timeseries-prediction.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 288,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from keras.models import Sequential\n",
    "from keras.layers import GRU, Dense\n",
    "from keras.layers import LSTM\n",
    "from keras  import callbacks\n",
    "from keras import optimizers\n",
    "import pandas as pd \n",
    "import tensorflow as tf\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "         Passengers\n",
      "Month              \n",
      "1949-01       112.0\n",
      "1949-02       118.0\n",
      "1949-03       132.0\n",
      "1949-04       129.0\n",
      "1949-05       121.0\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x19007778da0>"
      ]
     },
     "execution_count": 289,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.read_csv('data/international-airline-passengers.csv', index_col='Month')\n",
    "print(df.head())\n",
    "df.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data cleansing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total rows: 146\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Passengers</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Month</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1949-01</th>\n",
       "      <td>112000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-02</th>\n",
       "      <td>118000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-03</th>\n",
       "      <td>132000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-04</th>\n",
       "      <td>129000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1949-05</th>\n",
       "      <td>121000.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Passengers\n",
       "Month              \n",
       "1949-01    112000.0\n",
       "1949-02    118000.0\n",
       "1949-03    132000.0\n",
       "1949-04    129000.0\n",
       "1949-05    121000.0"
      ]
     },
     "execution_count": 290,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "columns_to_keep = ['Passengers']\n",
    "df = df[columns_to_keep]\n",
    "df['Passengers'] = df['Passengers'].apply(lambda x: x*1000)\n",
    "df.index.names = ['Month']\n",
    "df.sort_index(inplace=True)\n",
    "print('Total rows: {}'.format(len(df)))\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Passengers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>144.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>280298.611111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>119966.316943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>104000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>180000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>265500.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>360500.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>622000.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Passengers\n",
       "count     144.000000\n",
       "mean   280298.611111\n",
       "std    119966.316943\n",
       "min    104000.000000\n",
       "25%    180000.000000\n",
       "50%    265500.000000\n",
       "75%    360500.000000\n",
       "max    622000.000000"
      ]
     },
     "execution_count": 291,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1900d39dd68>"
      ]
     },
     "execution_count": 292,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Null values?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Passengers    2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 293,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 294,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Passengers    2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 294,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "null_columns=df.columns[df.isnull().any()]\n",
    "df[null_columns].isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                                    Passengers\n",
      "Month                                                         \n",
      "International airline passengers: monthly total...         NaN\n",
      "NaN                                                        NaN\n"
     ]
    }
   ],
   "source": [
    "print(df[df.isnull().any(axis=1)][null_columns].head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 296,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.dropna(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Passengers    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 297,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x000001900D40E550>]],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 298,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEICAYAAABPgw/pAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAFGBJREFUeJzt3XuQpXV95/H3R26yNMslYNdwKQYqhArFbAi0iIVrenR1AU0wVWYLNqsQTY2VxI0ppyo76NaGJOUW2Q1mY22ygtGE2lUnLBdlQRcJ2hKrvPUYyMCOLLdZGSAzsnIblyQOfPeP87ScjD3dpy+nL79+v6pOnef8nt/zPN9f95lPP/07Tz+TqkKStPq9YrkLkCQtDgNdkhphoEtSIwx0SWqEgS5JjTDQJakRBrokNcJA17JLsjPJC0n2Jtmd5E+TjCx3XdJqY6BrpfjZqhoBzgFeDfzbZa5naJIcvNw1qE0GulaUqnoc+DxwVpJfSrIjyfNJHknynql+SY5LcluSZ5J8L8lfJnlFt+7fJHm82+6BJG/s2l+RZEuSh5P83yQ3JDm2W7c+SSW5PMl3kjyV5IN9xzs8yfVJnu5q+s0ku/rWn5DkpiTfTfJokl/vW3dVkhuT/LckzwFXJDkvyWSS57rfSj489C+umueZglaUJCcDFwM3A3uAtwKPAK8HPp/km1X1LWAzsAs4vtv0fKCSnAG8F3h1VT2RZD1wUNfn14G3AT8DfBf4CPBHwGV9JbwOOAP4CeAbSW6uqh3AbwHrgdOAI4DP9dX8CuB/AJ/t9nUS8BdJHqiqO7pulwC/ALwTOAz4IvCHVfVfu+mls+b/VZN6PEPXSvGZJM8AXwG+DPz7qrq9qh6uni8DXwD+adf/B8A64JSq+kFV/WX1bkz0Ir3APDPJIVW1s6oe7rZ5D/DBqtpVVX8HXAW8fb8pkN+uqheq6l7gXuCnuvZ/0dX0dFXtovfDYMqrgeOr6neq6u+r6hHgY8ClfX2+WlWfqaqXquqFrv4fT3JcVe2tqq8t+CuoNc9A10rxtqo6uqpOqapfraoXklyU5GvdlMoz9M7cj+v6/0fgIeAL3XTMFoCqegj4DXphvSfJ1iQndNucAtzSTdM8A+yg9wNgtK+Ov+lb/n/A1IezJwCP9a3rXz4FOGFqv92+P7Dffvv7A7yb3m8B307yzSRvHeSLJM3EQNeKlOQw4Cbg94HRqjqa3jRHAKrq+araXFWnAT8LvH9qrryqPlVVr6MXtAX8Xrfbx4CLuh8cU49XdvP2s3mS3lTKlJP7lh8DHt1vv0dW1cV9ff7BbU2r6sGqugx4VVffjUmOGKAO6YAMdK1Uh9KbOvkusC/JRcCbp1YmeWuSH08S4Dl6Z9ovJjkjyRu6Hwh/C7zQrQP4KPChJKd0+zg+ySUD1nMDcGWSY5KcSG+efso3gOe6D2MPT3JQkrOSvPpAO0vyr5IcX1UvAc90zS8eqL80CANdK1JVPU/vQ8wbgKeBfwnc2tfldOAvgL3AV4E/rqoJej8Ergaeojd98ip60x8Af9jt4wtJnge+BrxmwJJ+h96HsI92x70R+Luu1hfp/ZZwdrf+KeBPgKNm2N+FwP1J9nZ1XVpVfztgLdK04n9wIc1dkl+hF8I/s9y1SFM8Q5cGkGRdkgu6a9nPoHfZ5C3LXZfUz+vQpcEcClwLnEpvznsr8MfLWpG0H6dcJKkRTrlIUiOWdMrluOOOq/Xr189pm+9///scccTauDx3rYx1rYwTHGurlnqs27Zte6qqjp+t35IG+vr165mcnJzTNhMTE4yPjw+noBVmrYx1rYwTHGurlnqsSf7PIP2ccpGkRhjoktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqREGuiQ1wkCXpEZ4t8UBrN9y+5IcZ/OGfVzRd6ydV79lSY4rqQ2eoUtSIwx0SWqEgS5JjTDQJakRBrokNcJAl6RGGOiS1AgDXZIaYaBLUiMMdElqhIEuSY2YNdCTnJzkS0l2JLk/yfu69quSPJ7knu5x8fDLlSQdyCA359oHbK6qbyU5EtiW5M5u3R9U1e8PrzxJ0qBmDfSqehJ4slt+PskO4MRhFyZJmptU1eCdk/XA3cBZwPuBK4DngEl6Z/FPT7PNJmATwOjo6Llbt26dU4F79+5lZGRkTtsstu2PP7skxxk9HHa/8PLrDScetSTHXWor4Xu6VBxrm5Z6rBs3btxWVWOz9Rs40JOMAF8GPlRVNycZBZ4CCvhdYF1VvWumfYyNjdXk5ORAx5syMTHB+Pj4nLZZbEt5P/Rrtr/8S1Or90NfCd/TpeJY27TUY00yUKAPdJVLkkOAm4BPVtXNAFW1u6perKqXgI8B5y2kYEnSwgxylUuAjwM7qurDfe3r+rr9PHDf4pcnSRrUIFe5XAC8A9ie5J6u7QPAZUnOpjflshN4z1AqlCQNZJCrXL4CZJpVn1v8ciRJ8+VfikpSIwx0SWrEIHPoK8JSXTooSauVZ+iS1AgDXZIaYaBLUiMMdElqhIEuSY0w0CWpEQa6JDXCQJekRhjoktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqRGr5va5WhuW6zbJO69+y7IcV1pMnqFLUiMMdElqhIEuSY0w0CWpEQa6JDXCQJekRhjoktQIA12SGmGgS1IjDHRJaoSBLkmNmDXQk5yc5EtJdiS5P8n7uvZjk9yZ5MHu+ZjhlytJOpBBztD3AZur6ieB84FfS3ImsAW4q6pOB+7qXkuSlsmsgV5VT1bVt7rl54EdwInAJcD1XbfrgbcNq0hJ0uxSVYN3TtYDdwNnAd+pqqP71j1dVT8y7ZJkE7AJYHR09NytW7fOqcC9e/cyMjLC9sefndN2q9Ho4bD7hZdfbzjxqOUrZoimvqfTWa7v87C+1jONtTWOdXg2bty4rarGZus3cKAnGQG+DHyoqm5O8swggd5vbGysJicnBzrelImJCcbHx5ftPtlLafOGfVyz/eVb1Ld6j+6p7+l0Wrsf+kxjbY1jHZ4kAwX6QFe5JDkEuAn4ZFXd3DXvTrKuW78O2DPfYiVJCzfIVS4BPg7sqKoP9626Fbi8W74c+OzilydJGtQg/wXdBcA7gO1J7unaPgBcDdyQ5N3Ad4BfGE6JkqRBzBroVfUVIAdY/cbFLUeSNF/+pagkNcJAl6RGDDKHrmXS2iV8kobLM3RJaoSBLkmNMNAlqREGuiQ1wkCXpEYY6JLUCANdkhrhdej6EcO+/n3zhn1csQZuhywtNc/QJakRBrokNcJAl6RGGOiS1AgDXZIaYaBLUiMMdElqhIEuSY0w0CWpEQa6JDXCQJekRhjoktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqREGuiQ1wkCXpEbMGuhJPpFkT5L7+tquSvJ4knu6x8XDLVOSNJtBztD/DLhwmvY/qKqzu8fnFrcsSdJczRroVXU38L0lqEWStACpqtk7JeuB26rqrO71VcAVwHPAJLC5qp4+wLabgE0Ao6Oj527dunVOBe7du5eRkRG2P/7snLZbjUYPh90vLHcVw7cSx7nhxKOGst+p9+9a4FiHZ+PGjduqamy2fvMN9FHgKaCA3wXWVdW7ZtvP2NhYTU5Oznq8fhMTE4yPj7N+y+1z2m412rxhH9dsP3i5yxi6lTjOnVe/ZSj7nXr/rgWOdXiSDBTo87rKpap2V9WLVfUS8DHgvPnsR5K0eOYV6EnW9b38eeC+A/WVJC2NWX/vTfJpYBw4Lsku4LeA8SRn05ty2Qm8Z4g1SpIGMGugV9Vl0zR/fAi1SJIWwL8UlaRGGOiS1AgDXZIaYaBLUiMMdElqhIEuSY0w0CWpEQa6JDXCQJekRqysW95Jy2RYd/PcvGEfV8yy72Hd6VFrj2foktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqREGuiQ1wkCXpEYY6JLUCANdkhphoEtSIwx0SWqEgS5JjTDQJakRBrokNcJAl6RGGOiS1AgDXZIaYaBLUiNmDfQkn0iyJ8l9fW3HJrkzyYPd8zHDLVOSNJtBztD/DLhwv7YtwF1VdTpwV/dakrSMZg30qrob+N5+zZcA13fL1wNvW+S6JElzlKqavVOyHritqs7qXj9TVUf3rX+6qqaddkmyCdgEMDo6eu7WrVvnVODevXsZGRlh++PPzmm71Wj0cNj9wnJXMXxrZZywsse64cSjFnV/U/9W14KlHuvGjRu3VdXYbP0OHnYhVXUdcB3A2NhYjY+Pz2n7iYkJxsfHuWLL7UOobmXZvGEf12wf+rdk2a2VccLKHuvOXxxf1P1N/VtdC1bqWOd7lcvuJOsAuuc9i1eSJGk+5hvotwKXd8uXA59dnHIkSfM1yGWLnwa+CpyRZFeSdwNXA29K8iDwpu61JGkZzTq5V1WXHWDVGxe5FknSAviXopLUCANdkhphoEtSIwx0SWqEgS5JjTDQJakRBrokNcJAl6RGGOiS1AgDXZIaYaBLUiMMdElqhIEuSY0w0CWpEQa6JDXCQJekRhjoktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqREGuiQ1wkCXpEYY6JLUiIOXuwBJy2P9ltsXdX+bN+zjigH3ufPqtyzqsdXjGbokNcJAl6RGGOiS1IgFzaEn2Qk8D7wI7KuqscUoSpI0d4vxoejGqnpqEfYjSVoAp1wkqRGpqvlvnDwKPA0UcG1VXTdNn03AJoDR0dFzt27dOqdj7N27l5GREbY//uy861wtRg+H3S8sdxXDt1bGCY71QDaceNRwixmyqVxaKhs3btw2yJT2QgP9hKp6IsmrgDuBf11Vdx+o/9jYWE1OTs7pGBMTE4yPjy/6NbMr0eYN+7hme/t/GrBWxgmO9UBW+3XoU7m0VJIMFOgLmnKpqie65z3ALcB5C9mfJGn+5h3oSY5IcuTUMvBm4L7FKkySNDcL+V1wFLglydR+PlVV/3NRqpIkzdm8A72qHgF+ahFrkSQtgJctSlIjDHRJaoSBLkmNMNAlqREGuiQ1wkCXpEYY6JLUCANdkhphoEtSIwx0SWrE2rivp6QVZbluh73ab9s7G8/QJakRBrokNcJAl6RGGOiS1AgDXZIaYaBLUiO8bFHSmrFYl0tu3rCPK+a4r6W4ZNIzdElqhIEuSY0w0CWpEQa6JDXCQJekRhjoktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqREGuiQ1YkGBnuTCJA8keSjJlsUqSpI0d/MO9CQHAX8EXAScCVyW5MzFKkySNDcLOUM/D3ioqh6pqr8HtgKXLE5ZkqS5SlXNb8Pk7cCFVfXL3et3AK+pqvfu128TsKl7eQbwwBwPdRzw1LyKXH3WyljXyjjBsbZqqcd6SlUdP1unhfwHF5mm7Ud+OlTVdcB18z5IMllVY/PdfjVZK2NdK+MEx9qqlTrWhUy57AJO7nt9EvDEwsqRJM3XQgL9m8DpSU5NcihwKXDr4pQlSZqreU+5VNW+JO8F7gAOAj5RVfcvWmUvm/d0zSq0Vsa6VsYJjrVVK3Ks8/5QVJK0sviXopLUCANdkhoxtEBP8okke5Lc19d2bJI7kzzYPR/TtSfJR7pbCPx1knP6trm86/9gksv72s9Nsr3b5iNJMtMxhinJyUm+lGRHkvuTvK/F8SZ5ZZJvJLm3G+dvd+2nJvl6V8Ofdx+Sk+Sw7vVD3fr1ffu6smt/IMk/72uf9nYSBzrGsCU5KMlfJbmt5bEm2dm9v+5JMtm1NfX+7avl6CQ3Jvl292/2tc2MtaqG8gBeD5wD3NfX9h+ALd3yFuD3uuWLgc/Tu7b9fODrXfuxwCPd8zHd8jHdum8Ar+22+Txw0UzHGOYDWAec0y0fCfxverdDaGq83bFHuuVDgK939d8AXNq1fxT4lW75V4GPdsuXAn/eLZ8J3AscBpwKPEzvg/WDuuXTgEO7Pmd220x7jCX43r4f+BRw20x1rPaxAjuB4/Zra+r92zeu64Ff7pYPBY5uZazD/sKt5x8G+gPAum55HfBAt3wtcNn+/YDLgGv72q/t2tYB3+5r/2G/Ax1jKR/AZ4E3tTxe4B8B3wJeQ+8v5g7u2l8L3NEt3wG8tls+uOsX4Ergyr593dFt98Ntu/Yru0cOdIwhj/Ek4C7gDcBtM9XRwFh38qOB3tz7F/jHwKN0F4S0NtalnkMfraonAbrnV3XtJwKP9fXb1bXN1L5rmvaZjrEkul+1f5re2Wtz4+2mIO4B9gB30jvLfKaq9k1T2w/H061/Fvgx5j7+H5vhGMP0n4DfBF7qXs9Ux2ofawFfSLItvdt1QIPvX3q/EX0X+NNuKu1PkhwxQx2raqwr5UPRA91GYK7tyyrJCHAT8BtV9dxMXadpWxXjraoXq+psemev5wE/OV237nmxxrnk40/yVmBPVW3rb56hjlU71s4FVXUOvbun/lqS18/Qd7WMaToH05sK/i9V9dPA9+lNfxzIqhrrUgf67iTrALrnPV37gW4jMFP7SdO0z3SMoUpyCL0w/2RV3TxLLat+vFX1DDBBb17x6CRTf6TWX9sPx9OtPwr4HnMf/1MzHGNYLgB+LslOencSfQO9M/YWx0pVPdE97wFuoffDusX37y5gV1V9vXt9I72Ab2KsSx3otwJTnwZfTm+uear9nd0nyucDz3a/ktwBvDnJMd0nwm+mN5/4JPB8kvO7T5Dfud++pjvG0HQ1fBzYUVUf7lvV1HiTHJ/k6G75cOCfATuALwFvP8A4p2p7O/DF6k0g3gpcmt6VIacCp9P7IGna20l02xzoGENRVVdW1UlVtb6r44tV9Ysz1LFqx5rkiCRHTi3Te9/dR2PvX4Cq+hvgsSRndE1vBP7XDHWsrrEO8cOHTwNPAj+g91Pr3fTmB+8CHuyej+36ht5/lvEwsB0Y69vPu4CHuscv9bWP0XvTPQz8Z17+q9dpjzHMB/A6er9W/TVwT/e4uLXxAv8E+KtunPcB/65rP41eSD0E/HfgsK79ld3rh7r1p/Xt64PdWB6guwqga7+Y3lVCDwMf7Guf9hhL8QDGefkql+bG2h3v3u5x/1Qtrb1/+2o5G5js3sefoXeVShNj9U//JakRK+VDUUnSAhnoktQIA12SGmGgS1IjDHRJaoSBLkmNMNAlqRH/H/Cz8NPw6RHRAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.hist(bins=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 299,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df[df['Passengers'] == 0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Scaled data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Min Passengers    104000.0\n",
      "dtype: float64\n",
      "Max Passengers    622000.0\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print('Min', np.min(df))\n",
    "print('Max', np.max(df))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can then extract the NumPy array from the dataframe and convert the integer values to floating point values, which are more suitable for modeling with a neural network."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = df.astype('float32')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "LSTMs are sensitive to the scale of the input data, specifically when the sigmoid (default) or tanh activation functions are used. It can be a good practice to rescale the data to the range of 0-to-1, also called normalizing. We can easily normalize the dataset using the MinMaxScaler preprocessing class from the scikit-learn library."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler(feature_range=(0, 1))\n",
    "scaled = scaler.fit_transform(dataset) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Min 0.0\n",
      "Max 1.0\n"
     ]
    }
   ],
   "source": [
    "print('Min', np.min(scaled))\n",
    "print('Max', np.max(scaled))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.01544402]\n",
      " [0.02702703]\n",
      " [0.05405405]\n",
      " [0.04826255]\n",
      " [0.03281853]\n",
      " [0.05984556]\n",
      " [0.08494208]\n",
      " [0.08494208]\n",
      " [0.06177606]\n",
      " [0.02895753]]\n"
     ]
    }
   ],
   "source": [
    "print(scaled[:10])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create the RNN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A simple method that we can use is to split the ordered dataset into train and test datasets. The code below calculates the index of the split point and separates the data into the training datasets with 70% of the observations that we can use to train our model, leaving the remaining 30% for testing the model.\n",
    "Split into train and test sets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train: 100\n",
      "test: 44\n"
     ]
    }
   ],
   "source": [
    "train_size = int(len(scaled) * 0.70)\n",
    "test_size = len(scaled - train_size)\n",
    "train, test = scaled[0:train_size, :], scaled[train_size: len(scaled), :]\n",
    "print('train: {}\\ntest: {}'.format(len(train), len(test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "convert an array of values into a dataset matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The function takes two arguments: the dataset, which is a NumPy array that we want to convert into a dataset, and the look_back, which is the number of previous time steps to use as input variables to predict the next time period — in this case defaulted to 1."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This default will create a dataset where X is the energy quantity at a given time (t) and Y is the qty of energy at the next time (t + 1)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_dataset(dataset, look_back=1):\n",
    "    print(len(dataset), look_back)\n",
    "    dataX, dataY = [], []\n",
    "    for i in range(len(dataset)-look_back-1):\n",
    "        a = dataset[i:(i+look_back), 0]\n",
    "        print(i)\n",
    "        print('X {} to {}'.format(i, i+look_back))\n",
    "        print(a)\n",
    "        print('Y {}'.format(i + look_back))\n",
    "        print(dataset[i + look_back, 0])\n",
    "        dataset[i + look_back, 0]\n",
    "        dataX.append(a)\n",
    "        dataY.append(dataset[i + look_back, 0])\n",
    "    return np.array(dataX), np.array(dataY)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "reshape into X=t and Y=t+1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100 1\n",
      "0\n",
      "X 0 to 1\n",
      "[0.01544402]\n",
      "Y 1\n",
      "0.02702702702702703\n",
      "1\n",
      "X 1 to 2\n",
      "[0.02702703]\n",
      "Y 2\n",
      "0.05405405405405403\n",
      "2\n",
      "X 2 to 3\n",
      "[0.05405405]\n",
      "Y 3\n",
      "0.04826254826254825\n",
      "3\n",
      "X 3 to 4\n",
      "[0.04826255]\n",
      "Y 4\n",
      "0.03281853281853281\n",
      "4\n",
      "X 4 to 5\n",
      "[0.03281853]\n",
      "Y 5\n",
      "0.05984555984555981\n",
      "5\n",
      "X 5 to 6\n",
      "[0.05984556]\n",
      "Y 6\n",
      "0.08494208494208491\n",
      "6\n",
      "X 6 to 7\n",
      "[0.08494208]\n",
      "Y 7\n",
      "0.08494208494208491\n",
      "7\n",
      "X 7 to 8\n",
      "[0.08494208]\n",
      "Y 8\n",
      "0.06177606177606179\n",
      "8\n",
      "X 8 to 9\n",
      "[0.06177606]\n",
      "Y 9\n",
      "0.028957528957528955\n",
      "9\n",
      "X 9 to 10\n",
      "[0.02895753]\n",
      "Y 10\n",
      "0.0\n",
      "10\n",
      "X 10 to 11\n",
      "[0.]\n",
      "Y 11\n",
      "0.02702702702702703\n",
      "11\n",
      "X 11 to 12\n",
      "[0.02702703]\n",
      "Y 12\n",
      "0.02123552123552122\n",
      "12\n",
      "X 12 to 13\n",
      "[0.02123552]\n",
      "Y 13\n",
      "0.04247104247104247\n",
      "13\n",
      "X 13 to 14\n",
      "[0.04247104]\n",
      "Y 14\n",
      "0.07142857142857142\n",
      "14\n",
      "X 14 to 15\n",
      "[0.07142857]\n",
      "Y 15\n",
      "0.05984555984555981\n",
      "15\n",
      "X 15 to 16\n",
      "[0.05984556]\n",
      "Y 16\n",
      "0.04054054054054054\n",
      "16\n",
      "X 16 to 17\n",
      "[0.04054054]\n",
      "Y 17\n",
      "0.08687258687258684\n",
      "17\n",
      "X 17 to 18\n",
      "[0.08687259]\n",
      "Y 18\n",
      "0.1274131274131274\n",
      "18\n",
      "X 18 to 19\n",
      "[0.12741313]\n",
      "Y 19\n",
      "0.1274131274131274\n",
      "19\n",
      "X 19 to 20\n",
      "[0.12741313]\n",
      "Y 20\n",
      "0.10424710424710423\n",
      "20\n",
      "X 20 to 21\n",
      "[0.1042471]\n",
      "Y 21\n",
      "0.055984555984555956\n",
      "21\n",
      "X 21 to 22\n",
      "[0.05598456]\n",
      "Y 22\n",
      "0.019305019305019294\n",
      "22\n",
      "X 22 to 23\n",
      "[0.01930502]\n",
      "Y 23\n",
      "0.0694980694980695\n",
      "23\n",
      "X 23 to 24\n",
      "[0.06949807]\n",
      "Y 24\n",
      "0.07915057915057913\n",
      "24\n",
      "X 24 to 25\n",
      "[0.07915058]\n",
      "Y 25\n",
      "0.08880308880308882\n",
      "25\n",
      "X 25 to 26\n",
      "[0.08880309]\n",
      "Y 26\n",
      "0.14285714285714288\n",
      "26\n",
      "X 26 to 27\n",
      "[0.14285714]\n",
      "Y 27\n",
      "0.11389961389961387\n",
      "27\n",
      "X 27 to 28\n",
      "[0.11389961]\n",
      "Y 28\n",
      "0.13127413127413126\n",
      "28\n",
      "X 28 to 29\n",
      "[0.13127413]\n",
      "Y 29\n",
      "0.14285714285714288\n",
      "29\n",
      "X 29 to 30\n",
      "[0.14285714]\n",
      "Y 30\n",
      "0.1833976833976834\n",
      "30\n",
      "X 30 to 31\n",
      "[0.18339768]\n",
      "Y 31\n",
      "0.1833976833976834\n",
      "31\n",
      "X 31 to 32\n",
      "[0.18339768]\n",
      "Y 32\n",
      "0.15444015444015444\n",
      "32\n",
      "X 32 to 33\n",
      "[0.15444015]\n",
      "Y 33\n",
      "0.11196911196911194\n",
      "33\n",
      "X 33 to 34\n",
      "[0.11196911]\n",
      "Y 34\n",
      "0.08108108108108106\n",
      "34\n",
      "X 34 to 35\n",
      "[0.08108108]\n",
      "Y 35\n",
      "0.1196911196911197\n",
      "35\n",
      "X 35 to 36\n",
      "[0.11969112]\n",
      "Y 36\n",
      "0.12934362934362933\n",
      "36\n",
      "X 36 to 37\n",
      "[0.12934363]\n",
      "Y 37\n",
      "0.14671814671814673\n",
      "37\n",
      "X 37 to 38\n",
      "[0.14671815]\n",
      "Y 38\n",
      "0.17181467181467183\n",
      "38\n",
      "X 38 to 39\n",
      "[0.17181467]\n",
      "Y 39\n",
      "0.14864864864864866\n",
      "39\n",
      "X 39 to 40\n",
      "[0.14864865]\n",
      "Y 40\n",
      "0.1525096525096525\n",
      "40\n",
      "X 40 to 41\n",
      "[0.15250965]\n",
      "Y 41\n",
      "0.22007722007722005\n",
      "41\n",
      "X 41 to 42\n",
      "[0.22007722]\n",
      "Y 42\n",
      "0.24324324324324323\n",
      "42\n",
      "X 42 to 43\n",
      "[0.24324324]\n",
      "Y 43\n",
      "0.26640926640926643\n",
      "43\n",
      "X 43 to 44\n",
      "[0.26640927]\n",
      "Y 44\n",
      "0.20270270270270271\n",
      "44\n",
      "X 44 to 45\n",
      "[0.2027027]\n",
      "Y 45\n",
      "0.16795366795366792\n",
      "45\n",
      "X 45 to 46\n",
      "[0.16795367]\n",
      "Y 46\n",
      "0.13127413127413126\n",
      "46\n",
      "X 46 to 47\n",
      "[0.13127413]\n",
      "Y 47\n",
      "0.17374517374517376\n",
      "47\n",
      "X 47 to 48\n",
      "[0.17374517]\n",
      "Y 48\n",
      "0.1776061776061776\n",
      "48\n",
      "X 48 to 49\n",
      "[0.17760618]\n",
      "Y 49\n",
      "0.1776061776061776\n",
      "49\n",
      "X 49 to 50\n",
      "[0.17760618]\n",
      "Y 50\n",
      "0.25482625482625487\n",
      "50\n",
      "X 50 to 51\n",
      "[0.25482625]\n",
      "Y 51\n",
      "0.25289575289575295\n",
      "51\n",
      "X 51 to 52\n",
      "[0.25289575]\n",
      "Y 52\n",
      "0.2413127413127413\n",
      "52\n",
      "X 52 to 53\n",
      "[0.24131274]\n",
      "Y 53\n",
      "0.26833976833976836\n",
      "53\n",
      "X 53 to 54\n",
      "[0.26833977]\n",
      "Y 54\n",
      "0.3088803088803088\n",
      "54\n",
      "X 54 to 55\n",
      "[0.30888031]\n",
      "Y 55\n",
      "0.32432432432432434\n",
      "55\n",
      "X 55 to 56\n",
      "[0.32432432]\n",
      "Y 56\n",
      "0.2567567567567568\n",
      "56\n",
      "X 56 to 57\n",
      "[0.25675676]\n",
      "Y 57\n",
      "0.20656370656370657\n",
      "57\n",
      "X 57 to 58\n",
      "[0.20656371]\n",
      "Y 58\n",
      "0.14671814671814673\n",
      "58\n",
      "X 58 to 59\n",
      "[0.14671815]\n",
      "Y 59\n",
      "0.18725868725868725\n",
      "59\n",
      "X 59 to 60\n",
      "[0.18725869]\n",
      "Y 60\n",
      "0.19305019305019303\n",
      "60\n",
      "X 60 to 61\n",
      "[0.19305019]\n",
      "Y 61\n",
      "0.16216216216216214\n",
      "61\n",
      "X 61 to 62\n",
      "[0.16216216]\n",
      "Y 62\n",
      "0.25289575289575295\n",
      "62\n",
      "X 62 to 63\n",
      "[0.25289575]\n",
      "Y 63\n",
      "0.23745173745173745\n",
      "63\n",
      "X 63 to 64\n",
      "[0.23745174]\n",
      "Y 64\n",
      "0.2509652509652509\n",
      "64\n",
      "X 64 to 65\n",
      "[0.25096525]\n",
      "Y 65\n",
      "0.3088803088803088\n",
      "65\n",
      "X 65 to 66\n",
      "[0.30888031]\n",
      "Y 66\n",
      "0.38223938223938225\n",
      "66\n",
      "X 66 to 67\n",
      "[0.38223938]\n",
      "Y 67\n",
      "0.3648648648648648\n",
      "67\n",
      "X 67 to 68\n",
      "[0.36486486]\n",
      "Y 68\n",
      "0.2992277992277992\n",
      "68\n",
      "X 68 to 69\n",
      "[0.2992278]\n",
      "Y 69\n",
      "0.2413127413127413\n",
      "69\n",
      "X 69 to 70\n",
      "[0.24131274]\n",
      "Y 70\n",
      "0.1911196911196911\n",
      "70\n",
      "X 70 to 71\n",
      "[0.19111969]\n",
      "Y 71\n",
      "0.2413127413127413\n",
      "71\n",
      "X 71 to 72\n",
      "[0.24131274]\n",
      "Y 72\n",
      "0.26640926640926643\n",
      "72\n",
      "X 72 to 73\n",
      "[0.26640927]\n",
      "Y 73\n",
      "0.249034749034749\n",
      "73\n",
      "X 73 to 74\n",
      "[0.24903475]\n",
      "Y 74\n",
      "0.3146718146718146\n",
      "74\n",
      "X 74 to 75\n",
      "[0.31467181]\n",
      "Y 75\n",
      "0.31853281853281845\n",
      "75\n",
      "X 75 to 76\n",
      "[0.31853282]\n",
      "Y 76\n",
      "0.3204633204633204\n",
      "76\n",
      "X 76 to 77\n",
      "[0.32046332]\n",
      "Y 77\n",
      "0.4073359073359073\n",
      "77\n",
      "X 77 to 78\n",
      "[0.40733591]\n",
      "Y 78\n",
      "0.5019305019305019\n",
      "78\n",
      "X 78 to 79\n",
      "[0.5019305]\n",
      "Y 79\n",
      "0.46911196911196906\n",
      "79\n",
      "X 79 to 80\n",
      "[0.46911197]\n",
      "Y 80\n",
      "0.4015444015444015\n",
      "80\n",
      "X 80 to 81\n",
      "[0.4015444]\n",
      "Y 81\n",
      "0.3281853281853282\n",
      "81\n",
      "X 81 to 82\n",
      "[0.32818533]\n",
      "Y 82\n",
      "0.2567567567567568\n",
      "82\n",
      "X 82 to 83\n",
      "[0.25675676]\n",
      "Y 83\n",
      "0.3359073359073359\n",
      "83\n",
      "X 83 to 84\n",
      "[0.33590734]\n",
      "Y 84\n",
      "0.34749034749034746\n",
      "84\n",
      "X 84 to 85\n",
      "[0.34749035]\n",
      "Y 85\n",
      "0.333976833976834\n",
      "85\n",
      "X 85 to 86\n",
      "[0.33397683]\n",
      "Y 86\n",
      "0.41119691119691115\n",
      "86\n",
      "X 86 to 87\n",
      "[0.41119691]\n",
      "Y 87\n",
      "0.40347490347490345\n",
      "87\n",
      "X 87 to 88\n",
      "[0.4034749]\n",
      "Y 88\n",
      "0.4131274131274131\n",
      "88\n",
      "X 88 to 89\n",
      "[0.41312741]\n",
      "Y 89\n",
      "0.5212355212355212\n",
      "89\n",
      "X 89 to 90\n",
      "[0.52123552]\n",
      "Y 90\n",
      "0.5965250965250966\n",
      "90\n",
      "X 90 to 91\n",
      "[0.5965251]\n",
      "Y 91\n",
      "0.581081081081081\n",
      "91\n",
      "X 91 to 92\n",
      "[0.58108108]\n",
      "Y 92\n",
      "0.4845559845559845\n",
      "92\n",
      "X 92 to 93\n",
      "[0.48455598]\n",
      "Y 93\n",
      "0.38996138996138996\n",
      "93\n",
      "X 93 to 94\n",
      "[0.38996139]\n",
      "Y 94\n",
      "0.3223938223938224\n",
      "94\n",
      "X 94 to 95\n",
      "[0.32239382]\n",
      "Y 95\n",
      "0.38996138996138996\n",
      "95\n",
      "X 95 to 96\n",
      "[0.38996139]\n",
      "Y 96\n",
      "0.4073359073359073\n",
      "96\n",
      "X 96 to 97\n",
      "[0.40733591]\n",
      "Y 97\n",
      "0.3803088803088803\n",
      "97\n",
      "X 97 to 98\n",
      "[0.38030888]\n",
      "Y 98\n",
      "0.4864864864864865\n",
      "44 1\n",
      "0\n",
      "X 0 to 1\n",
      "[0.48455598]\n",
      "Y 1\n",
      "0.6138996138996139\n",
      "1\n",
      "X 1 to 2\n",
      "[0.61389961]\n",
      "Y 2\n",
      "0.6969111969111969\n",
      "2\n",
      "X 2 to 3\n",
      "[0.6969112]\n",
      "Y 3\n",
      "0.7007722007722007\n",
      "3\n",
      "X 3 to 4\n",
      "[0.7007722]\n",
      "Y 4\n",
      "0.5791505791505791\n",
      "4\n",
      "X 4 to 5\n",
      "[0.57915058]\n",
      "Y 5\n",
      "0.46911196911196906\n",
      "5\n",
      "X 5 to 6\n",
      "[0.46911197]\n",
      "Y 6\n",
      "0.38803088803088803\n",
      "6\n",
      "X 6 to 7\n",
      "[0.38803089]\n",
      "Y 7\n",
      "0.44787644787644787\n",
      "7\n",
      "X 7 to 8\n",
      "[0.44787645]\n",
      "Y 8\n",
      "0.4555984555984556\n",
      "8\n",
      "X 8 to 9\n",
      "[0.45559846]\n",
      "Y 9\n",
      "0.4131274131274131\n",
      "9\n",
      "X 9 to 10\n",
      "[0.41312741]\n",
      "Y 10\n",
      "0.4980694980694981\n",
      "10\n",
      "X 10 to 11\n",
      "[0.4980695]\n",
      "Y 11\n",
      "0.471042471042471\n",
      "11\n",
      "X 11 to 12\n",
      "[0.47104247]\n",
      "Y 12\n",
      "0.5\n",
      "12\n",
      "X 12 to 13\n",
      "[0.5]\n",
      "Y 13\n",
      "0.6389961389961389\n",
      "13\n",
      "X 13 to 14\n",
      "[0.63899614]\n",
      "Y 14\n",
      "0.747104247104247\n",
      "14\n",
      "X 14 to 15\n",
      "[0.74710425]\n",
      "Y 15\n",
      "0.7741312741312741\n",
      "15\n",
      "X 15 to 16\n",
      "[0.77413127]\n",
      "Y 16\n",
      "0.5791505791505791\n",
      "16\n",
      "X 16 to 17\n",
      "[0.57915058]\n",
      "Y 17\n",
      "0.4922779922779923\n",
      "17\n",
      "X 17 to 18\n",
      "[0.49227799]\n",
      "Y 18\n",
      "0.39768339768339767\n",
      "18\n",
      "X 18 to 19\n",
      "[0.3976834]\n",
      "Y 19\n",
      "0.4498069498069498\n",
      "19\n",
      "X 19 to 20\n",
      "[0.44980695]\n",
      "Y 20\n",
      "0.4942084942084942\n",
      "20\n",
      "X 20 to 21\n",
      "[0.49420849]\n",
      "Y 21\n",
      "0.45945945945945943\n",
      "21\n",
      "X 21 to 22\n",
      "[0.45945946]\n",
      "Y 22\n",
      "0.583011583011583\n",
      "22\n",
      "X 22 to 23\n",
      "[0.58301158]\n",
      "Y 23\n",
      "0.5637065637065637\n",
      "23\n",
      "X 23 to 24\n",
      "[0.56370656]\n",
      "Y 24\n",
      "0.61003861003861\n",
      "24\n",
      "X 24 to 25\n",
      "[0.61003861]\n",
      "Y 25\n",
      "0.7104247104247104\n",
      "25\n",
      "X 25 to 26\n",
      "[0.71042471]\n",
      "Y 26\n",
      "0.8571428571428572\n",
      "26\n",
      "X 26 to 27\n",
      "[0.85714286]\n",
      "Y 27\n",
      "0.8783783783783783\n",
      "27\n",
      "X 27 to 28\n",
      "[0.87837838]\n",
      "Y 28\n",
      "0.693050193050193\n",
      "28\n",
      "X 28 to 29\n",
      "[0.69305019]\n",
      "Y 29\n",
      "0.5849420849420849\n",
      "29\n",
      "X 29 to 30\n",
      "[0.58494208]\n",
      "Y 30\n",
      "0.4980694980694981\n",
      "30\n",
      "X 30 to 31\n",
      "[0.4980695]\n",
      "Y 31\n",
      "0.581081081081081\n",
      "31\n",
      "X 31 to 32\n",
      "[0.58108108]\n",
      "Y 32\n",
      "0.6042471042471043\n",
      "32\n",
      "X 32 to 33\n",
      "[0.6042471]\n",
      "Y 33\n",
      "0.5540540540540541\n",
      "33\n",
      "X 33 to 34\n",
      "[0.55405405]\n",
      "Y 34\n",
      "0.6081081081081081\n",
      "34\n",
      "X 34 to 35\n",
      "[0.60810811]\n",
      "Y 35\n",
      "0.6891891891891891\n",
      "35\n",
      "X 35 to 36\n",
      "[0.68918919]\n",
      "Y 36\n",
      "0.7104247104247104\n",
      "36\n",
      "X 36 to 37\n",
      "[0.71042471]\n",
      "Y 37\n",
      "0.832046332046332\n",
      "37\n",
      "X 37 to 38\n",
      "[0.83204633]\n",
      "Y 38\n",
      "1.0\n",
      "38\n",
      "X 38 to 39\n",
      "[1.]\n",
      "Y 39\n",
      "0.9691119691119692\n",
      "39\n",
      "X 39 to 40\n",
      "[0.96911197]\n",
      "Y 40\n",
      "0.7799227799227799\n",
      "40\n",
      "X 40 to 41\n",
      "[0.77992278]\n",
      "Y 41\n",
      "0.6891891891891891\n",
      "41\n",
      "X 41 to 42\n",
      "[0.68918919]\n",
      "Y 42\n",
      "0.5521235521235521\n"
     ]
    }
   ],
   "source": [
    "look_back = 1\n",
    "X_train, y_train = create_dataset(train, look_back)\n",
    "X_test, y_test = create_dataset(test, look_back)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The LSTM network expects the input data (X) to be provided with a specific array structure in the form of: [samples, time steps, features].\n",
    "\n",
    "\n",
    "reshape input to be [samples, time steps, features]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(98, 1, 1)\n",
      "(42, 1, 1)\n"
     ]
    }
   ],
   "source": [
    "X_train = np.reshape(X_train, (X_train.shape[0], X_train.shape[1], 1))\n",
    "X_test = np.reshape(X_test, (X_test.shape[0], X_test.shape[1], 1))\n",
    "print(X_train.shape)\n",
    "print(X_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The network has a visible layer with 1 input, a hidden layer with 4 LSTM blocks or neurons, and an output layer that makes a single value prediction. The default sigmoid activation function is used for the LSTM blocks. The network is trained for 100 epochs and a batch size of 1 is used.\n",
    "\n",
    "\n",
    "create and fit the LSTM network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      " - 1s - loss: 0.0292\n",
      "Epoch 2/100\n",
      " - 0s - loss: 0.0187\n",
      "Epoch 3/100\n",
      " - 0s - loss: 0.0176\n",
      "Epoch 4/100\n",
      " - 0s - loss: 0.0173\n",
      "Epoch 5/100\n",
      " - 0s - loss: 0.0163\n",
      "Epoch 6/100\n",
      " - 0s - loss: 0.0150\n",
      "Epoch 7/100\n",
      " - 0s - loss: 0.0141\n",
      "Epoch 8/100\n",
      " - 0s - loss: 0.0147\n",
      "Epoch 9/100\n",
      " - 0s - loss: 0.0155\n",
      "Epoch 10/100\n",
      " - 0s - loss: 0.0151\n",
      "Epoch 11/100\n",
      " - 0s - loss: 0.0138\n",
      "Epoch 12/100\n",
      " - 0s - loss: 0.0161\n",
      "Epoch 13/100\n",
      " - 0s - loss: 0.0151\n",
      "Epoch 14/100\n",
      " - 0s - loss: 0.0128\n",
      "Epoch 15/100\n",
      " - 0s - loss: 0.0127\n",
      "Epoch 16/100\n",
      " - 0s - loss: 0.0145\n",
      "Epoch 17/100\n",
      " - 0s - loss: 0.0125\n",
      "Epoch 18/100\n",
      " - 0s - loss: 0.0124\n",
      "Epoch 19/100\n",
      " - 0s - loss: 0.0117\n",
      "Epoch 20/100\n",
      " - 0s - loss: 0.0110\n",
      "Epoch 21/100\n",
      " - 0s - loss: 0.0114\n",
      "Epoch 22/100\n",
      " - 0s - loss: 0.0112\n",
      "Epoch 23/100\n",
      " - 0s - loss: 0.0105\n",
      "Epoch 24/100\n",
      " - 0s - loss: 0.0095\n",
      "Epoch 25/100\n",
      " - 0s - loss: 0.0095\n",
      "Epoch 26/100\n",
      " - 0s - loss: 0.0100\n",
      "Epoch 27/100\n",
      " - 0s - loss: 0.0083\n",
      "Epoch 28/100\n",
      " - 0s - loss: 0.0083\n",
      "Epoch 29/100\n",
      " - 0s - loss: 0.0088\n",
      "Epoch 30/100\n",
      " - 0s - loss: 0.0069\n",
      "Epoch 31/100\n",
      " - 0s - loss: 0.0080\n",
      "Epoch 32/100\n",
      " - 0s - loss: 0.0079\n",
      "Epoch 33/100\n",
      " - 0s - loss: 0.0086\n",
      "Epoch 34/100\n",
      " - 0s - loss: 0.0077\n",
      "Epoch 35/100\n",
      " - 0s - loss: 0.0071\n",
      "Epoch 36/100\n",
      " - 0s - loss: 0.0069\n",
      "Epoch 37/100\n",
      " - 0s - loss: 0.0061\n",
      "Epoch 38/100\n",
      " - 0s - loss: 0.0065\n",
      "Epoch 39/100\n",
      " - 0s - loss: 0.0065\n",
      "Epoch 40/100\n",
      " - 0s - loss: 0.0075\n",
      "Epoch 41/100\n",
      " - 0s - loss: 0.0072\n",
      "Epoch 42/100\n",
      " - 0s - loss: 0.0061\n",
      "Epoch 43/100\n",
      " - 0s - loss: 0.0056\n",
      "Epoch 44/100\n",
      " - 0s - loss: 0.0054\n",
      "Epoch 45/100\n",
      " - 0s - loss: 0.0056\n",
      "Epoch 46/100\n",
      " - 0s - loss: 0.0057\n",
      "Epoch 47/100\n",
      " - 0s - loss: 0.0051\n",
      "Epoch 48/100\n",
      " - 0s - loss: 0.0057\n",
      "Epoch 49/100\n",
      " - 0s - loss: 0.0050\n",
      "Epoch 50/100\n",
      " - 0s - loss: 0.0053\n",
      "Epoch 51/100\n",
      " - 0s - loss: 0.0054\n",
      "Epoch 52/100\n",
      " - 0s - loss: 0.0042\n",
      "Epoch 53/100\n",
      " - 0s - loss: 0.0053\n",
      "Epoch 54/100\n",
      " - 0s - loss: 0.0052\n",
      "Epoch 55/100\n",
      " - 0s - loss: 0.0044\n",
      "Epoch 56/100\n",
      " - 0s - loss: 0.0042\n",
      "Epoch 57/100\n",
      " - 0s - loss: 0.0046\n",
      "Epoch 58/100\n",
      " - 0s - loss: 0.0042\n",
      "Epoch 59/100\n",
      " - 0s - loss: 0.0046\n",
      "Epoch 60/100\n",
      " - 0s - loss: 0.0040\n",
      "Epoch 61/100\n",
      " - 0s - loss: 0.0052\n",
      "Epoch 62/100\n",
      " - 0s - loss: 0.0041\n",
      "Epoch 63/100\n",
      " - 0s - loss: 0.0047\n",
      "Epoch 64/100\n",
      " - 0s - loss: 0.0039\n",
      "Epoch 65/100\n",
      " - 0s - loss: 0.0044\n",
      "Epoch 66/100\n",
      " - 0s - loss: 0.0045\n",
      "Epoch 67/100\n",
      " - 0s - loss: 0.0038\n",
      "Epoch 68/100\n",
      " - 0s - loss: 0.0042\n",
      "Epoch 69/100\n",
      " - 0s - loss: 0.0034\n",
      "Epoch 70/100\n",
      " - 0s - loss: 0.0043\n",
      "Epoch 71/100\n",
      " - 0s - loss: 0.0039\n",
      "Epoch 72/100\n",
      " - 0s - loss: 0.0034\n",
      "Epoch 73/100\n",
      " - 0s - loss: 0.0043\n",
      "Epoch 74/100\n",
      " - 0s - loss: 0.0044\n",
      "Epoch 75/100\n",
      " - 0s - loss: 0.0048\n",
      "Epoch 76/100\n",
      " - 0s - loss: 0.0035\n",
      "Epoch 77/100\n",
      " - 0s - loss: 0.0040\n",
      "Epoch 78/100\n",
      " - 0s - loss: 0.0038\n",
      "Epoch 79/100\n",
      " - 0s - loss: 0.0037\n",
      "Epoch 80/100\n",
      " - 0s - loss: 0.0039\n",
      "Epoch 81/100\n",
      " - 0s - loss: 0.0029\n",
      "Epoch 82/100\n",
      " - 0s - loss: 0.0029\n",
      "Epoch 83/100\n",
      " - 0s - loss: 0.0034\n",
      "Epoch 84/100\n",
      " - 0s - loss: 0.0043\n",
      "Epoch 85/100\n",
      " - 0s - loss: 0.0037\n",
      "Epoch 86/100\n",
      " - 0s - loss: 0.0034\n",
      "Epoch 87/100\n",
      " - 0s - loss: 0.0035\n",
      "Epoch 88/100\n",
      " - 0s - loss: 0.0048\n",
      "Epoch 89/100\n",
      " - 0s - loss: 0.0037\n",
      "Epoch 90/100\n",
      " - 0s - loss: 0.0038\n",
      "Epoch 91/100\n",
      " - 0s - loss: 0.0041\n",
      "Epoch 92/100\n",
      " - 0s - loss: 0.0040\n",
      "Epoch 93/100\n",
      " - 0s - loss: 0.0037\n",
      "Epoch 94/100\n",
      " - 0s - loss: 0.0037\n",
      "Epoch 95/100\n",
      " - 0s - loss: 0.0038\n",
      "Epoch 96/100\n",
      " - 0s - loss: 0.0034\n",
      "Epoch 97/100\n",
      " - 0s - loss: 0.0033\n",
      "Epoch 98/100\n",
      " - 0s - loss: 0.0035\n",
      "Epoch 99/100\n",
      " - 0s - loss: 0.0038\n",
      "Epoch 100/100\n",
      " - 0s - loss: 0.0033\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x19013ec8eb8>"
      ]
     },
     "execution_count": 309,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "batch_size = 1\n",
    "model = Sequential()\n",
    "model.add(LSTM(4, batch_input_shape=(batch_size, look_back, 1), stateful=True))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mean_squared_error', optimizer='adam')\n",
    "model.fit(X_train, y_train, epochs=100, batch_size=batch_size, verbose=2, shuffle=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that we invert the predictions before calculating error scores to ensure that performance is reported in the same units as the original data (thousands of passengers per month).\n",
    "\n",
    "Make preditions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Score: 33646.74 RMSE\n",
      "Test Score: 58630.98 RMSE\n"
     ]
    }
   ],
   "source": [
    "import math\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "trainPredict = model.predict(X_train, batch_size=batch_size)\n",
    "model.reset_states()\n",
    "\n",
    "testPredict = model.predict(X_test, batch_size=batch_size)\n",
    "# invert predictions\n",
    "trainPredict = scaler.inverse_transform(trainPredict)\n",
    "y_train = scaler.inverse_transform([y_train])\n",
    "testPredict = scaler.inverse_transform(testPredict)\n",
    "y_test = scaler.inverse_transform([y_test])\n",
    "# calculate root mean squared error\n",
    "trainScore = math.sqrt(mean_squared_error(y_train[0], trainPredict[:,0]))\n",
    "print('Train Score: %.2f RMSE' % (trainScore))\n",
    "testScore = math.sqrt(mean_squared_error(y_test[0], testPredict[:,0]))\n",
    "print('Test Score: %.2f RMSE' % (testScore))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Because of how the dataset was prepared, we must shift the predictions so that they align on the x-axis with the original dataset. Once prepared, the data is plotted, showing the original dataset in blue, the predictions for the training dataset in green, and the predictions on the unseen test dataset in red.\n",
    "\n",
    "\n",
    "shift train predictions for plotting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "trainPredictPlot = np.empty_like(scaled)\n",
    "trainPredictPlot[:, :] = np.nan\n",
    "trainPredictPlot[look_back:len(trainPredict)+look_back, :] = trainPredict\n",
    "# shift test predictions for plotting\n",
    "testPredictPlot = np.empty_like(scaled)\n",
    "testPredictPlot[:, :] = np.nan\n",
    "testPredictPlot[len(trainPredict)+(look_back*2)+1:len(scaled)-1, :] = testPredict\n",
    "# plot baseline and predictions\n",
    "plt.figure(figsize=(20,10))\n",
    "plt.plot(scaler.inverse_transform(scaled))\n",
    "plt.plot(trainPredictPlot)\n",
    "plt.plot(testPredictPlot)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}