Chapter 24 Time series prediction with MLP.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas\n",
    "import matplotlib.pyplot as plt\n",
    "dataset = pandas.read_csv('AirPassengers.csv',usecols=[1],engine='python')\n",
    "plt.plot(dataset)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# MLP regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "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"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "numpy.random.seed(7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = dataset.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = dataset.astype('float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[112.],\n",
       "       [118.],\n",
       "       [132.],\n",
       "       [129.],\n",
       "       [121.],\n",
       "       [135.],\n",
       "       [148.],\n",
       "       [148.],\n",
       "       [136.],\n",
       "       [119.],\n",
       "       [104.],\n",
       "       [118.],\n",
       "       [115.],\n",
       "       [126.],\n",
       "       [141.],\n",
       "       [135.],\n",
       "       [125.],\n",
       "       [149.],\n",
       "       [170.],\n",
       "       [170.],\n",
       "       [158.],\n",
       "       [133.],\n",
       "       [114.],\n",
       "       [140.],\n",
       "       [145.],\n",
       "       [150.],\n",
       "       [178.],\n",
       "       [163.],\n",
       "       [172.],\n",
       "       [178.],\n",
       "       [199.],\n",
       "       [199.],\n",
       "       [184.],\n",
       "       [162.],\n",
       "       [146.],\n",
       "       [166.],\n",
       "       [171.],\n",
       "       [180.],\n",
       "       [193.],\n",
       "       [181.],\n",
       "       [183.],\n",
       "       [218.],\n",
       "       [230.],\n",
       "       [242.],\n",
       "       [209.],\n",
       "       [191.],\n",
       "       [172.],\n",
       "       [194.],\n",
       "       [196.],\n",
       "       [196.],\n",
       "       [236.],\n",
       "       [235.],\n",
       "       [229.],\n",
       "       [243.],\n",
       "       [264.],\n",
       "       [272.],\n",
       "       [237.],\n",
       "       [211.],\n",
       "       [180.],\n",
       "       [201.],\n",
       "       [204.],\n",
       "       [188.],\n",
       "       [235.],\n",
       "       [227.],\n",
       "       [234.],\n",
       "       [264.],\n",
       "       [302.],\n",
       "       [293.],\n",
       "       [259.],\n",
       "       [229.],\n",
       "       [203.],\n",
       "       [229.],\n",
       "       [242.],\n",
       "       [233.],\n",
       "       [267.],\n",
       "       [269.],\n",
       "       [270.],\n",
       "       [315.],\n",
       "       [364.],\n",
       "       [347.],\n",
       "       [312.],\n",
       "       [274.],\n",
       "       [237.],\n",
       "       [278.],\n",
       "       [284.],\n",
       "       [277.],\n",
       "       [317.],\n",
       "       [313.],\n",
       "       [318.],\n",
       "       [374.],\n",
       "       [413.],\n",
       "       [405.],\n",
       "       [355.],\n",
       "       [306.],\n",
       "       [271.],\n",
       "       [306.],\n",
       "       [315.],\n",
       "       [301.],\n",
       "       [356.],\n",
       "       [348.],\n",
       "       [355.],\n",
       "       [422.],\n",
       "       [465.],\n",
       "       [467.],\n",
       "       [404.],\n",
       "       [347.],\n",
       "       [305.],\n",
       "       [336.],\n",
       "       [340.],\n",
       "       [318.],\n",
       "       [362.],\n",
       "       [348.],\n",
       "       [363.],\n",
       "       [435.],\n",
       "       [491.],\n",
       "       [505.],\n",
       "       [404.],\n",
       "       [359.],\n",
       "       [310.],\n",
       "       [337.],\n",
       "       [360.],\n",
       "       [342.],\n",
       "       [406.],\n",
       "       [396.],\n",
       "       [420.],\n",
       "       [472.],\n",
       "       [548.],\n",
       "       [559.],\n",
       "       [463.],\n",
       "       [407.],\n",
       "       [362.],\n",
       "       [405.],\n",
       "       [417.],\n",
       "       [391.],\n",
       "       [419.],\n",
       "       [461.],\n",
       "       [472.],\n",
       "       [535.],\n",
       "       [622.],\n",
       "       [606.],\n",
       "       [508.],\n",
       "       [461.],\n",
       "       [390.],\n",
       "       [432.]], dtype=float32)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "96 48\n"
     ]
    }
   ],
   "source": [
    "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": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "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": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "look_back=1\n",
    "trainX, trainY = create_dataset(train,look_back)\n",
    "testX, testY = create_dataset(test, look_back)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "W0426 16:58:43.587568 140691999496000 deprecation.py:506] From /home/arvind/.local/lib/python3.6/site-packages/tensorflow_core/python/ops/resource_variable_ops.py:1633: calling BaseResourceVariable.__init__ (from tensorflow.python.ops.resource_variable_ops) with constraint is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "If using Keras pass *_constraint arguments to layers.\n",
      "W0426 16:58:43.784261 140691999496000 deprecation.py:323] From /home/arvind/.local/lib/python3.6/site-packages/keras/optimizers.py:550: BaseResourceVariable.constraint (from tensorflow.python.ops.resource_variable_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Apply a constraint manually following the optimizer update step.\n",
      "W0426 16:58:43.928082 140691999496000 module_wrapper.py:136] From /home/arvind/.local/lib/python3.6/site-packages/tensorflow_core/python/util/module_wrapper.py:163: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      " - 0s - loss: 47555.9284\n",
      "Epoch 2/200\n",
      " - 0s - loss: 41653.3919\n",
      "Epoch 3/200\n",
      " - 0s - loss: 35699.0043\n",
      "Epoch 4/200\n",
      " - 0s - loss: 29483.6718\n",
      "Epoch 5/200\n",
      " - 0s - loss: 23393.8455\n",
      "Epoch 6/200\n",
      " - 0s - loss: 17691.3700\n",
      "Epoch 7/200\n",
      " - 0s - loss: 12745.1545\n",
      "Epoch 8/200\n",
      " - 0s - loss: 8651.8868\n",
      "Epoch 9/200\n",
      " - 0s - loss: 5562.2341\n",
      "Epoch 10/200\n",
      " - 0s - loss: 3415.3183\n",
      "Epoch 11/200\n",
      " - 0s - loss: 2055.5277\n",
      "Epoch 12/200\n",
      " - 0s - loss: 1282.6550\n",
      "Epoch 13/200\n",
      " - 0s - loss: 863.3661\n",
      "Epoch 14/200\n",
      " - 0s - loss: 673.2044\n",
      "Epoch 15/200\n",
      " - 0s - loss: 584.3995\n",
      "Epoch 16/200\n",
      " - 0s - loss: 554.8133\n",
      "Epoch 17/200\n",
      " - 0s - loss: 546.6272\n",
      "Epoch 18/200\n",
      " - 0s - loss: 547.3683\n",
      "Epoch 19/200\n",
      " - 0s - loss: 538.8246\n",
      "Epoch 20/200\n",
      " - 0s - loss: 538.5317\n",
      "Epoch 21/200\n",
      " - 0s - loss: 536.9810\n",
      "Epoch 22/200\n",
      " - 0s - loss: 536.0283\n",
      "Epoch 23/200\n",
      " - 0s - loss: 537.1247\n",
      "Epoch 24/200\n",
      " - 0s - loss: 535.7827\n",
      "Epoch 25/200\n",
      " - 0s - loss: 538.2108\n",
      "Epoch 26/200\n",
      " - 0s - loss: 538.4736\n",
      "Epoch 27/200\n",
      " - 0s - loss: 538.2303\n",
      "Epoch 28/200\n",
      " - 0s - loss: 537.3810\n",
      "Epoch 29/200\n",
      " - 0s - loss: 538.2823\n",
      "Epoch 30/200\n",
      " - 0s - loss: 539.0556\n",
      "Epoch 31/200\n",
      " - 0s - loss: 536.2942\n",
      "Epoch 32/200\n",
      " - 0s - loss: 539.0038\n",
      "Epoch 33/200\n",
      " - 0s - loss: 536.8501\n",
      "Epoch 34/200\n",
      " - 0s - loss: 539.1813\n",
      "Epoch 35/200\n",
      " - 0s - loss: 537.3514\n",
      "Epoch 36/200\n",
      " - 0s - loss: 538.6051\n",
      "Epoch 37/200\n",
      " - 0s - loss: 538.0627\n",
      "Epoch 38/200\n",
      " - 0s - loss: 541.0699\n",
      "Epoch 39/200\n",
      " - 0s - loss: 538.1966\n",
      "Epoch 40/200\n",
      " - 0s - loss: 536.2566\n",
      "Epoch 41/200\n",
      " - 0s - loss: 536.8805\n",
      "Epoch 42/200\n",
      " - 0s - loss: 538.0673\n",
      "Epoch 43/200\n",
      " - 0s - loss: 542.7847\n",
      "Epoch 44/200\n",
      " - 0s - loss: 538.2604\n",
      "Epoch 45/200\n",
      " - 0s - loss: 537.6739\n",
      "Epoch 46/200\n",
      " - 0s - loss: 541.4314\n",
      "Epoch 47/200\n",
      " - 0s - loss: 540.7052\n",
      "Epoch 48/200\n",
      " - 0s - loss: 540.6792\n",
      "Epoch 49/200\n",
      " - 0s - loss: 535.4652\n",
      "Epoch 50/200\n",
      " - 0s - loss: 539.8989\n",
      "Epoch 51/200\n",
      " - 0s - loss: 538.1576\n",
      "Epoch 52/200\n",
      " - 0s - loss: 535.3379\n",
      "Epoch 53/200\n",
      " - 0s - loss: 536.7900\n",
      "Epoch 54/200\n",
      " - 0s - loss: 542.9296\n",
      "Epoch 55/200\n",
      " - 0s - loss: 537.2221\n",
      "Epoch 56/200\n",
      " - 0s - loss: 541.2078\n",
      "Epoch 57/200\n",
      " - 0s - loss: 549.5546\n",
      "Epoch 58/200\n",
      " - 0s - loss: 544.1331\n",
      "Epoch 59/200\n",
      " - 0s - loss: 540.0629\n",
      "Epoch 60/200\n",
      " - 0s - loss: 540.7305\n",
      "Epoch 61/200\n",
      " - 0s - loss: 540.3415\n",
      "Epoch 62/200\n",
      " - 0s - loss: 543.9038\n",
      "Epoch 63/200\n",
      " - 0s - loss: 537.8564\n",
      "Epoch 64/200\n",
      " - 0s - loss: 540.5008\n",
      "Epoch 65/200\n",
      " - 0s - loss: 537.9627\n",
      "Epoch 66/200\n",
      " - 0s - loss: 538.8419\n",
      "Epoch 67/200\n",
      " - 0s - loss: 549.7381\n",
      "Epoch 68/200\n",
      " - 0s - loss: 558.7654\n",
      "Epoch 69/200\n",
      " - 0s - loss: 537.8396\n",
      "Epoch 70/200\n",
      " - 0s - loss: 537.5499\n",
      "Epoch 71/200\n",
      " - 0s - loss: 543.5098\n",
      "Epoch 72/200\n",
      " - 0s - loss: 540.4309\n",
      "Epoch 73/200\n",
      " - 0s - loss: 540.0852\n",
      "Epoch 74/200\n",
      " - 0s - loss: 538.8094\n",
      "Epoch 75/200\n",
      " - 0s - loss: 540.9883\n",
      "Epoch 76/200\n",
      " - 0s - loss: 537.0864\n",
      "Epoch 77/200\n",
      " - 0s - loss: 540.0628\n",
      "Epoch 78/200\n",
      " - 0s - loss: 535.3524\n",
      "Epoch 79/200\n",
      " - 0s - loss: 541.5892\n",
      "Epoch 80/200\n",
      " - 0s - loss: 540.8637\n",
      "Epoch 81/200\n",
      " - 0s - loss: 537.5930\n",
      "Epoch 82/200\n",
      " - 0s - loss: 536.5664\n",
      "Epoch 83/200\n",
      " - 0s - loss: 549.5660\n",
      "Epoch 84/200\n",
      " - 0s - loss: 545.8253\n",
      "Epoch 85/200\n",
      " - 0s - loss: 538.5540\n",
      "Epoch 86/200\n",
      " - 0s - loss: 538.1985\n",
      "Epoch 87/200\n",
      " - 0s - loss: 537.5314\n",
      "Epoch 88/200\n",
      " - 0s - loss: 539.7148\n",
      "Epoch 89/200\n",
      " - 0s - loss: 541.6901\n",
      "Epoch 90/200\n",
      " - 0s - loss: 548.9254\n",
      "Epoch 91/200\n",
      " - 0s - loss: 543.7339\n",
      "Epoch 92/200\n",
      " - 0s - loss: 540.1132\n",
      "Epoch 93/200\n",
      " - 0s - loss: 538.2214\n",
      "Epoch 94/200\n",
      " - 0s - loss: 536.7756\n",
      "Epoch 95/200\n",
      " - 0s - loss: 543.4021\n",
      "Epoch 96/200\n",
      " - 0s - loss: 538.1066\n",
      "Epoch 97/200\n",
      " - 0s - loss: 539.9250\n",
      "Epoch 98/200\n",
      " - 0s - loss: 537.8863\n",
      "Epoch 99/200\n",
      " - 0s - loss: 541.8814\n",
      "Epoch 100/200\n",
      " - 0s - loss: 538.2110\n",
      "Epoch 101/200\n",
      " - 0s - loss: 537.9105\n",
      "Epoch 102/200\n",
      " - 0s - loss: 536.2739\n",
      "Epoch 103/200\n",
      " - 0s - loss: 542.0661\n",
      "Epoch 104/200\n",
      " - 0s - loss: 534.6675\n",
      "Epoch 105/200\n",
      " - 0s - loss: 542.8574\n",
      "Epoch 106/200\n",
      " - 0s - loss: 540.4594\n",
      "Epoch 107/200\n",
      " - 0s - loss: 538.4268\n",
      "Epoch 108/200\n",
      " - 0s - loss: 541.5730\n",
      "Epoch 109/200\n",
      " - 0s - loss: 536.9202\n",
      "Epoch 110/200\n",
      " - 0s - loss: 541.7031\n",
      "Epoch 111/200\n",
      " - 0s - loss: 537.9534\n",
      "Epoch 112/200\n",
      " - 0s - loss: 540.5577\n",
      "Epoch 113/200\n",
      " - 0s - loss: 539.6316\n",
      "Epoch 114/200\n",
      " - 0s - loss: 534.2774\n",
      "Epoch 115/200\n",
      " - 0s - loss: 544.9202\n",
      "Epoch 116/200\n",
      " - 0s - loss: 538.8419\n",
      "Epoch 117/200\n",
      " - 0s - loss: 539.7734\n",
      "Epoch 118/200\n",
      " - 0s - loss: 539.0312\n",
      "Epoch 119/200\n",
      " - 0s - loss: 540.0628\n",
      "Epoch 120/200\n",
      " - 0s - loss: 538.5755\n",
      "Epoch 121/200\n",
      " - 0s - loss: 539.9155\n",
      "Epoch 122/200\n",
      " - 0s - loss: 539.8793\n",
      "Epoch 123/200\n",
      " - 0s - loss: 538.2963\n",
      "Epoch 124/200\n",
      " - 0s - loss: 537.4978\n",
      "Epoch 125/200\n",
      " - 0s - loss: 546.9351\n",
      "Epoch 126/200\n",
      " - 0s - loss: 534.9721\n",
      "Epoch 127/200\n",
      " - 0s - loss: 538.2620\n",
      "Epoch 128/200\n",
      " - 0s - loss: 542.0791\n",
      "Epoch 129/200\n",
      " - 0s - loss: 540.1608\n",
      "Epoch 130/200\n",
      " - 0s - loss: 540.9398\n",
      "Epoch 131/200\n",
      " - 0s - loss: 540.5172\n",
      "Epoch 132/200\n",
      " - 0s - loss: 541.9306\n",
      "Epoch 133/200\n",
      " - 0s - loss: 546.7708\n",
      "Epoch 134/200\n",
      " - 0s - loss: 536.5134\n",
      "Epoch 135/200\n",
      " - 0s - loss: 537.8884\n",
      "Epoch 136/200\n",
      " - 0s - loss: 546.7969\n",
      "Epoch 137/200\n",
      " - 0s - loss: 539.0936\n",
      "Epoch 138/200\n",
      " - 0s - loss: 543.9665\n",
      "Epoch 139/200\n",
      " - 0s - loss: 540.2032\n",
      "Epoch 140/200\n",
      " - 0s - loss: 554.7345\n",
      "Epoch 141/200\n",
      " - 0s - loss: 535.4821\n",
      "Epoch 142/200\n",
      " - 0s - loss: 539.8004\n",
      "Epoch 143/200\n",
      " - 0s - loss: 535.6923\n",
      "Epoch 144/200\n",
      " - 0s - loss: 536.6488\n",
      "Epoch 145/200\n",
      " - 0s - loss: 540.3596\n",
      "Epoch 146/200\n",
      " - 0s - loss: 547.7705\n",
      "Epoch 147/200\n",
      " - 0s - loss: 546.7946\n",
      "Epoch 148/200\n",
      " - 0s - loss: 542.7995\n",
      "Epoch 149/200\n",
      " - 0s - loss: 538.0800\n",
      "Epoch 150/200\n",
      " - 0s - loss: 539.9116\n",
      "Epoch 151/200\n",
      " - 0s - loss: 538.2345\n",
      "Epoch 152/200\n",
      " - 0s - loss: 542.6563\n",
      "Epoch 153/200\n",
      " - 0s - loss: 542.0500\n",
      "Epoch 154/200\n",
      " - 0s - loss: 534.9969\n",
      "Epoch 155/200\n",
      " - 0s - loss: 541.8801\n",
      "Epoch 156/200\n",
      " - 0s - loss: 535.1742\n",
      "Epoch 157/200\n",
      " - 0s - loss: 543.3245\n",
      "Epoch 158/200\n",
      " - 0s - loss: 544.6594\n",
      "Epoch 159/200\n",
      " - 0s - loss: 536.2708\n",
      "Epoch 160/200\n",
      " - 0s - loss: 537.5012\n",
      "Epoch 161/200\n",
      " - 0s - loss: 536.8786\n",
      "Epoch 162/200\n",
      " - 0s - loss: 561.2330\n",
      "Epoch 163/200\n",
      " - 0s - loss: 548.3076\n",
      "Epoch 164/200\n",
      " - 0s - loss: 537.9429\n",
      "Epoch 165/200\n",
      " - 0s - loss: 537.0720\n",
      "Epoch 166/200\n",
      " - 0s - loss: 540.8805\n",
      "Epoch 167/200\n",
      " - 0s - loss: 535.9838\n",
      "Epoch 168/200\n",
      " - 0s - loss: 540.2042\n",
      "Epoch 169/200\n",
      " - 0s - loss: 537.2613\n",
      "Epoch 170/200\n",
      " - 0s - loss: 550.7933\n",
      "Epoch 171/200\n",
      " - 0s - loss: 541.4729\n",
      "Epoch 172/200\n",
      " - 0s - loss: 540.1929\n",
      "Epoch 173/200\n",
      " - 0s - loss: 547.6941\n",
      "Epoch 174/200\n",
      " - 0s - loss: 542.1946\n",
      "Epoch 175/200\n",
      " - 0s - loss: 537.1703\n",
      "Epoch 176/200\n",
      " - 0s - loss: 541.2273\n",
      "Epoch 177/200\n",
      " - 0s - loss: 533.9241\n",
      "Epoch 178/200\n",
      " - 0s - loss: 536.9766\n",
      "Epoch 179/200\n",
      " - 0s - loss: 538.5023\n",
      "Epoch 180/200\n",
      " - 0s - loss: 542.0446\n",
      "Epoch 181/200\n",
      " - 0s - loss: 555.3934\n",
      "Epoch 182/200\n",
      " - 0s - loss: 541.5840\n",
      "Epoch 183/200\n",
      " - 0s - loss: 535.3345\n",
      "Epoch 184/200\n",
      " - 0s - loss: 534.6165\n",
      "Epoch 185/200\n",
      " - 0s - loss: 556.3776\n",
      "Epoch 186/200\n",
      " - 0s - loss: 536.3121\n",
      "Epoch 187/200\n",
      " - 0s - loss: 536.3267\n",
      "Epoch 188/200\n",
      " - 0s - loss: 539.2417\n",
      "Epoch 189/200\n",
      " - 0s - loss: 541.8297\n",
      "Epoch 190/200\n",
      " - 0s - loss: 539.0216\n",
      "Epoch 191/200\n",
      " - 0s - loss: 537.5241\n",
      "Epoch 192/200\n",
      " - 0s - loss: 539.5370\n",
      "Epoch 193/200\n",
      " - 0s - loss: 546.4690\n",
      "Epoch 194/200\n",
      " - 0s - loss: 542.1810\n",
      "Epoch 195/200\n",
      " - 0s - loss: 535.3081\n",
      "Epoch 196/200\n",
      " - 0s - loss: 551.2705\n",
      "Epoch 197/200\n",
      " - 0s - loss: 543.7846\n",
      "Epoch 198/200\n",
      " - 0s - loss: 538.5891\n",
      "Epoch 199/200\n",
      " - 0s - loss: 539.1442\n",
      "Epoch 200/200\n",
      " - 0s - loss: 533.8352\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x7ff50844e780>"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "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,epochs=200,batch_size=2,verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Score 531.71 MSE (23.06 RMSE)\n",
      "Test Score 2355.06 MSE (48.53 RMSE)\n"
     ]
    }
   ],
   "source": [
    "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)))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainPredict = model.predict(trainX)\n",
    "testPredict = model.predict(testX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainPredictPlot = numpy.empty_like(dataset)\n",
    "trainPredictPlot[:,:] = numpy.nan\n",
    "trainPredictPlot[look_back:len(trainPredict)+look_back,:] = trainPredict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "testPredictPlot = numpy.empty_like(dataset)\n",
    "testPredictPlot[:,:] = numpy.nan\n",
    "testPredictPlot[len(trainPredict)+(look_back*2)+1:len(dataset)-1,:] = testPredict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD4CAYAAAAXUaZHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOy9eZhcV3mv+66a567q6uq5pZZagy15kGVhYww2g80QkpgYEjIc4iQQk0MOBy6HGxJyyYHcHC45GYBwA4kDMUMIxBDABmzwCMYGD/IkyxpbUqvnrq6pa57X+WPt3WpJPVRXlzWu93n0qGrtvdde24/121996xuElBKNRqPRXFhYzvYCNBqNRtN6tLhrNBrNBYgWd41Go7kA0eKu0Wg0FyBa3DUajeYCxHa2FwDQ0dEhBwcHz/YyNBqN5rzimWeeiUkpI4sdOyfEfXBwkN27d5/tZWg0Gs15hRDi+FLHtFtGo9FoLkC0uGs0Gs0FiBZ3jUajuQDR4q7RaDQXIFrcNRqN5gJEi7tGo9FcgGhx12g0mguQcyLOXaPRaC4UXhhLUa3X2dbnJl/JE3aHz8o6tLhrNBpNC/n/7tvPSCzPr930DPeP/JgfveNH2C32M74O7ZbRaDSaFpLKV5hOF9kXPU60EOXJqSfPyjq0uGs0Gk0LmStUABhNxQG49+i9Z2UdWtw1Go2mhaQNcY/lUwA8NPoQxWrxjK9Di7tGo9G0iGqtTq5c48r+NuqiQNARIV/N89Pxn57xtTQk7kKIoBDi20KIA0KI/UKI64QQ7UKIB4QQh42/Q8a5QgjxD0KIYSHEHiHEzpf3ETQajebcIF2sAvDWK3qwWAt4q5cTcUfOimumUcv9s8CPpJSXAFcC+4E/BR6SUm4GHjK+A7wF2Gz8uR34QktXrNFoNOcopkum3WsHa5GZlJWb19/MzyZ+RrVePaNrWVHchRBtwA3AlwCklGUpZQq4BfiKcdpXgLcZn28BvioVTwBBIURPy1eu0Wg05xjmZqrTXgEkpbKTAf8AlXqFbDl7RtfSiOW+AZgF7hRCPCeE+KIQwgt0SSmnjHOmgS7jcx8wtuD6cWNMo9FoLmjSRSXuVrvaQC2XnbhtXgCylXNP3G3ATuALUsqrgBwnXDAASCklIFdzYyHE7UKI3UKI3bOzs6u5VKPRaM5J0gXlehHWghqou7ELN3Buivs4MC6lNCPxv40S+xnT3WL8HTWOTwADC67vN8ZOQkp5h5Ryl5RyVySyaAtAjUajOa8w3TJYlLjLmhtRV+KeKWfO6FpWFHcp5TQwJoTYagy9AdgH3APcZozdBtxtfL4H+F0jauaVwNwC941Go9FcsJhumbrIA0rcZd0FQK6SO6NrabS2zPuBrwshHMBR4PdRL4a7hBDvBo4Dv2Gcey/wS8AwkDfO1Wg0mgueuUIFm0VQqikhlzU3GOJ+pi33hsRdSvk8sGuRQ29Y5FwJ/PEa16XRaDTnHelChTa3nUxF7SPKupta1Qmcmz53jUaj0TTAXKFCwG0nXU5jwQJ1B1VD3M+0W0aLu0aj0bSIdLFKwGUjXUrjtfsAC4US2Cy2c29DVaPRaDSNkTYs90wlQ8AZACBbruG3+8/JJCaNRqPRNIAp7ulSmoDDj90qyBSr+Bw+7XPXaDSa85V00dhQLSvL3ee0kSlW8Nm1uGs0Gs15iZRSbai61IZqwBHA77KTNS137ZbRaDSa849ipU6lJk9Y7g7Tcq9qy12j0WjOV8zSAwG3jXQ5jd/hx++ykSkZ4q4td41Gozn/MEsPeJySUq0075YxN1QzFR0KqdFoNC8b9brk+y9Mki9XOZo6SqFaaMm8ZqMOm60EcMJyNzZUc5UcKoH/zKDFXaPRXFR8f88k7//Gc3z9md3ces+t3HXwrpbMa7plrDZVy11Z7jayJWW512W9ZS+SRtDirtFoLhqqtTqfffAwAD+e/Co1WSNWiLVkbtMtY9Zy9zv8J22owpmtL6PFXaPRXDTc88IkR2M5LM5pDmYfA1pXrXEur8RdGuV+A07lc6/VJU6L0Y3pDG6qanHXaDQXBdVanc8+dJhtPQHckYewCRcRd6Rl1nS6qLowVVHi7nf48blU4V2LNMr+nsFNVS3uGo3mouDJYwmOx/O869VBLP4XGbS/iW5vd8us6XShgsdhJV9V8wUcAQKGuAtD3HPlM1cZUou7RqO5KJjNqCiWNr8SWFdtIz5760IU54xa7ulyGjixoQqqrjtoy12j0WhaTjJfVh8sym1SKbtbWhYgXVSlBzLlDC6rC4fVgc9pB6BecwDa567RaDQtJ5mvIARUUAJbLDlamjmqGnWcyE4F5i33WlW5ZXS0jEaj0bSYuXzZKOo1B0Ch6Gpp5mi6UD2prgyAz6nEvVxWf2tx12g0mhaTzFcIeuykSilAkCva8dv9FKoFqvXqmuefW1DL3bTcAy7llsmVJV67V7tlNBqNptUk82WCHgdzpTkcwke2WMfnUMlFa+1vKqUkli3R4XMSL8Zpd7UD4HVaAc5KTXct7hqN5qIgla8QMix3l8VPtlTFY1PivtZEpmypSqlaJ+x1kCgmCLvDANisFjwO64myv9py12g0mtaSzJcJeRykSim8NuU2sQsVorhWizqeVZE4Ia+NZDE5L+6gNlWzZ6EypBZ3jUZzUTBn+NznSnP47G0ACGnEn6/Rco9lVQy921VEIgm7Toi7z2kjU6rgc/h0EpNGo9G0kkqtTqZUJehWlnvAYYh73cgcXaPP3RR3i029JEyfO3Ciprv2uWs0Gk1rSRlFvUJeZbkHnUrc5zNH12y5K7eMsKmXxKluGVPcW1WkrBG0uGs0mguelJGd6nVBoVog5A4BCzJH12hRm5Z7WaoY+oVuGbOmu9/hX/MvhNWgxV2j0VzwJA3L3eFQtdY73EEAqhUnsPayALFsSfnzy0ngZMtd1XSv4LV7KdaKVOqVNd2rUbS4azSaCx7TcrcYjTQ6vconni9bcFgca45iiWfLKsa9EMdusc8354ATPnczselMhUNqcddoNBc8ps8di3KLdPuUZZ0pVlpSPEwlMDmIF+OE3WGEEPPH/C4b+XINt/XMNuxoSNyFECNCiBeFEM8LIXYbY+1CiAeEEIeNv0PGuBBC/IMQYlgIsUcIsfPlfACNRnPh8C+PHuVvf3yQh0cf5lNPfapl85oVIeuGuEc8IexW0bLkoli2TNjITl3obwfo9KuIHCuq3ky0EF3TvRplNZb766SUO6SUu4zvfwo8JKXcDDxkfAd4C7DZ+HM78IVWLVaj0Vy45EpVPvPgIX64d5JPP/NpvnXwWy2bO5mvYLcKijXlfgm6goa7pNKS5KJYtkTE5yRRSJzkbwfoC6mIHFFVm7hTuak13atR1uKWuQX4ivH5K8DbFox/VSqeAIJCiJ413Eej0VwEfP+FSXLlGun6IUbSI5TrZcq1ckvmTpl1ZYyKkEFncD5E0W/3r8lyL1ZqZIpV5ZYpnG659wWV5V4qKMt9KntuibsE7hdCPCOEuN0Y65JSmqucBrqMz33A2IJrx42xkxBC3C6E2C2E2D07O9vE0jUazYXEN54aBaDofnx+rFWhg6l8haDbTqqYwml14rK5TioLsJZQyEROvYDavXYSxcRJCUwAvUFluUfTkpAzdM5Z7q+WUu5EuVz+WAhxw8KDUkqJegE0jJTyDinlLinlrkgksppLNRrNBcbeiTleGJ8j0lZFePcQcioXRqs2HxfWlWkzEphUiOLak4vMGHePu0JVVk9zy3gcNkIeO5OpAt3ebiZzk80/yCpoSNyllBPG31Hgu8A1wIzpbjH+NncJJoCBBZf3G2MajUazKN98ehSnzcLlWw8jLDXevuk3gdY1t0iZdWXKcwSdKsbd77KTLlbwO/xruo8p7ja7muNUtwwo630iVaDX18t0drrpe62GFcVdCOEVQvjNz8Abgb3APcBtxmm3AXcbn+8BfteImnklMLfAfaPRaDSn8fPhODdsiZCo76VW7GZT22VA68TdtNxV6QFT3A3L3eEjV8lRq9eamnu+9IDVEHf34uI+mSrQ4+1hMjeJcna8vDRiuXcBjwkhXgCeAn4opfwR8CngZiHEYeAm4zvAvcBRYBj4F+B9LV+1RqO5oJhJFxkIecjXk9QrIWhRQS9QjTRSC7owmW6ZgBktYyQc5arN3cu03Gvi9KJhJn1BNxNJJe6FaoG50lxT91oNtpVOkFIeBa5cZDwOvGGRcQn8cUtWp9FoLniypSq5co3OgJNMOo6sXoKsKXFvRaGtQqVGuVZX0TKZky33bKl6QtzLufnep6shlinjdVjJVOPA4pZ7X9BNrlyjzd4JqHDIoCvY7CM1hM5Q1Wg0Z5VoughA2GchW51DVgPUjIJerbDczboyQbftNLdMXYLd4gFoOtY9niupBKZCHKuwzs+/EDPW3VJTG8VnYlNVi7tGozmrRDPKreFy5QGQVT+1qlHQqwU+96QRquhyVqjJ2rxbxm80r7ZKJe7NRuaYpQcSxQQhVwiLOF1WzXDISlndezr38m+qanHXaDRnlRnDcrfY0gDUKwGKZQs2i60llrtZV8ZmVy+PE+KuvNJmw45mXySxzImiYYv52wF6jUSmdNaJ0+pkMqstd41Gc4Eza1juNYsSd1n1kyvVWtZQ2qwrI4yKkKbbxOdU4l6rqV8Jzfr3TbdMophYNAwSoMPrxGGzMDVXpMfbc0YSmbS4azSas8pMuojLbiFXTaiBWmC+/nlL3DKGuEsjVHFhnDtAvdZ8TffxZJ54rkzYLzmWPkaXt2vR8ywWQW+bi3EjHPJMlCDQ4q7RaFYklS/z2OEY2XKWxycep1gttmzuaKZEp9/FbGEWq7DitbaRKbWu5+hMuojVIijU1cujy6MEOGC4ZaqGf7+ZDdV/fGQYu8WCaHuMTDnDb2z5jSXPnY91950Zy33FUEiNRnPxEk0X+djde3n44BjWzu/gCe2nUi/xiVd9gls339qSe8yki3QFnMwWZgm7w+RdDrLFKl6ftyU+96m5Il1+J7HCCAJBh6cDOGG5F0tWbMK2ast9LJHnW7vH+Y1rOvjOkb/ixv4buTxy+ZLn9wXd/PTQLG/y9hAvxinVSjitzuYfbAW05a7RaJbkx/tm+PFLM1x76Rz2tue51H89AtHSaI95yz0/S6e7E5/zRM/RVvjcZ9JFutpczORnCLvD2C1K1M0N1WypueJhn3v4MBaLINTzJOlymvftWD5fszfoJpopEXGrXw4vd8SMFneNRrMkY4k8DpuFm65Qgni1/78QdAaJF+Itu0c0XaLTsNwjnsh8clGrfO5Tc0V6DHHv9HTOj3scVpw2C4lcmZArRKKYaHjOeLbEfz47wW9f0893j36D1w+8nm3hbcte02eEQzql+uXwckfMaHHXaDRLMpbI0x9yM1ucRNZt5Asewu7wqoRwOXKlKtlSdd5yj7gj+Iyeoz67b81uGSkl03NFugIuZnIz8/52ACEEnQEnM+kine5OZvONlx4/MpujVpdctcFCppzh+r7rV7ymw68SswLWfj567UcZDAyu+nlWgxZ3jUazJGPJPAMhD5PZSaz1MIlchXZXO/Fiayx3M4Ep7LOQLCWV5W64ZbwO75rdMplSlXy5Nm+5LxR3gC6/i2imRIeng9lC4+I+mlAx8xaHeskN+AeWOx2Adq/yr1erHn7rkt+ix/fy9jDS4q7RaJZkNJ5noN3NeHYcJxFmsyXCrnDL3DJm6QEzOzXijhh11lVBr7V2Y5qZU/OHvCqO/dRQxYWWezQfbbha42gij0VACfVC6Pf3r3hNu0dZ7vFsa7pLrYQWd41GsyhzhQrpYpV17R4mMhMEbF2qEbQ73DLLfcaw3C12lcAU8UTwmR2SzIJea3DNTBni7nSpXwCnWu6dhuUe8USo1Cuky+mG5h1L5OlpczOVn8AqrHR7u1e8JuRV+xZm3P3LjRZ3jUazKGOG6yHsr5GpZGh3dBPPlgi7w+QquZbEupuWe93ITu30qGiZXLmGx+YF1taNadqYX9hUid3TxD3gJFOs0mZXmaXRfJRGGE3kWdfuYTwzTre3ez4CZzl8Tht2qyCRq6zmEZpGi7tGo1mU8aQSd7srCUCXp4d4rkzIqeqntGJTNZop4bSdyE7tcHfMhyjahIouWUvEjOmWqQj1DAujZUBZ7gBWqerNNLqpOi/u2fGGXDKgNnBDHsd8IbOXGy3uGo1mUcYSqhZL3Xpi07BWlziFqnneCr97NF2cD4O0Civtrvb5mi8WubaCXgBT6SLtXgeJkhLtU8W9K6A2OWVVPVMjm6qFco3ZTEntRWTG6fc1Ju6AWot2y2g0mrPJaCJPwGUjUVap8huCKiLEIv0ALfG7z6RLdBmlB8LuMBZhmc8cRa69G9PMXJFuIwwy4AjgsXtOOm5a7uWycgE1Iu5jxi+azjb166VRyx3QlrtGozn7jCXzDBibqW3ONvraVKOJekVtdLbEcs8YlruRnQrgM9wyZjemNVnuc0W6F0lgMjEt92RW4Lf7G/K5j8aVuDvdyo+/GnHXlrtGoznrjCVUjPtEdoJ+Xz8RnxLCkmHltsJyX1g0zKz5Ml+Kt6pCB9eyoTqTVuIezUcXrdjY5rbjsFmIZopEPBFihdiKc5ox7tKqnr+RGHeTdu8Jy11KSbHSXFPuRtDirtFoTqNel4wnC/Mx7n2+PjoMcZ/LSXx235o3VMvVOplilbDXQaqYmm90YW6omnXWm7XcS9Ua8VxZuWUWSWACI0vV7ySaLhFxRxqz3BN5vA4rc1VVG2Y1PveQ10GqUKFWl8wVKlzysR/xtV+MNHz9atDirtFoTmM2W6JUrdMfUl2D+vx9tLntWC1CNadwrz2RKVVQFmzQYydZSp7WRKNYsmATzXdjiqZVDH0kYCNeiC8q7oAS91VY7mMJ5a4az4zjd/jnOzs1QrvHjpQqh8D8BdAZcDV8/WrQ4q7RaE7DjHH3+wpU6hX6ff1YLIKw10EsU1ZZqmt0yySNeG+fu0alXiHkVD590+eeK9fWVILATGByu3JI5JLi3hVwMZNWiUyNZKmaYZBj2bFVWe2gLHeARK48L+7r2j3LXdI0Wtw1Gs1pmBEhVoeKDzdFrMPnJJZtjeWeMHzPVqO3adClLHevQ4m7WTysWbeMmcBks59IkFqMTr8qQRBxqyzVudLcknNKKefFfSIzsarNVFA+d1BZqqa4D2hx12g0Z4qJpIpxr1mUX73X1wtA2OcglivT7mpfs8/dTMO32JTImZa71SLma7qvRdzNBKaqNQWwZAu8zoBLZak6VJbqcuGQsxnDXdXuVBvNqxT3kOeE5T6WyBP2OubdUK1Gi7tGozmNWLaM32UjXVaWe4dbRbJEfE5iGVU8LFVKUak3n0pvWu5SGL1NDcsdmC8e5rU3343paCyH32VjKnccgaDX27voeZ1+tXFrWyRLtVipcSyWo1Qr8cTUExw3rG2fNzfvrloN7ae4ZV4uqx20uGs0mkWIZUt0+JwkSgnsFjteuwp/7PArt4wZ2ZIsJpu+hxkSWEWJu2m5g/K7z3dIatLn/szxBDvXhXh+9nk2hzbjc/gWPa/L2NA0s1SjhRMRM3/5g3380md/xj8+9wX+8P4/5NFje9W5jnEAtrZvXdWaThX3l8vfDlrcNZrzlr+//yA/3DMFj3wSHviLls6dyJVVTHYxScgVQggBQNjroFSt47UrIV6L3z2Zr+B32shUlI/7dMu92rTlPpevcGgmy851AZ6PPs9VnVcteW6nkchUKqkXmBkxE80U+fbucQq1HN888B8APD31PB0+B1OFYazCytbQ6sTdZbficViZzZSYTBVfVnHXDbI1mvOQWl3yTz89isUCbwx9A3t6FPquhm23tGT+RK7MQLuHRDFB2BWeHzdj3a21tZcgSObLKu67lMIqrPjt/vljZqs9v93flM/9mVG1H9AdSZGfyLOzc+eS53YZJQhSp2Sp3vn4CJV6HXvoKQq1LHaLnWOZfVzW90r2JfYxFBzCZVt9GGPI4+ClyTlqdaktd41GczKTqQLlWp1apYwlrVwEfP8DkJ5qyfyxbJkO3wnL3aTD8E/Xa8rKXcumaiJXJuSxkyqlCDqD878OQFnu2eLqujFJKanVJYVqgceOjmGzCIrWIwDLWu5Bjx2H1cKMEes+m58lU6zwb08c583bO3B3PE5QbGNn5y4yHOWy3gD7YvtW7Jm6FO1eBy9OqF8r54TPXQhhFUI8J4T4gfF9gxDiSSHEsBDiP4QQDmPcaXwfNo4PvjxL12guXo4b9U0+sNOGlTpPDbwbKkW498NrnrtelyTzyi2TKCZOEvew4TOulNdeX2ah5b7wHsBJ0TKNdmN679ee4Y/+7Rk+9dSn+N7Mh7m0z83e+PN0e7uXbWln9lKdnivS6elkOjfNN58aI1OsctnWY0jrHOXYjXQ5t2BxTtMZniNZSjYt7iGvg2KlDsC68Dkg7sAHgP0Lvv818Gkp5SYgCbzbGH83kDTGP22cp9FoWsixuPJD/86mKgDfzmyHK98Jxx9f89zpokqPb/c6SRQT85unABHDck/nLbht7oYyOpcikSvT7lG/DszsVJOA204qX5l31aRKqRXn2z+d5oF9Mzw/PUzFEscXeZxnZ55d1mo36Qu6mUgW6PP1MZmb5BdH42zt8pOoHsZp8TETXUcu3YsQksOFBwHYHt7exFOrLFUAu1XQ/TJlp0KD4i6E6AfeCnzR+C6A1wPfNk75CvA24/MtxneM428QC39vaTSaNTMSy+GyW2grjALwXK4dguugkIRyfk1zx4wen353nUK1cJK4h70OhFDnhF3hVTWVPpVkbmnLvTfoplCpEbCrxKPJ7OSK85m9SY/GVc2XA8XvEC1El/W3m/SHPEykCvT6ekkUE4wmUqwPq2YcPd4+QPDYS8rKfmjsh1iFlS2hLat53HnMLNX+kAer5eWTxkYt988AfwLUje9hICWlrBrfx4E+43MfMAZgHJ8zzj8JIcTtQojdQojds7PN/w+i0VyMHI/nGAx7sSSPUbQFGM46qBqJRqQn1jS3GX/ucKiXxEJxt1kthL1O1WTD09m05V6q1siVa/MROada7v0h1YXJUlX3Xknc8+Uq+XKNHQNBpCVLNbsZUGUEGrHc+0NuptNFOt3KfTOenVTljrMTDAXXYbUIYnN2nLKTTCXT9GYqnGiU/XL626EBcRdC/DIQlVI+08obSynvkFLuklLuikQirZxao7ngORbLsT7sgcQRct71SAkJm/HvaG58TXMnckbT6lMyR01Uoa0SHe6OhtvSnUoqr5Kfgu4TG6oLMcW9VFSJRZO55cXdtNrf+YpehC1P0LKZ27bfRre3m03BTSuupy/kRkpwGnZoRcTpCzqYzE6yrq2fS7qVe6jXraz1Zv3tcMJyX9fubnqORmjEcr8e+FUhxAjwTZQ75rNAUAhhhlL2A6a5MAEMABjH24DWtErXaDTU6pKxRIHBDi/Ej1INbgBgGpVFulbLPW5mjlqM5KJTXCadgdVVUVwM89eBx1WmJmuni3tQWbWzafVyWclyj2XVC8nlUi+kP7z+Cj6w8wPce+u9WC3WFddjvkzqZfWsFnuSNv+Jomk7BtT6rohcAaxN3MPz4n6WLXcp5Z9JKfullIPAbwIPSyl/B3gEeIdx2m3A3cbne4zvGMcfliuVWdNoNA1jhkFuDNpgbgxbZAiA42Wj9OzcGsXdsIKrIgNwUpw7MF//vMPdQbaSJV9ZvY8/eUrRsFNfIAG3Db/TNu8HX0nczTVb7WqjeV1bF0II7BZ7Q+sZCCmhzeY8WIUNYU+eVDTtNZsjOG0Wbr3kDfR4e7iu57oGn/R0wkauwMst7mtJYvoI8E0hxF8BzwFfMsa/BHxNCDEMJFAvBI1G0yJGjEiZrY44IPH0KFfBZLYO3k5Ir9Uto+rKZMrK5XKa5e53qcqQLvVLIVaIsc6+bnX3OKVo2KmWuxCCvpCb8WSe3vW9HE4eXnY+03KXVvVrY+E+QSN0t7mwCJhIFfFaIxTtSSpCORz6/H1c19vFMx+7GZ/Txv3vuH9Vc5/K1etD/NXbLuP1lyxeyKxVrErcpZQ/AX5ifD4KXLPIOUXg11uwNo1GswgjRoz7eqESltxdW/E7Y6p+eVvf2i33XJmwEePusDjm68qYdAac1CU4hRLk2cIs6wKrE3fTcq8L9aI69QUCKppkPJnn5u19PDr+KFJKlgq8M11JVanK+4bdp8VwLIvdaqE74GI8WcBeb8fhmiNanJwvOCaEaFn1RqtF8F9eub4lcy2HzlDVaM4zToRBjqmB8EZ6gi4mUwUI9K3d554tnZTAdKqgmlUUqalCW82EQyaNDdUKyvVzquUOyg8+nizQ6+2lVCstW+ogli0ZdWpUPPxqLXd1Pw/jqQL1SgiLPclEZoIubxd2a2OunXMNLe4azXnGSMwIg0wcBXc7uEP0tLkNy71/zZZ7Ilcm7Ds9gckkYtRiqZdVBEksv/pN1USuTMBlI1NWafiLW+5usqUqQYdyX0xkl36ueLZM2Hfi14bPvngFyOXoD6lEpnw+QM2S5sjckVWX9D2X0OKu0ZxnjMRPhEESVpupvUEXU3OG5V7OQHHpbkIrYbplksXkouJuWu7ZggO7xX5SidxGMcsbJEtJ7BY7Htvpm4tmBAtVJfzLbarGsiXCPifxYpx2d/uS7pvl6Au5mZorkM6ql9aBxAH6fH0rXHXuosVdozmPmA+DDHshcQzaNwLQHXATy5apmDVUmrTe63VJMndCeBezqM0SBLOZMh3ujqYt96DHyE51nu76AeUmASgXlctmJcu9w+cgXoyfFt3TKP0hN3UJtZJ65rqs0+fX4q7RaM4A8WyJcq1Of9AJmSllqQM9QeUqiVuNRKYm/e7pYoVqXS7rlnHZrbS57UQzJSLuSJM+d/UCSRVTJ9VxX0hfUFnusYzyyU9ll654Gc8pyz1RWHzNjWC+TOqVEy807ZbRaDRnhGhGhfz1OQpQr4JfWeq9bUoIJ+qG1dpklqoZdWLWlVnMcgczS7WoLPcmEpmSuQqhBZb7YgQ9drwOqwqH9PUykVv8hVWrSxK5Mh1ew3JfZaSMifkykVU/NiM/c7U9UiSJSngAACAASURBVM8ltLhrNC8DUkqQEuJHWjrvTFo1fe61Gu3t/N3ACct9rOIHYWnacjczR+320+vKLERlqZZU/fOmLXc7yVJySctdCGGEQxrVGpfwuSfzZeqS+Q3VZi33nqALIcBqsdLjVS9N7XPXaDTz3PfiFNd+8iFy++6Hz+2E8d0tm9u03DswSuAa4j5vuacNa75Jn3s8a9aVUfHnS4q730U0rdwyc6W5Zeut1+uSg9MZ6rLOdw9/l1QhR75cI+ixEy/EFw2DNFkYDjmZnWSxZHczO9XjqlCtV5sWd6fNSpffRW/QRa+vF6fVOd8Y/HxEi7tG02KeH0sRzZQ4st+otbf3P1s2t2m5B6tGzLch7m6HlaDHfiJipsksVdMtg3Xp5CJQbplZo3gYLB/r/uD+Gd70mUf5xguP8hc//wv+7ok7AbC5j5Mup7m84/Ilr+0LuZkw3DKnxrp/+5lxoukid+77Z6zeQ9gdas3NumUALunxc2l3gGt7ruVVva/CIs5fidQ9VDWaFjM5pwR4euwIVwDsuxve+L/AsnahiGZUgpEtP6MGfCdS2Hva3EyljCzVqT2rmjdTVslECcMKrqBCKZd2y7go1+q4LUr8Z/OzS7owDkdVSYB7XlJr+vHo3bR5Psx49VHcNjc3r795yXX1h9yki1WCjhN13TvcHYzEcnz4Wy/wW9eGuTf9VRwd6xC2XcuuuRE+91tXGdmou5qe41xBi7tG02KmUgUAqslx9ds4PQGTz0L/2gUjmi6qOPPMtEpgsjnnj/W0udSLpacPDt6nfP4rxHv//p1PsT7sJeb9AiPpEa4QH8fvtDGdPwYs55ZR9xV1laW63KbqeFL999gzM4ItDAWmedW2KA+OPsDN62/GY1+6gFavsclpral1TOemuSJyBU8dU71bHzr2JDIssbpHiZePA6cXOlsNftf5mY26GOfvbw6N5hxlaq7I+rCHbmKkgtvBYoeXvtuSuaOZEp0BlxJ3/8l9QXvajESmtn6oFiG/fPNqKSVPHkvw5Z+PsG/2CMfTx7l79J/Ysi7FnXvv5Lqe606rK2NiinutohJ+lnPLjCfztHsdSGscS92PrLl5qfJFspUsb9v0tiWvU8+kxL1SVi+R6ZzqsvSkIe5zHAQpEEJy33FVmHYtbpkLCS3uGk0LqdUl0+kiv3R5D/2WBPtq/TD0eth3j7Kk18hMukiX34hx959cVbA36CaVr1DyGKK/gt89W1Ldi4RQ4izqTixtTzLl/hwd7g7++oal2x93Gr0/CwUXFmFZtmnHeLLAdUNh/P4M5WIYb/la5soJ+nx9XN119bJr7DWigFJZG26bm+m8EvenRuK8YjCE1XOMWmEQUW3nQOIAArHsBu3FhBZ3jaaFRDNFanVJf8BGB0mem/NS2vrLMDcKk8+tae5aXRLLlukMOCE7s6jlDhCzmLHuy0fMzKRVZMx7b+wDS4ly/LUMeDdRqZf4zOs+s+RmKpyw3GPZKh2ujiUt93pdMpEs0B9y43SlkOUQtwy9HYHglqFbVtyw7PS7sFoE03Mlur3dTOemmUwVGEsUeP22NqyuCaq5DXhrVwIq2clm0d5m0D53jaalTKbUZuoGZwYLkrFaO3sdO7kaYOIZ6Fu5WfNSxHMlanVJl9+hxN13suXesyCRqQ9WjHWPGpE3W/skzMAfXreL9+z6GMliksG2wWWv9TpVM42puQLd3m6mcotnj0YzKqO2N+ggMxvn+sHX8d9veDW3ZL/B5uDmFZ/ZahF0+Z1MzhXoDipxf3pEuWQCwQkQklp+A13tIbI8sqbN1AsNbblrNC1kak5tHvZZVMjepAyzP+cDhx9ih9Y0d9SwtPsd+ZOyU01MF8bxklf5+VfIUp3JKHEXdlUD/YahIdqcbSsKu8lAu4exRJ4+fx/jmdPvJaVkOKaE2ONRce6/vO0yfE4b28PbcVgdDd2nJ6iigEzL/cljCXxOG9HKPqzCSq2wnkHPZbQ527S/fQHactdoWsiUYblH6spNkbR1Mjybg47NMHtwTXNHDTHusZ6cwGTSZfjBp9JlCPSsaLmbbpmaUGGPEffqGtWva/dwOJrhqiv7uX/kfir1CpWq4K/vO0C0EKUc+jovzu4Hy4ex2lVGbTMZnz1tLvZOzPEabzexQownx6PsGgzxXPRZtndsZ90rhnjN5gi/2vaXTZX6vVDR4q7RtJDJuQJehxW3sfHn7ljHkdksRLbC0Z+saW7Tco/Ik0sPmLjsVsJeh5HItHJd95l0EZ/TRrqiSvZGPKsU97CHhw9G6fcNUJM1fnH8CJ/4zjQj2ZdwD3wFa7lMXdawuo9RFjWApqos9gXd3L9vhi5PFxLJ0eQkt+zYyZ0TL/Kube/iQ1dfYZzZs+w8FxvaLaPRtJCpVJGeoBuRngBXG31dnQxHs9CxRUW4rKHOumlpt9VOzk5diOrIZCQyrRAtE02XVI2YfBSv3btk2ONSDLR7KFfreK0qwehvHvo5yXyFK7btQUor79v8OQQ2/KHjRAtTWIWVLs/q+4b2tLmM+6hsWIstRUd7nGq9ypWRK1c938WCFneNpoVMzRVU1Ep6AgL9bOr0MTVXpBDcpE6ILd/oeTmimSLtXgf2vNEcw3e6UKqOTGYJgimo15ecT4VVuojmo6t2yYByywDIsvJzH58b482XdVOxTWApbeDYZBvu+kZsniNMZCfo9nY3FcnSEzy5aYfFMUfJOgLAZeHLVj3fxYIWd42mhUykiqp07NwYtPUxFFHW8HGLUTp2DX73mXTJyE6dOi071aS3zXWi3V69ArmluyTNZIp0BpzMFmbp9HSuej2muKdzbuwWO0VmGeq0M5oepc+7kV8cjVPPb6RkHVtTVyOzKFq5pBKmIsEih1L7ibgjdHlX/0vgYkGLu0bTIkrVGrFsSYUkzk1Am7LcAfYX2sHqgFjz4j6bKRrZqafHuJv0BN1kilUKHsNls4TfXUrJTLpEV8Cw3FfpbwflCxcCxpMlQo4uLI4EPn8cieSqrm0cj+dJxNcBkqNzR+n19a76HuqZ1EZxIgPU3AQDWfbG9nJZh7bal0OLu0bTImbmjFBFv4RCAgJ9rA97sVkEh2NFaB+C2ebDIU+y3P2LW6xmItMsRqnaJfzuc4UK5WqdiM/BbH6WTvfqLXeHzUJvm5uxRB4nESz2BBWrut8bN10FQK2wDptQIY/NWu5hrwOHzcLTx5PUKkGkfZqR9IgW9xXQ4q7RtIhJI8Z9vc3YNG3rx261sD7sUZuqkS1NW+71umQ2W6JriexUEzORaaxuJPMsYbmbm7MBb4VyvdyU5Q4w0O5mNJGnVm7H6kwwmj2Cx+bhNRu2EPTYQdoYCigRblbchRD0trl45EAUWWljorAf0P72ldDirtG0CDOBqVcYqfhGf9OhiE+FQ3ZsheQIVIoNzVet1dWG6KH7iaXS1OqSdY7cotmpJqblPlZwgc21ZKy7WRfe7lTleJsV93VGIlMmEwBLgd0zu9kc2ozNauXaDeoFc033NcDauhr1tLnJl2tQCyJRm8TbO7Y3Pd/FgI5z12hahFl6oKNmlL9tU5uomzp9PHwgSjW8GZusQ+IIdC0vTD/aO837v/Es/7Bzhre8+EHytg283vJ23vbsXcp3f8kvL3pdd5tqFTeZLqmXy4IsVbMK5OZuK1/c/9dYfT0Im7L0m3HLgBL3aKaEQ/hxeuFw8jC/vuXXAbhlRx9HZ3P89rZbKcr4msTY9LtH3F0kgQH/AG3OtqbnuxjQlrtG0yImUwWCHjuO3CQgIKA2EDd1+qjWJZP2derEBiJmnh1NUqlJnnv2SQAClVn+1fG3OGs5uO0H0L94NUW71ULE51Q15dv6TrLc79s7zW/e8QTffDLKocxuHKGfUzXa9TXvllERM9XSiSJjW0NbAfily3t44EM30h/o4S+u+wuc1tOjexrFjJjZEFTWv3bJrIwWd42mRUykCgyEPJAaVT5xI1TRjJg5UDFcKQ3Euo/Ecmzq9PHOoQopEWD4HQ/Cqz8E735gSWE36Qm6VTjkgizVTLHCJ77/EgB37Z6g33YjNu9hXkqo7kjNxLnDiXDIeuVEwa4t7Vuamms5TMv9sq71gHbJNIIWd42mRYwnCyrGPTUKwXXz4xs6VKz7kWQd2gYgPrziXMfjeQbDHoasswT7tnLN5ZfCTf8TwkMrXts737SjD7LTUKvyd/cfIpop8QfXb+B4PM+xY8ryvfvI3QQcAVw2V1PPbIq70+Im5FTWeyPVHlfLlf1BAi4bb7v0lfzKxl/hTYNvavk9LjRWFHchhEsI8ZQQ4gUhxEtCiE8Y4xuEEE8KIYaFEP8hhIp3EkI4je/DxvHBl/cRNJqzj5SS8WSe/pAbUsdPEne/y07E7+RYLKvEeQVxl1JyPJFjfdgLiWPQvnFVa1FZqkVkoA9knePHj/DVX4zwX65dz5+8eSshj534nAe/vJxqvdpUApNJu9eB12FlS5efAf8Afb4+fI7WF++6rK+NPR9/E5s72/nkaz5Jt/f00guak2nEci8Br5dSXgnsAN4shHgl8NfAp6WUm4Ak8G7j/HcDSWP808Z5Gs05wdMjCZK5Mvzi8/DS91o2bzxXplipM9BmV66QBeIOyno/FstBeLMS92W6MkUzJYqVOhtDNpXpukpxXx/2kC/XiPXcCL97N8/MCuoSfu/6QVx2K2/fqTZ6h1xvAFiTuAsheOP2bt58WTfvvfK9fPDqDzY9l6a1rBgtI6WUQNb4ajf+SOD1wG8b418BPg58AbjF+AzwbeD/F0IIYx6N5qxRqtb4nX95kl3rg3w98SlErQp9V0NwYM1zm02gh1wZkLXTxH1jh5cH98/Ajk1QSkNuFnyLi+pILAfAZkcCkKsWd9PHf6gQILLptRw8uB+H1cJ6w4Xym9es44uPHeOy0LUk8uvY2La6+U/l0+/cYd55TfNoWktDPnchhFUI8TwQBR4AjgApKWXVOGUcMINY+4AxAOP4HHBaBX0hxO1CiN1CiN2zs0v3X9RoWsVYIk+5Vmf/0RFEcQ4qObjvT1oy94Qh7gMW4//lRSz3WLZMzj+oBpZxzRxP5AEYFKpsMKENq1qLKe7DUWWTHYlmGezwYLNa5o//87uu5rbrhrjrV+7iQ1d/aFXza84PGhJ3KWVNSrkD6AeuAS5Z642llHdIKXdJKXdFIs3t1Gs0q+FYTInmVV5VMre67jVw8F7Y/4M1zz2eVHN31pcWd4DjGPVVlhP3eA6bRRAuG2GMq7TcO/1O/E6bSpxCibwp+CZv2t5Nd5sLr92L3Wpf1fya84NVRctIKVPAI8B1QFAIYbp1+gEzoHYCGAAwjrcB8ZasVqNZA8diSuw+eq2qdfKltveprNHHPr3muceTBdrcdtw545+BkcBkstGoDnmw1KaSkJYR95F4nr6QG2vyGDjbwLO6vqBCCIY6fQxHsxQrNUYTeTZFdIeii41GomUiQoig8dkN3AzsR4n8O4zTbgPuNj7fY3zHOP6w9rdrzgWOxXK0ex1ssk5TxcqDM34Yeh1E9y9b97wRxpP5E2GQC2LcTQbaPVgEHIsVlSUeW1rcR+N5FSmTPAbtG0CIVa9nKKLE/Xg8T13CUKcW94uNRiz3HuARIcQe4GngASnlD4CPAB8SQgyjfOpfMs7/EhA2xj8E/Gnrl63RrJ5jsZxyj8SHSTp7ORjNIzu2Kt/7Cl2LVmIiVVg0DNLEabPSH/JwNJaD8KYlLXcpJSPxHINhDySOrtolY7Kp00c0U+K5UdWSb0hb7hcdjUTL7AGuWmT8KMr/fup4Efj1lqxOo2khx2I5XrM5ArEjFP0bSM9VSXk3EgJVEmARUW4EFeNe4NWbInBkFAZO+2cBLAiHvHQTHPox1GtgsQKq3EAqX+ba6Lf4dO3bTPo/q34FbL+1qTWZPvYfvzSNEFrcL0Z0hqrmoiBXqjKTLrEh7Ib4EawdKmzvkDSCvGYPND13Ml8hX66pGPf06THuJqa4y/CQ6pKUGgXgqWMJfuuOJ3j4a5/C+/Cfc5P1OX513wehXl2T5Q7w+HCcvqAbt8Pa3MNpzlu0uGsuCkbiKnb8Um8OqgV8/Srga3/KBt5OiDYv7makzEZ3RgnyEuI+FPGSL9dIulV9FOLD7J2Y491ffpp3+p7nL+138mDtKv535Z0EY8+qc5oU94GQG4fVQrlWPy1SRnNxoMVdc1FwzEgM2mhRseP+3q20ue0cimYhsnVNlrsZ477OYpT6XdJyVyJ7pGakzseH+djde/E6bXy08xfU2gb5M+v/4Av1X6W27VYQFuWfbwKb1TIffqkjZS5OtLhrzk3qdRjfvWya/mowsz57a2rjVIQ3s7nTx/BMFiKXKJ97k/cys1O76kYz6uD6Rc/bYIRDHs66wNWGjA1zeCbLm7Z34UqPYO+/is//3vX86Zsvxfr2O+CPHgdf8zkgQ51e428t7hcjWtw15xQ/2jvFmz/zKOWf/A188Q0w+kRL5j0ay9EdcOFIHQObG/w9bO7ycSiaQUa2QjkD6clVzyulZCo6wxtd+/AcuU8NBhbvONQTcOG0WTgWVxEzldlDZEtV1occqqlG+wZeMdjOe28cAqsdurat5ZHnLXbtlrk40Z2YNOcMk6kCf/LtPawvHcL+s/+tBmf2wvrr1jz3wjBIwkNgsbCp008qP8acfxNBgNn9qkzuChQrNR7YN8MLTz3CKye+ykfkMzhFBQ4Ag68B++Llcy0WsaCA2CY48jMAtjhTqh7NKssMrMRrL+nkkYOzbOsJtHRezfmBFnfNOUG9Lvnwt15AVvJ8xv6PFJxhPPUcxA61ZP6RWI63XN4DY8PQrWqZbzYs2sP1Pl4ByjWz6aZl55FS8muf/zn7p9I86vp/CVtzHOl9B6Gdb6NnyyvAe1oZpZPY0OHl4EwGBjfh2PMfuCgxaJlRB9tbK+4714X4/vtf3dI5NecP2i2jOSe4a/cYPz8S51+vPMiQZYrvrv9z6NjSUEu6lUjlyyTzFYZCdpVkZGxSbu4yOiSlHeAJN7SpOpspsX8qzf91Qx8DTON9zX9j27v/iZ6r3ryisIMS99F4nlpINd0YFDN0VafUwRZb7pqLGy3umnOCe/dOMxTxsqv6HBOWHh4qbVNRLC2w3A9MZwC40jmlQhW7lOXeHXDhc9o4HF2wqboCh41KizeGUwikWuMq2NDhpVqXzNhV7ZmrvDHs6eNgdaqyBRpNi9DirjnrlKo1njoW58ZNIcTI4xz1v0KVq+3YApkpKM6taf59k2kAtnBcDXRfDqgCW5s6fRyayShxj+5fMWLm8Ix6UWyQRrmCzktXtRazgNjhmuqneoVrFpIjEBoEi/7nqGkd+v8mzVnn2eMpipU6bwlNQDlDqvt6xpJ5yiGjF2cDDaWXY/9Umg6fg0DqANg9JyUGXdLt58B0Btl1mWqikTq+7FyHo1kCLhuBzDBYbKtOMjJj3YdTkhnCbLZOG630tEtG01q0uGvOOo8Px7BaBFeUnwMEtk03IiWMWowOSWv0u++bSnNpTwCmX4Su7fP1XAC29QZI5SvE/YZ7ZfrFZec6HM2yucuPiB1UvvtV1kIPeey0ue3sn0ozXOumvz5hWO5a3DWtRYu75qzz2HCMHQNBnKM/g96rWN+nwhH3l9pV7fNY8+JeqdU5PJNlW7dfCbfhkjG51AgTfLHSpzJCp/cuO99wNKuibGYPrNrfDsoVtDHi5dFDsxyT3XRmD6iqlNpy17QYLe6as8pcocKe8RSvHXTD+NOw8bVsjHgRAoZjRWgfgtnmN1WPzGYp1+pcHcxCaW5+M9Xkkm4/AHujFWWJL2O5x7MlErkyW8M2ZW1HVudvN9nQ4SWaKXFM9mAxO1Vqy13TYrS4a84qTxyNU5fwJu+wimTZ+FpcdisDIQ/Ds1mIbFmT5b5/Sm2mbreqCox0X3HScb/Lzvqwh/3TaWXVLyPuZqTM5e4YyHpTljuoZtkAR+SC6BhtuWtajBZ3TcM8uG+G0Xge7v5jePRvWjLno4dm8TisDCUeVZudA9cCKmX+SDSr2uAlR6BSbGr+fZNpHDYL3fnDgFg0pf/S7oCKqOm6DOZGoZCcP1ap1fn3J0fJpxNs/t5beaPlaYbkmDoYaa6VsLmpGrUbewqIpmvJazRLocVd0xAjsRy3f20337v3+/Dcv8FL31vznMVKjR/smeItm71YX/pPuOzW+dT9TZ0+jsZy1Du2KCs5caSpe+yfyrC1y481ule5XRze087Z1hvgeCJPIbxdDSzwu3/32Qk++t0XefQ/Pk04vZ9P2v+VYPw5EFZVxqAJzGqN1vb1YLFD28Bpbfk0mrWixV3TEHf87Ch1CVePfVUNxA5Brdrw9dNzReLZEjz1L3D//wMvfJOHn3mJuUKF94WfU5uKV//+/PlDES/lap1pu2HRRvc3vthaFWLDSCnZN5VWtVWm95y2mWpyaU8AKeGQGDQWq1wzUkru/PkIVlHn0vFvMip6CIkM4ukvKmFvUpAHOzwADHT41Tzh5mq2azTLoWvLaFYkminy7WfGGbJMc13pcWR4IyJxVPX4jGxZ8fpCucbr/vYnFCsV9rs+iosyANdaOrg29Ek2jt4FXZdD39Xz12ztVlEse8rd9FqdMPkcXP6ORec3eXw4xr89cZzPXn4Ux3ffQ+pX7iSRc/KKYBr2jsLVv7foddt6jXulnFzp7VTFylAdkvZPpfnna2Ks3xPlfaX/zjt7prkxflfT/nYAj8PGW6/o4Q2XdELn58HubnoujWYptOWuWZEvPz5CpVbnb/t+SgUbsVf/pToQ3dfQ9ccTOQqVGrdd5sRFmY9Vfo+vXfpP2Gp5/rX2UcT0i3D1bSDE/DWX9vhxWC08N5GHnitg4pkV7/PAvhnu2zvN40+oMsH2ez9Iv5jlrQc+As4AbP+1Ra/rbXPR5razbypjbKruUc/98xGCHjs3pb9Lwd3Nj+uv4Oj2D6gX0dDrG3r2pfjH397JrTv7of/qNZf21WgWQ4u7ZlkyxQpfe+I477zEzpXx+7irdiMvOS4HRMPdi0Ziqg3dbVvrAAT6t/Gx5wK8p/YRPPW82ki94jdOusZps3Jpb4DnxlLQtwsmn4daZdn7TKRU04zo2DBFiwdbNc9Dno/iju2FW+9YMptUCMGlPX72TRkRM9EDTMRS/Pilaf7r5WAd+Smu627nC++6hre/aiv818dg1x809OwazdlCi7tmWb7x1CiZYpUP+R9CyCp31N7KgXhNhe41arkb/Uu766oZxvt//U3cvK2LK175RsQf3Ae/+e/gajvtuqsGgrw4Pket92qoFmDmpWXvM5Es8KqhMJucSfZVe/lW++04azl47Udh61uWvXZ7bxsHptJUu6+EeoXnn/4ZdQlvbx8BQGy7hTdu7ybgWl1GqkZzttDirlmSUrXGlx47xk0bnHQe+Dpi+69R9q/j0HRGJfA02FR6JJ6n3evAnT4OVgeu9gH+5Xd38bFf3ga9O2DodYtet2MgSKFS46jTSBaa2L3sfSbnCgxFfFzuS2MNredX3/Nx+G/PwI1/suIadwwEKVXrHHGoe5VGniLidxJOvQiuYNORMRrN2UKLu2ZJ7n5ukpl0iT/v/IVqQ3f9B9jS5edQNAOdl6iuRtXSivOMJnKsD3vUBmxw/Um1XZZjx0AQgN1zfvB0wPjSfvdcqUoqX6G3zYkjO8mV27fT5nVAx6aTfPkr3euphBt83fjjL7BjIIiY2K02ehuYQ6M5l9DirlmUel3yT48e4YoeD4PDX4WhN0DPlWzp8jMczVKPXKpaw8WHV5xrJJZnMOw1qh82Hva3Puwh5LHz/Ngc9O9a1nKfNPztG9x5qJVWnRTUH3IT9jp4fmyOcs9ONpUP8IoeuwrB7H/FqubSaM4FtLhrFmXPxBxHZ3O8/0qByEXhincCsKXLR7FSZ8oxqE5cIf68VK0xOVdgXcitLPdViLsQgisHgjxvbqrGDkEhtei55mbqeltcDbT1N3wf8147BoI8P5Zk3LONDZYZbrC+CEj1YtFozjO0uGsW5ZDRlOIKl9Hf04jr3tylCm3tK3epLM0VxH08WUBK2OrPG9UPV5ews2MgyKFohkLXDjUw+eyi55ni3i1jaqBtYNHzVrrXkdkcjxUGAdh07OvqwIL4e43mfEGLu2ZRjkSzOKwWOgojaqBDNc4wm0ofipfVJuMK4m5GygzZDNFtQtylhD1S9T1dyu8+mSpgswjaytNqYJWWO8COdcrv/vlDfuoIbGM/V+v1tK96Lo3mbKPFXbMow9EsGzq8WOOHoG3dfE0Wv8tOX9Ctqi12X6784PX6kvOYMe59NRUGudrqh1f0K8F9MSZVK7qZxeutT6aKdLe5sKQnVMKSO7iq+yy813TRzoxzUA32aZeM5vxkRXEXQgwIIR4RQuwTQrwkhPiAMd4uhHhACHHY+DtkjAshxD8IIYaFEHuEEDtf7ofQtJ4js1k2dfpUud1TSgzsGgzx+HCM2tDNkJ1Z0lUCynL3O214c6PKjbPKjc52r4OugFMlGHVdtnis+8hjtE8/Tm/QDamxpqx2gDa3nSGjx2kuYvxvqzdTNecpjVjuVeB/SCm3Aa8E/lgIsQ34U+AhKeVm4CHjO8BbgM3Gn9uBL7R81ZqXlWKlxmgiz1CHG2LDquzuAt64rZtkvsJzrmuUYB/44ZJzHU/kWd/hQSSPQXBg1W3pAC7pDnBgKqPEPXEEWc7x8IEZ/ugz/86hv70JvvxW/u/kx9kYkDA31pS/3WTHQAgA99B1amDgmqbn0mjOJisWDpNSTgFTxueMEGI/0AfcArzWOO0rwE+AjxjjX5VSSuAJIURQCNFjzKNpIVJK4rky1sQw/ue/hE3UwR2C1/4Z2BxNzzsSz1GXcLkvrTJDDX+7yY1bIzisFn50pMSuweuVuN/0Pxed63g8r6oyrjJSZiGX9Pj5xZE41cg2bLLO33/9Hj53MMD3E+xBvQAAF2hJREFUHX9Dl5ihsOP3cL/wZV5Te1qJ+xqs7Vt29BLLlui6/iZYt0klWWk05yGr8rkLIQaBq4Anga4Fgj0NdBmf+4CxBZeNG2OnznW7EGK3EGL37OzsKpetAfjUfQfY9VcPcu8dH0M8+6/U990Dj/09HFzakm6EYaPj0Bar4Sc/pQKiz2njVZvCPLB/Brn1rcp1Ezs93r0aH+GG1Pd4R+k7EDvctLhf2h2gXKsz5lDXR4ef4T07fVxuOcod1bfyec8fMSHD7ErcoxptBJu33G/YEuErf3ANNodzycxZjeZ8oGFxF0L4gP8EPiilTC88ZljpcjU3llLeIaXcJaXcFYlEVnOpxuD+fTNc0d/GW30HeLi2k/983cPg74Xnv7H6ySpFmFLVEIejWYSA3orRmq7j9PK2N2/r4ng8z7GO16qBgz9ESskjB6L82dce4amPvwrb567kE7Y7ed3o56Cchf7mXByX9Kjwyz25IFWrh0vEKL/TeRyAY227+MovRvl+7VV0JYwkpzW4ZTSaC4WGxF0IYUcJ+9ellN8xhmeEED3G8R4gaoxPAAv/dfUbY5oWMp7McyyW411bJMHiBAfcO/nWs1Nw5Tth+EHITDc0z70vTvHn330R+cj/gn++AaL7OTKboy/oxp44DJ4weMOnXXfTpeqH2r2jVtWX9MAP+dYz4/z+l58mdPR7XMNL/KTnPfzzjm+T+uAx+KixtibY2OHDbhUcmMkx6dzAZdYx1qeeAmcbW6+6gXSxyt21V524QIu7RtNQtIwAvgTsl1L+/YJD9wC3GZ9vA+5eMP67RtTMK4E57W9vPY8Pq7jxG+wqeqT9yjfx1EiCyfW/psoC7LlrxTlqdckn793Pd548RG33lwEJj/8Dw1EzUubQolY7QFfAxY6BIA/sm4FL3gpjT/HEiwfpD7n58OYohAZ57Xv/jv/T3p3HR1WeCxz/PZOVbGSHhBAgIRrCjqwCyi6goCIqVqxaW7W1atW2iq21vbefq21vqdoit1oVrYp1ARdkUUFFdsImOwkkZBGyEZKwZJt57x/ngIFsEBIyE57v5zMfZs6Zc3jOC/PMmfe8533uu2E8oaHh4BvQ5GP19XaQGBXEnkOlbKmMI8UrC0fG19BtJFP6WYl8t4nHFWGP6mniaBml2pJzOXMfDtwBjBGRrfZjMvAsMF5E0oBx9muAxcABIB14GfhZ84etVqUXERXsR3TBWgiOZfSVwxGB/2S2sy4obpsPpuGesuW788gpPslN3qvwriyF+GGY7e9yoiCT7pGBUFB7GGRN43pEsy2nhOK4MYCh3cEvGZ4QiiNrNXQd2azH2yMmhPUZR9h4MpZA1zGrkHXiaBKigujdqT2hAb44Bt0DwTEQ3LFZ/26lPFGjyd0Ys8oYI8aYPsaYfvZjsTGmyBgz1hiTZIwZZ4w5Yr/fGGMeMMYkGmN6G2ManqdVnTeXy7AmvZCRieFIxteQMIrYsACGJ0ayYEsOrj63WXOt2xWF6jNvTSaxIX48Evwl37q6sW/4XwH4IZ8yxHc/lB+t98wdYHRyNABfFHegKiCaYc5UrokshPIS6HZVsx0vQHLHYE5UOtnjqtHlkmBd8Hx6Sgq/n9IThtwHj+w851knlWrL9A5VD7TncBlFxyuZHFVgjQ6xR3Xc2L8T2UdOsjvsauuN6cvr3cfew2Ws2V/ErOQ8Ik5m8DaT+e1XpXzjP4p7vJcwfu0d4O0PCaPq3UdKTAgdQ/xZsbeQA6FXcpVjG4Oqt1orm/nMPTnGqnNaHGQPy2zf+fTom4Fdw7mhfydrWl5N7EoBmtzdmstld6usmwt531c9OtXfPsRpz7PSzUrmI5MiAVj1nVjFNDK/qXffr6w6QDtvmFTwLwjqQPjgW9mQcYTnqqaxL2Is5trZ8OjuBut7igijk6P4Jq2QpZV9CZGTBG99CSKSICTmQg69lh4drREz/ZK6WBdwe0zVOdaVakCjNzGp1lF0rIJRf/mK56fEMWbpE3DZRPjBfygrr+L9TTlMD0sneP3frO6PYGvkSnSIP4lRgaw9UMR93UbCljehurLWDU1bs4/y3qYc5nTfjHf2Zpj2Lx5O6cXEfl3oGdseL8ft5xzn6Mujmb8hm5dy4nmgnQ/exwugx5RmbQuAqGA/fjnhMsaldIDI5XqGrlQj9MzdTW3JOkpZRTUb166wFqR9TnnxIX7yRiqBhVt5tupZCE+Em18/Y7thiRFszDhCdZeRUHUCcs+cRbHa6eLJBdtJCTrOpPx/Wt0uvafj5+1Fn7hQvBzndzY8vHskvl4OjtOOI1H2naHN3CUD1q+En49JIrljiPVlpcldqQZpcm8hTpehyunCufrv8OkvrUf2hnPefntuCQDeedusBcbJoree59sD3/FmyBy8g6LgjoW1pqMdlhDJ8UonO317A1Kra+a11ZlkHsrn9dBXkOpKuHb2BXVvBPp5MyTBisG/703gE9jsF1OVUudPu2VaQH5pOWNnf01oRS7f+P2WSq9AfKmGjK/hZ+vB0fh36o7cEiKDfOlVkcHRdvGYdmH0zF/E291LCcg5DLctq7Nfe6idaFflOOnbsRdkrDxdIHr+hixeXrKOxSF/I6JwP1z/j2Yp/PzA6O70jw8j5MrJMOAma34bpVSr0jP3FvDVvgLKyqt58nLrLtHbHc9QNeUf1k1Bexef0z6255YwMimKK3wOklrZhZdKBtPDkU3fnLdhwJ0QP7TO7SKC/Li8QzDrDhRB16usXwtV5cz5Mp1ZC7bzZsg/6OLKRma8Df1+0CzHOzQhgkfHX2b9AtDErpRb0OTeAlanFxIZ5MfEgD1UBHRkY1kEH1YMgtAusPq5Rm8uyi8tJ7+sgoFRTiJdBawrj+ftE4NwOXyRgAgY9/sGtx+WGEFqZjFVXUaAs4Ky9DX89bO93Jrsw2UVO5GrfwWXT2y+A1ZKuR1N7s3MGMPq9EJGJoYiGSvxvWwMPWPbM3fVQVzDHoScjXBwTYP7ONXfPtDHnhzLJ4kbh/XCceNcuPXfjZZ9G5oQwckqJ9skBbx8yU/9EJeBn3a29kf38Rd+oEopt6bJvZntzSuj8Fglk6MK4WQxkjCG+65O5EDBcZa3Gw+BUbDuxQb3sT23BBHoVrUPgNm/+CG/uy4Fek+HLlc2uC3AiKRI/LwdfLL3GCRNIPLgIsL8HcQfWQtBHazyeEqpNk2TezNblWbfYGTsW/8TrmZyr450DPFnwbdF1nj1zFUN1h3dkVtCQmQgvvnbITyR9qEROM5jiGKQnzdje0Tz6fZDOFOm0b66iLs75eA4sAISx+rNP0pdAjS5N7M1+4tIiAok5LtVEN0TgqLx9nIwJCGczVnFmM6DrTlbimoXtzhle26JVaz5u60Q279JcUztG0vhsUreOJLMMePPHWWvWFMVdB/b1ENTSnmQSze5V520RpJkra+zilCTdul0se5AEaMSgiFr3RmVfAbEh5FXWkF+aF9rQU7dY97zy8rJK61ggv9uKM1pcpm3UZdHE+znzZ+XZ/OZayBhpbsBOT3ZllKqbbvkkvuKPXk89u421s35MbwyHl6dAHMGN0uC35p9lBOVTia2zwFnxRk38wyIt4YIbiiNAP/2kL2+1vbHKqp57N1tTHWs5pptD1rzw/S9rUmx+Pt4MaFnR05WOdkUbJ+tdxpQZ+ENpVTbc0kl9yqniyc+2M7Gnfvof3QZnziHcnDCK+DwhjUvXPD+N2QcAaC3y57kq/P3ZeWSY4Lx93GwObvEKjeXvfGMbfPLypnxzzWkZLzOC75zcHQeDD9aCoGRTY5nar9YAIJ6jLOm7u19S5P3pZTyLJdUcl+y4zD5ZRXM67MLP6p42XELf8lMgP4zreIW51ia7rSi/bDp9dPj1jcdLKZ7dBDtDqdCdMoZN/T4eDnoExfK5qyjVtIv2AMnjwKQUXic6S+u4ubCuczyfgtSboCZC6Bd6AUd74jukTw0NomZw7vDzzfA0PsvaH9KKc9xSSX3easzSAz3pVvmO5A4hmFDr2Tx9kPkpvwYXNWNDlE85e/L07jz1Q1UL34CPnkIPv8dLqeL1MwjDIoPsfry67iDdEB8GLu+K6Ey5grAQG4qew+XcdPcNYwoX8mdjsUw5H6Y/hr4+F/w8Xo5hEfHX0bn8KaXuFNKeaZLJrlvyz7K5qyjPJW4Hyn7Dobczz3Du+HtcDD3Wyf0vBE2vmpVEWpAeZWTl1YeYNe+NGT/F5jgWFjzAkeWPUtpeTVjwougwipZd7YB8aFUOQ3b6Q7igOyNvL42k4oqJ0+kFIFfe7jmmXOae0YppRrSdrNIeekZY8lfX5NJoK8XI4oXQlg36D6e6BB/pg3oxHupORzrdy9UlsHuRQ3udtnOw5RVVPOHbjvxwsV/hf4R0+cWIjf8iYGyhwHssd5Y15l7F6ubJvVQtdVtk7OBrVlHrUm3CrdZFzw1sSulmkGbyiRVThf/XptJzsE0+GsyvHQ1pH9B8bEKFm0/xH0pVXjnrIOBd59OojMGx1NR7WLZ0U7QPh52fdTg3/HB5lw6tfdnkvNL8oJ78VqaPysSZ1HiHcnv/eYTXpQKwbFWGbizRAb5ER8ewKaDxVYx6qz1ZOQdYVCsL+TvhLiBLdEsSqlLUJtK7h9v/Y6nPtrJ+y8/C1XHcZ0ohjdvIvu9X1NZ7eJ27y/B4QP9vq801DeuPbHt/Vmy8zCkTIX9K05f6DxbXmk5q9IKuP/yMiR/F5Ej7yYxKpA/fp7F3+U2epGG7PrYOmuv5y7QKxMjWLu/iOrEcUjVcQayi+EBOWBcEDeoRdpFKXXpaVPJff6GLBLC/bir3UpWuvrw88iXMf1m0ufgPO6J3kdE+gdWCbgawwtFhEm9Y1i5r5DjSVPAVQX7lta5/4VbcnEZuN6sAC8/vHrfxJOTe5BReJxXyoZQGJwMxllnf/spo5OjKauoJlV6Ue3wZ4xjCz1ce62Vna5o1vZQSl262kxy35dXRurBYn7dPZfQqnzKe9/B4l1HeLr6LtJdsTxZ9kfrtv+Bd9fadnLvjlQ6XXxREgchcbDzw1rvqXa6mL8hi/FxTkJ2vQN9boZ2YYxJjmZYQgQGB8VX/TcERjd4i/8Iuyzd8vRSdrfrxzU+Wwks2AphXS9oTLtSStXUZpL7/A1Z+Ho5GHNiMQRGM/aGuxjcLZw3UvN5Uh7CIUB4Qp31Pft3DqNDiB+LdxyGlOth/3LrgmwNC7bkcrDoBE+FLrXOzq/6FWCd+T8zrTc/HZVIwhXj4VdpDVY3OlWWbsWefJZU9CXW5EHaF9BJ+9uVUs2nTST38ionCzbnclfSCXz3fwb9Z+Ll48vsW/oSGuBDn0FXIzM/gJvn1dkX7nAIk3rF8NXeAk4mXQfOStjzKcXHK6muduL66EEcSx9nWoc8Ome8Z930FNb19PZdIwN5fGLyOReXHpMczf6C4yw81staUH1SL6YqpZpVm6ihOvvzfQSX5/JY3jNWt8jgewGICwtg1eNjCPDxAkdKg/u4tk8M89ZksuRoZ6aFdaVy85sMXxDGqKAsXjzxBtOB6SWfgpcvjPzlBcU7JjmaP3yyi0NEcCIsmYDiPXoxVSnVrDz+zH3uV/tZuHITHwf/GV9TAXcsPKNwdJCf9znNhT6wSxjx4QF8sCUX+v4An6zVhFUe5trKZZwwfjwa+jym7wwY+zSE1h7meD66RASSEBWIt0Pw6TMN/EO1gIZSqll59Jn7Oxuy+NPSPfw24QhhRceQmQuhQ8Nn6PUREaYN6MTzy9M4NP4GYr76H56IWMXkyjVkd53Cw1NvRiLuarbY7x2ZwJ7DZfhcNQGG/AS8/Zpt30op5dHJvUdMCNP6d+LO6ZOQipmN1hZtzE0D4njuizQe+6yYB50pXHf8fQRD/ISfQURgM0VtmTE4/vsXFxi3UkqdrdFuGRF5VUTyRWRHjWXhIvK5iKTZf4bZy0VEXhCRdBH5VkQGtGTwfTuHMvvWfvh4OZolQXYOD2Bwt3DW7C9iifdoBGN1l8S26GEopVSzO5c+93nAxLOWPQEsN8YkAcvt1wCTgCT7cS8wt3nCvHimD4gDIKj/NIi8DEY8ojVHlVIep9FuGWPMShHpetbi64FR9vPXga+Ax+3lbxhjDLBOREJFJMYYc6i5Am5pU/rGsjevjDuvSoCQjY1voJRSbqipfe4daiTsw0AH+3knILvG+3LsZbWSu4jci3V2T3x8/NmrW007Xy+euq5pF2WVUspdXPBQSPss3TRhu5eMMQONMQOjoqIuNAyllFI1NDW554lIDID9Z769PBeoOQg8zl6mlFLqImpqcv8YuNN+fifwUY3lP7RHzQwFSjypv10ppdqKRvvcRWQ+1sXTSBHJAZ4GngXeFZF7gIPALfbbFwOTgXTgBFB7CkallFIt7lxGy9xWz6pa89ra/e8PXGhQSimlLozHzy2jlFKqNk3uSinVBmlyV0qpNkisbvJWDkKkAOvCbFNEAoXNGE5L8pRYPSVO0FhbgqfECZ4Ta0vF2cUYU+eNQm6R3C+EiKQaYzyijJGnxOopcYLG2hI8JU7wnFhbI07tllFKqTZIk7tSSrVBbSG5v9TaAZwHT4nVU+IEjbUleEqc4DmxXvQ4Pb7PXSmlVG1t4cxdKaXUWTS5K6VUG+TRyV1EJorIXrtm6xONb3FxiEhnEflSRHaJyE4RedheXmftWXcgIl4iskVEFtmvu4nIertt/yMivm4QY6iIvC8ie0Rkt4gMc9c2FZFH7H/7HSIyX0T83aVN3bku8jnE+Rf73/9bEVkoIqE11s2y49wrItdcrDjri7XGusdExIhIpP36orSpxyZ3EfEC5mDVbU0BbhMRdymhVA08ZoxJAYYCD9ix1Vd71h08DOyu8fpPwN+MMd2BYuCeVonqTM8DS40xyUBfrHjdrk1FpBPwEDDQGNML8AJm4D5tOg/PqIs8j9pxfg70Msb0AfYBswDsz9cMoKe9zYt2jrhY5lE7VkSkMzAByKqx+OK0qTHGIx/AMGBZjdezgFmtHVc9sX4EjAf2AjH2shhgb2vHZscSh/WBHgMsAgTrbjrvutq6lWJsD2RgDwKosdzt2pTvy02GY828ugi4xp3aFOgK7GisHYF/ArfV9b7WiPOsdTcCb9nPz/j8A8uAYa3Zpvay97FORDKByIvZph575k799Vrdil1cvD+wnvprz7a254BfAy77dQRw1BhTbb92h7btBhQAr9ndR/8SkUDcsE2NMbnA/2KdrR0CSoBNuF+b1nS+dZHdwY+AJfZzt4tTRK4Hco0x285adVFi9eTk7vZEJAj4APiFMaa05jpjfWW3+jhUEbkOyDfGbGrtWBrhDQwA5hpj+gPHOasLxo3aNAy4HusLKRYIpI6f7O7KXdqxISLyG6zuz7daO5a6iEgA8CTwu9aKwZOTu1vXaxURH6zE/pYxZoG9uL7as61pODBVRDKBd7C6Zp4HQkXkVDEXd2jbHCDHGLPefv0+VrJ3xzYdB2QYYwqMMVXAAqx2drc2rclj6iKLyF3AdcDt9hcRuF+ciVhf7tvsz1YcsFlEOnKRYvXk5L4RSLJHIPhiXUz5uJVjAqyr4cArwG5jzOwaq+qrPdtqjDGzjDFxxpiuWG24whhzO/AlMN1+W6vHaow5DGSLyOX2orHALtywTbG6Y4aKSID9f+FUrG7VpmfxiLrIIjIRqwtxqjHmRI1VHwMzRMRPRLphXazc0BoxAhhjthtjoo0xXe3PVg4wwP5/fHHa9GJecGiBCxiTsa6Y7wd+09rx1IhrBNbP2m+BrfZjMlZf9nIgDfgCCG/tWM+KexSwyH6egPXhSAfeA/zcIL5+QKrdrh8CYe7apsAfgD3ADuDfgJ+7tCkwH+taQBVW0rmnvnbEurg+x/6MbccaAdSacaZj9Vef+lz9X433/8aOcy8wqbXb9Kz1mXx/QfWitKlOP6CUUm2QJ3fLKKWUqocmd6WUaoM0uSulVBukyV0ppdogTe5KKdUGaXJXSqk2SJO7Ukq1Qf8Pxr+usUPL5bgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(dataset)\n",
    "plt.plot(trainPredictPlot)\n",
    "plt.plot(testPredictPlot)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "# MLP using the window method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "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": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "numpy.random.seed(7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "look_back = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainX,trainY = create_dataset(train,look_back)\n",
    "testX,testY = create_dataset(test,look_back)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/200\n",
      " - 0s - loss: 88109.7045\n",
      "Epoch 2/200\n",
      " - 0s - loss: 61204.0175\n",
      "Epoch 3/200\n",
      " - 0s - loss: 45815.9598\n",
      "Epoch 4/200\n",
      " - 0s - loss: 30762.1070\n",
      "Epoch 5/200\n",
      " - 0s - loss: 16606.9325\n",
      "Epoch 6/200\n",
      " - 0s - loss: 6863.6326\n",
      "Epoch 7/200\n",
      " - 0s - loss: 2538.2402\n",
      "Epoch 8/200\n",
      " - 0s - loss: 1375.1444\n",
      "Epoch 9/200\n",
      " - 0s - loss: 1190.9321\n",
      "Epoch 10/200\n",
      " - 0s - loss: 1165.6686\n",
      "Epoch 11/200\n",
      " - 0s - loss: 1165.8791\n",
      "Epoch 12/200\n",
      " - 0s - loss: 1156.1145\n",
      "Epoch 13/200\n",
      " - 0s - loss: 1146.5375\n",
      "Epoch 14/200\n",
      " - 0s - loss: 1152.3337\n",
      "Epoch 15/200\n",
      " - 0s - loss: 1148.5767\n",
      "Epoch 16/200\n",
      " - 0s - loss: 1138.7788\n",
      "Epoch 17/200\n",
      " - 0s - loss: 1148.2691\n",
      "Epoch 18/200\n",
      " - 0s - loss: 1131.1921\n",
      "Epoch 19/200\n",
      " - 0s - loss: 1143.7408\n",
      "Epoch 20/200\n",
      " - 0s - loss: 1133.2917\n",
      "Epoch 21/200\n",
      " - 0s - loss: 1125.7523\n",
      "Epoch 22/200\n",
      " - 0s - loss: 1128.9423\n",
      "Epoch 23/200\n",
      " - 0s - loss: 1120.3594\n",
      "Epoch 24/200\n",
      " - 0s - loss: 1113.8314\n",
      "Epoch 25/200\n",
      " - 0s - loss: 1120.7597\n",
      "Epoch 26/200\n",
      " - 0s - loss: 1122.4006\n",
      "Epoch 27/200\n",
      " - 0s - loss: 1122.6049\n",
      "Epoch 28/200\n",
      " - 0s - loss: 1116.2528\n",
      "Epoch 29/200\n",
      " - 0s - loss: 1123.1763\n",
      "Epoch 30/200\n",
      " - 0s - loss: 1104.0541\n",
      "Epoch 31/200\n",
      " - 0s - loss: 1101.1488\n",
      "Epoch 32/200\n",
      " - 0s - loss: 1100.0832\n",
      "Epoch 33/200\n",
      " - 0s - loss: 1110.2963\n",
      "Epoch 34/200\n",
      " - 0s - loss: 1101.7075\n",
      "Epoch 35/200\n",
      " - 0s - loss: 1106.7604\n",
      "Epoch 36/200\n",
      " - 0s - loss: 1081.7767\n",
      "Epoch 37/200\n",
      " - 0s - loss: 1083.2561\n",
      "Epoch 38/200\n",
      " - 0s - loss: 1091.4647\n",
      "Epoch 39/200\n",
      " - 0s - loss: 1074.4731\n",
      "Epoch 40/200\n",
      " - 0s - loss: 1071.3613\n",
      "Epoch 41/200\n",
      " - 0s - loss: 1074.1452\n",
      "Epoch 42/200\n",
      " - 0s - loss: 1078.8272\n",
      "Epoch 43/200\n",
      " - 0s - loss: 1074.2227\n",
      "Epoch 44/200\n",
      " - 0s - loss: 1070.6428\n",
      "Epoch 45/200\n",
      " - 0s - loss: 1063.0799\n",
      "Epoch 46/200\n",
      " - 0s - loss: 1052.6108\n",
      "Epoch 47/200\n",
      " - 0s - loss: 1058.0311\n",
      "Epoch 48/200\n",
      " - 0s - loss: 1044.1198\n",
      "Epoch 49/200\n",
      " - 0s - loss: 1053.3180\n",
      "Epoch 50/200\n",
      " - 0s - loss: 1054.2190\n",
      "Epoch 51/200\n",
      " - 0s - loss: 1029.1791\n",
      "Epoch 52/200\n",
      " - 0s - loss: 1043.9421\n",
      "Epoch 53/200\n",
      " - 0s - loss: 1045.6889\n",
      "Epoch 54/200\n",
      " - 0s - loss: 1025.7823\n",
      "Epoch 55/200\n",
      " - 0s - loss: 1039.5492\n",
      "Epoch 56/200\n",
      " - 0s - loss: 1017.3459\n",
      "Epoch 57/200\n",
      " - 0s - loss: 1044.4078\n",
      "Epoch 58/200\n",
      " - 0s - loss: 1024.8244\n",
      "Epoch 59/200\n",
      " - 0s - loss: 1010.5569\n",
      "Epoch 60/200\n",
      " - 0s - loss: 1005.7719\n",
      "Epoch 61/200\n",
      " - 0s - loss: 1008.8488\n",
      "Epoch 62/200\n",
      " - 0s - loss: 1005.1824\n",
      "Epoch 63/200\n",
      " - 0s - loss: 993.4732\n",
      "Epoch 64/200\n",
      " - 0s - loss: 993.6050\n",
      "Epoch 65/200\n",
      " - 0s - loss: 991.9991\n",
      "Epoch 66/200\n",
      " - 0s - loss: 985.0057\n",
      "Epoch 67/200\n",
      " - 0s - loss: 990.2446\n",
      "Epoch 68/200\n",
      " - 0s - loss: 978.4799\n",
      "Epoch 69/200\n",
      " - 0s - loss: 988.9839\n",
      "Epoch 70/200\n",
      " - 0s - loss: 987.1740\n",
      "Epoch 71/200\n",
      " - 0s - loss: 972.9336\n",
      "Epoch 72/200\n",
      " - 0s - loss: 968.8991\n",
      "Epoch 73/200\n",
      " - 0s - loss: 978.9653\n",
      "Epoch 74/200\n",
      " - 0s - loss: 959.2847\n",
      "Epoch 75/200\n",
      " - 0s - loss: 958.7106\n",
      "Epoch 76/200\n",
      " - 0s - loss: 972.2824\n",
      "Epoch 77/200\n",
      " - 0s - loss: 972.5175\n",
      "Epoch 78/200\n",
      " - 0s - loss: 956.7038\n",
      "Epoch 79/200\n",
      " - 0s - loss: 943.6586\n",
      "Epoch 80/200\n",
      " - 0s - loss: 948.6939\n",
      "Epoch 81/200\n",
      " - 0s - loss: 933.4097\n",
      "Epoch 82/200\n",
      " - 0s - loss: 946.8171\n",
      "Epoch 83/200\n",
      " - 0s - loss: 940.1210\n",
      "Epoch 84/200\n",
      " - 0s - loss: 936.7137\n",
      "Epoch 85/200\n",
      " - 0s - loss: 932.9696\n",
      "Epoch 86/200\n",
      " - 0s - loss: 930.2683\n",
      "Epoch 87/200\n",
      " - 0s - loss: 922.1439\n",
      "Epoch 88/200\n",
      " - 0s - loss: 923.7484\n",
      "Epoch 89/200\n",
      " - 0s - loss: 905.5510\n",
      "Epoch 90/200\n",
      " - 0s - loss: 903.5081\n",
      "Epoch 91/200\n",
      " - 0s - loss: 915.5458\n",
      "Epoch 92/200\n",
      " - 0s - loss: 954.7363\n",
      "Epoch 93/200\n",
      " - 0s - loss: 900.9466\n",
      "Epoch 94/200\n",
      " - 0s - loss: 896.3934\n",
      "Epoch 95/200\n",
      " - 0s - loss: 887.3830\n",
      "Epoch 96/200\n",
      " - 0s - loss: 891.2874\n",
      "Epoch 97/200\n",
      " - 0s - loss: 905.6440\n",
      "Epoch 98/200\n",
      " - 0s - loss: 913.2689\n",
      "Epoch 99/200\n",
      " - 0s - loss: 881.0864\n",
      "Epoch 100/200\n",
      " - 0s - loss: 889.7467\n",
      "Epoch 101/200\n",
      " - 0s - loss: 872.4957\n",
      "Epoch 102/200\n",
      " - 0s - loss: 873.4792\n",
      "Epoch 103/200\n",
      " - 0s - loss: 864.4079\n",
      "Epoch 104/200\n",
      " - 0s - loss: 866.5847\n",
      "Epoch 105/200\n",
      " - 0s - loss: 861.3070\n",
      "Epoch 106/200\n",
      " - 0s - loss: 854.8246\n",
      "Epoch 107/200\n",
      " - 0s - loss: 856.6695\n",
      "Epoch 108/200\n",
      " - 0s - loss: 864.3212\n",
      "Epoch 109/200\n",
      " - 0s - loss: 855.9216\n",
      "Epoch 110/200\n",
      " - 0s - loss: 866.7273\n",
      "Epoch 111/200\n",
      " - 0s - loss: 844.5831\n",
      "Epoch 112/200\n",
      " - 0s - loss: 841.3421\n",
      "Epoch 113/200\n",
      " - 0s - loss: 845.9477\n",
      "Epoch 114/200\n",
      " - 0s - loss: 842.4117\n",
      "Epoch 115/200\n",
      " - 0s - loss: 841.0921\n",
      "Epoch 116/200\n",
      " - 0s - loss: 823.3601\n",
      "Epoch 117/200\n",
      " - 0s - loss: 835.5037\n",
      "Epoch 118/200\n",
      " - 0s - loss: 819.8666\n",
      "Epoch 119/200\n",
      " - 0s - loss: 829.6376\n",
      "Epoch 120/200\n",
      " - 0s - loss: 820.9195\n",
      "Epoch 121/200\n",
      " - 0s - loss: 808.3402\n",
      "Epoch 122/200\n",
      " - 0s - loss: 839.0756\n",
      "Epoch 123/200\n",
      " - 0s - loss: 822.3472\n",
      "Epoch 124/200\n",
      " - 0s - loss: 817.0709\n",
      "Epoch 125/200\n",
      " - 0s - loss: 821.9765\n",
      "Epoch 126/200\n",
      " - 0s - loss: 814.8395\n",
      "Epoch 127/200\n",
      " - 0s - loss: 801.0941\n",
      "Epoch 128/200\n",
      " - 0s - loss: 795.8277\n",
      "Epoch 129/200\n",
      " - 0s - loss: 812.9776\n",
      "Epoch 130/200\n",
      " - 0s - loss: 787.8123\n",
      "Epoch 131/200\n",
      " - 0s - loss: 783.4977\n",
      "Epoch 132/200\n",
      " - 0s - loss: 780.4476\n",
      "Epoch 133/200\n",
      " - 0s - loss: 787.3004\n",
      "Epoch 134/200\n",
      " - 0s - loss: 786.6275\n",
      "Epoch 135/200\n",
      " - 0s - loss: 772.2812\n",
      "Epoch 136/200\n",
      " - 0s - loss: 768.7214\n",
      "Epoch 137/200\n",
      " - 0s - loss: 765.8321\n",
      "Epoch 138/200\n",
      " - 0s - loss: 776.0052\n",
      "Epoch 139/200\n",
      " - 0s - loss: 766.8218\n",
      "Epoch 140/200\n",
      " - 0s - loss: 774.9826\n",
      "Epoch 141/200\n",
      " - 0s - loss: 754.1395\n",
      "Epoch 142/200\n",
      " - 0s - loss: 768.0853\n",
      "Epoch 143/200\n",
      " - 0s - loss: 750.0253\n",
      "Epoch 144/200\n",
      " - 0s - loss: 756.3095\n",
      "Epoch 145/200\n",
      " - 0s - loss: 757.3155\n",
      "Epoch 146/200\n",
      " - 0s - loss: 755.6854\n",
      "Epoch 147/200\n",
      " - 0s - loss: 740.8942\n",
      "Epoch 148/200\n",
      " - 0s - loss: 740.7913\n",
      "Epoch 149/200\n",
      " - 0s - loss: 745.1263\n",
      "Epoch 150/200\n",
      " - 0s - loss: 740.8292\n",
      "Epoch 151/200\n",
      " - 0s - loss: 727.6746\n",
      "Epoch 152/200\n",
      " - 0s - loss: 731.0017\n",
      "Epoch 153/200\n",
      " - 0s - loss: 727.9053\n",
      "Epoch 154/200\n",
      " - 0s - loss: 739.8590\n",
      "Epoch 155/200\n",
      " - 0s - loss: 731.2719\n",
      "Epoch 156/200\n",
      " - 0s - loss: 724.5324\n",
      "Epoch 157/200\n",
      " - 0s - loss: 717.3151\n",
      "Epoch 158/200\n",
      " - 0s - loss: 724.7290\n",
      "Epoch 159/200\n",
      " - 0s - loss: 704.2126\n",
      "Epoch 160/200\n",
      " - 0s - loss: 706.3153\n",
      "Epoch 161/200\n",
      " - 0s - loss: 720.1406\n",
      "Epoch 162/200\n",
      " - 0s - loss: 709.7170\n",
      "Epoch 163/200\n",
      " - 0s - loss: 726.5037\n",
      "Epoch 164/200\n",
      " - 0s - loss: 694.4019\n",
      "Epoch 165/200\n",
      " - 0s - loss: 712.6890\n",
      "Epoch 166/200\n",
      " - 0s - loss: 713.3198\n",
      "Epoch 167/200\n",
      " - 0s - loss: 699.0713\n",
      "Epoch 168/200\n",
      " - 0s - loss: 695.9354\n",
      "Epoch 169/200\n",
      " - 0s - loss: 693.9168\n",
      "Epoch 170/200\n",
      " - 0s - loss: 730.6917\n",
      "Epoch 171/200\n",
      " - 0s - loss: 688.6809\n",
      "Epoch 172/200\n",
      " - 0s - loss: 694.1907\n",
      "Epoch 173/200\n",
      " - 0s - loss: 681.8024\n",
      "Epoch 174/200\n",
      " - 0s - loss: 676.8689\n",
      "Epoch 175/200\n",
      " - 0s - loss: 690.0550\n",
      "Epoch 176/200\n",
      " - 0s - loss: 671.0315\n",
      "Epoch 177/200\n",
      " - 0s - loss: 675.8267\n",
      "Epoch 178/200\n",
      " - 0s - loss: 673.7959\n",
      "Epoch 179/200\n",
      " - 0s - loss: 670.8289\n",
      "Epoch 180/200\n",
      " - 0s - loss: 671.5434\n",
      "Epoch 181/200\n",
      " - 0s - loss: 673.8727\n",
      "Epoch 182/200\n",
      " - 0s - loss: 657.0309\n",
      "Epoch 183/200\n",
      " - 0s - loss: 657.8520\n",
      "Epoch 184/200\n",
      " - 0s - loss: 648.5559\n",
      "Epoch 185/200\n",
      " - 0s - loss: 648.9846\n",
      "Epoch 186/200\n",
      " - 0s - loss: 662.2291\n",
      "Epoch 187/200\n",
      " - 0s - loss: 669.4110\n",
      "Epoch 188/200\n",
      " - 0s - loss: 649.4986\n",
      "Epoch 189/200\n",
      " - 0s - loss: 636.8045\n",
      "Epoch 190/200\n",
      " - 0s - loss: 633.0625\n",
      "Epoch 191/200\n",
      " - 0s - loss: 643.1732\n",
      "Epoch 192/200\n",
      " - 0s - loss: 638.7997\n",
      "Epoch 193/200\n",
      " - 0s - loss: 650.5050\n",
      "Epoch 194/200\n",
      " - 0s - loss: 633.2100\n",
      "Epoch 195/200\n",
      " - 0s - loss: 629.0484\n",
      "Epoch 196/200\n",
      " - 0s - loss: 629.2481\n",
      "Epoch 197/200\n",
      " - 0s - loss: 636.2134\n",
      "Epoch 198/200\n",
      " - 0s - loss: 621.6615\n",
      "Epoch 199/200\n",
      " - 0s - loss: 619.0273\n",
      "Epoch 200/200\n",
      " - 0s - loss: 620.1533\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.callbacks.History at 0x7ff51228b9e8>"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "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,epochs = 200,batch_size=2,verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainScore = model.evaluate(trainX,trainY,verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Score: 606.05 MSE (24.62 RMSE)\n"
     ]
    }
   ],
   "source": [
    "print('Train Score: %.2f MSE (%.2f RMSE)' %(trainScore,math.sqrt(trainScore)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Score: 2355.06 MSE (48.53 RMSE)\n"
     ]
    }
   ],
   "source": [
    "print('Test Score: %.2f MSE (%.2f RMSE)'%(testScore, math.sqrt(testScore)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainPredict = model.predict(trainX)\n",
    "testPredict = model.predict(testX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "trainPredictPlot = numpy.empty_like(dataset)\n",
    "trainPredictPlot[:,:] = numpy.nan\n",
    "trainPredictPlot[look_back:len(trainPredict)+look_back,:] = trainPredict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "testPredictPlot = numpy.empty_like(dataset)\n",
    "testPredictPlot[:,:] = numpy.nan\n",
    "testPredictPlot[len(trainPredict)+(look_back*2)+1:len(dataset)-1,:] = testPredict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(dataset)\n",
    "plt.plot(trainPredictPlot)\n",
    "plt.plot(testPredictPlot)\n",
    "plt.show()"
   ]
  },
  {
   "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}