 499.7097\n",
      "Epoch 162/200\n",
      "0s - loss: 491.1051\n",
      "Epoch 163/200\n",
      "0s - loss: 493.6081\n",
      "Epoch 164/200\n",
      "0s - loss: 491.8929\n",
      "Epoch 165/200\n",
      "0s - loss: 502.3509\n",
      "Epoch 166/200\n",
      "0s - loss: 490.7454\n",
      "Epoch 167/200\n",
      "0s - loss: 523.4744\n",
      "Epoch 168/200\n",
      "0s - loss: 503.4264\n",
      "Epoch 169/200\n",
      "0s - loss: 530.0948\n",
      "Epoch 170/200\n",
      "0s - loss: 500.3688\n",
      "Epoch 171/200\n",
      "0s - loss: 510.7172\n",
      "Epoch 172/200\n",
      "0s - loss: 500.8852\n",
      "Epoch 173/200\n",
      "0s - loss: 520.3327\n",
      "Epoch 174/200\n",
      "0s - loss: 500.4319\n",
      "Epoch 175/200\n",
      "0s - loss: 503.5199\n",
      "Epoch 176/200\n",
      "0s - loss: 493.0779\n",
      "Epoch 177/200\n",
      "0s - loss: 510.1977\n",
      "Epoch 178/200\n",
      "0s - loss: 492.6268\n",
      "Epoch 179/200\n",
      "0s - loss: 501.7902\n",
      "Epoch 180/200\n",
      "0s - loss: 498.4706\n",
      "Epoch 181/200\n",
      "0s - loss: 493.1729\n",
      "Epoch 182/200\n",
      "0s - loss: 490.9330\n",
      "Epoch 183/200\n",
      "0s - loss: 488.4096\n",
      "Epoch 184/200\n",
      "0s - loss: 499.5969\n",
      "Epoch 185/200\n",
      "0s - loss: 503.9610\n",
      "Epoch 186/200\n",
      "0s - loss: 488.1497\n",
      "Epoch 187/200\n",
      "0s - loss: 488.0814\n",
      "Epoch 188/200\n",
      "0s - loss: 497.3637\n",
      "Epoch 189/200\n",
      "0s - loss: 508.5562\n",
      "Epoch 190/200\n",
      "0s - loss: 504.3515\n",
      "Epoch 191/200\n",
      "0s - loss: 501.1038\n",
      "Epoch 192/200\n",
      "0s - loss: 525.1095\n",
      "Epoch 193/200\n",
      "0s - loss: 487.4438\n",
      "Epoch 194/200\n",
      "0s - loss: 520.4491\n",
      "Epoch 195/200\n",
      "0s - loss: 497.5151\n",
      "Epoch 196/200\n",
      "0s - loss: 490.3914\n",
      "Epoch 197/200\n",
      "0s - loss: 496.7289\n",
      "Epoch 198/200\n",
      "0s - loss: 497.4330\n",
      "Epoch 199/200\n",
      "0s - loss: 484.6063\n",
      "Epoch 200/200\n",
      "0s - loss: 494.3108\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x11f2ee950>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create and fit Multilayer Perceptron model\n",
    "model = Sequential()\n",
    "model.add(Dense(8, input_dim=look_back, activation='relu'))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "model.fit(trainX, trainY, nb_epoch=200, batch_size=2, verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Score: 481.13 MSE (21.93 RMSE)\n",
      "Test Score: 1986.48 MSE (44.57 RMSE)\n"
     ]
    }
   ],
   "source": [
    "# estimate model performance\n",
    "trainScore = model.evaluate(trainX, trainY, verbose=0)\n",
    "print('Train Score: %.2f MSE (%.2f RMSE)' % (trainScore, math.sqrt(trainScore)))\n",
    "testScore = model.evaluate(testX, testY, verbose=0)\n",
    "print('Test Score: %.2f MSE (%.2f RMSE)' % (testScore, math.sqrt(testScore)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11f409e90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# generate predictions for training\n",
    "trainPredict = model.predict(trainX)\n",
    "testPredict = model.predict(testX)\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(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
}