12-Month Forecast With LSTM.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "from statsmodels.tools.eval_measures import rmse\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from keras.preprocessing.sequence import TimeseriesGenerator\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import LSTM\n",
    "from keras.layers import Dropout\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "df = pd.read_csv('AirPassengers.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.Month = pd.to_datetime(df.Month)\n",
    "df = df.set_index(\"Month\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "train, test = df[:-12], df[-12:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler()\n",
    "scaler.fit(train)\n",
    "train = scaler.transform(train)\n",
    "test = scaler.transform(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_input = 12\n",
    "n_features = 1\n",
    "generator = TimeseriesGenerator(train, train, length=n_input, batch_size=6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\Grego\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\framework\\op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Colocations handled automatically by placer.\n",
      "WARNING:tensorflow:From C:\\Users\\Grego\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:3445: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(LSTM(200, activation='relu', input_shape=(n_input, n_features)))\n",
    "model.add(Dropout(0.15))\n",
    "model.add(Dense(1))\n",
    "model.compile(optimizer='adam', loss='mse')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\Grego\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\ops\\math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n",
      "Epoch 1/90\n",
      "20/20 [==============================] - 3s 151ms/step - loss: 0.0867\n",
      "Epoch 2/90\n",
      "20/20 [==============================] - 1s 28ms/step - loss: 0.0180\n",
      "Epoch 3/90\n",
      "20/20 [==============================] - 1s 29ms/step - loss: 0.0138\n",
      "Epoch 4/90\n",
      "20/20 [==============================] - 1s 28ms/step - loss: 0.0116\n",
      "Epoch 5/90\n",
      "20/20 [==============================] - 1s 29ms/step - loss: 0.0121\n",
      "Epoch 6/90\n",
      "20/20 [==============================] - 1s 28ms/step - loss: 0.0099\n",
      "Epoch 7/90\n",
      "20/20 [==============================] - 1s 29ms/step - loss: 0.0103\n",
      "Epoch 8/90\n",
      "20/20 [==============================] - 1s 28ms/step - loss: 0.0159\n",
      "Epoch 9/90\n",
      "20/20 [==============================] - 1s 29ms/step - loss: 0.0156\n",
      "Epoch 10/90\n",
      "20/20 [==============================] - 1s 28ms/step - loss: 0.0138\n",
      "Epoch 11/90\n",
      "20/20 [==============================] - 1s 28ms/step - loss: 0.0098\n",
      "Epoch 12/90\n",
      "20/20 [==============================] - 1s 29ms/step - loss: 0.0128\n",
      "Epoch 13/90\n",
      "20/20 [==============================] - 1s 28ms/step - loss: 0.0098\n",
      "Epoch 14/90\n",
      "20/20 [==============================] - 1s 29ms/step - loss: 0.0092\n",
      "Epoch 15/90\n",
      "20/20 [==============================] - 1s 27ms/step - loss: 0.0106\n",
      "Epoch 16/90\n",
      "20/20 [==============================] - 1s 30ms/step - loss: 0.0097\n",
      "Epoch 17/90\n",
      "20/20 [==============================] - 1s 37ms/step - loss: 0.0089\n",
      "Epoch 18/90\n",
      "20/20 [==============================] - 1s 34ms/step - loss: 0.0101\n",
      "Epoch 19/90\n",
      "20/20 [==============================] - 1s 50ms/step - loss: 0.0094: 0s - loss: 0.00\n",
      "Epoch 20/90\n",
      "20/20 [==============================] - 1s 36ms/step - loss: 0.0075\n",
      "Epoch 21/90\n",
      "20/20 [==============================] - 1s 35ms/step - loss: 0.0077\n",
      "Epoch 22/90\n",
      "20/20 [==============================] - 1s 31ms/step - loss: 0.0071\n",
      "Epoch 23/90\n",
      "20/20 [==============================] - 1s 29ms/step - loss: 0.0076\n",
      "Epoch 24/90\n",
      "20/20 [==============================] - 1s 30ms/step - loss: 0.0057: 0s - loss: 0.005\n",
      "Epoch 25/90\n",
      "20/20 [==============================] - 1s 28ms/step - loss: 0.0079\n",
      "Epoch 26/90\n",
      "20/20 [==============================] - 1s 34ms/step - loss: 0.0059: 0s - loss: 0.006\n",
      "Epoch 27/90\n",
      "20/20 [==============================] - 1s 25ms/step - loss: 0.0075\n",
      "Epoch 28/90\n",
      "20/20 [==============================] - 1s 29ms/step - loss: 0.0048\n",
      "Epoch 29/90\n",
      "20/20 [==============================] - 1s 31ms/step - loss: 0.0051\n",
      "Epoch 30/90\n",
      "20/20 [==============================] - 1s 28ms/step - loss: 0.0058\n",
      "Epoch 31/90\n",
      "20/20 [==============================] - 0s 25ms/step - loss: 0.0051\n",
      "Epoch 32/90\n",
      "20/20 [==============================] - 1s 35ms/step - loss: 0.0041\n",
      "Epoch 33/90\n",
      "20/20 [==============================] - 1s 39ms/step - loss: 0.0034\n",
      "Epoch 34/90\n",
      "20/20 [==============================] - 1s 41ms/step - loss: 0.0039\n",
      "Epoch 35/90\n",
      "20/20 [==============================] - 1s 35ms/step - loss: 0.0054\n",
      "Epoch 36/90\n",
      "20/20 [==============================] - 1s 36ms/step - loss: 0.0076\n",
      "Epoch 37/90\n",
      "20/20 [==============================] - 1s 26ms/step - loss: 0.0045\n",
      "Epoch 38/90\n",
      "20/20 [==============================] - 1s 49ms/step - loss: 0.0051\n",
      "Epoch 39/90\n",
      "20/20 [==============================] - 1s 39ms/step - loss: 0.0050\n",
      "Epoch 40/90\n",
      "20/20 [==============================] - 1s 47ms/step - loss: 0.0048\n",
      "Epoch 41/90\n",
      "20/20 [==============================] - 2s 82ms/step - loss: 0.0040\n",
      "Epoch 42/90\n",
      "20/20 [==============================] - 1s 63ms/step - loss: 0.0045\n",
      "Epoch 43/90\n",
      "20/20 [==============================] - 1s 42ms/step - loss: 0.0042\n",
      "Epoch 44/90\n",
      "20/20 [==============================] - 1s 38ms/step - loss: 0.0040\n",
      "Epoch 45/90\n",
      "20/20 [==============================] - 1s 38ms/step - loss: 0.0051\n",
      "Epoch 46/90\n",
      "20/20 [==============================] - 1s 51ms/step - loss: 0.0035\n",
      "Epoch 47/90\n",
      "20/20 [==============================] - 1s 39ms/step - loss: 0.0038\n",
      "Epoch 48/90\n",
      "20/20 [==============================] - 1s 38ms/step - loss: 0.0043: 1s - los\n",
      "Epoch 49/90\n",
      "20/20 [==============================] - 1s 38ms/step - loss: 0.0028\n",
      "Epoch 50/90\n",
      "20/20 [==============================] - 1s 37ms/step - loss: 0.0044\n",
      "Epoch 51/90\n",
      "20/20 [==============================] - 1s 43ms/step - loss: 0.0032\n",
      "Epoch 52/90\n",
      "20/20 [==============================] - 1s 36ms/step - loss: 0.0031\n",
      "Epoch 53/90\n",
      "20/20 [==============================] - 1s 37ms/step - loss: 0.0030\n",
      "Epoch 54/90\n",
      "20/20 [==============================] - 1s 44ms/step - loss: 0.0026\n",
      "Epoch 55/90\n",
      "20/20 [==============================] - 1s 37ms/step - loss: 0.0025\n",
      "Epoch 56/90\n",
      "20/20 [==============================] - 1s 32ms/step - loss: 0.0050\n",
      "Epoch 57/90\n",
      "20/20 [==============================] - 1s 30ms/step - loss: 0.0039\n",
      "Epoch 58/90\n",
      "20/20 [==============================] - 1s 29ms/step - loss: 0.0047\n",
      "Epoch 59/90\n",
      "20/20 [==============================] - 1s 32ms/step - loss: 0.0053\n",
      "Epoch 60/90\n",
      "20/20 [==============================] - 1s 43ms/step - loss: 0.0022\n",
      "Epoch 61/90\n",
      "20/20 [==============================] - 1s 42ms/step - loss: 0.0026\n",
      "Epoch 62/90\n",
      "20/20 [==============================] - 1s 34ms/step - loss: 0.0050\n",
      "Epoch 63/90\n",
      "20/20 [==============================] - 1s 30ms/step - loss: 0.0058\n",
      "Epoch 64/90\n",
      "20/20 [==============================] - 1s 30ms/step - loss: 0.0038\n",
      "Epoch 65/90\n",
      "20/20 [==============================] - 1s 30ms/step - loss: 0.0034\n",
      "Epoch 66/90\n",
      "20/20 [==============================] - 1s 29ms/step - loss: 0.0049\n",
      "Epoch 67/90\n",
      "20/20 [==============================] - 1s 32ms/step - loss: 0.0028\n",
      "Epoch 68/90\n",
      "20/20 [==============================] - 0s 24ms/step - loss: 0.0027\n",
      "Epoch 69/90\n",
      "20/20 [==============================] - 1s 26ms/step - loss: 0.0035\n",
      "Epoch 70/90\n",
      "20/20 [==============================] - 1s 41ms/step - loss: 0.0036\n",
      "Epoch 71/90\n",
      "20/20 [==============================] - 1s 44ms/step - loss: 0.0025\n",
      "Epoch 72/90\n",
      "20/20 [==============================] - 1s 40ms/step - loss: 0.0024\n",
      "Epoch 73/90\n",
      "20/20 [==============================] - 1s 33ms/step - loss: 0.0027\n",
      "Epoch 74/90\n",
      "20/20 [==============================] - 1s 35ms/step - loss: 0.0018\n",
      "Epoch 75/90\n",
      "20/20 [==============================] - 1s 32ms/step - loss: 0.0030\n",
      "Epoch 76/90\n",
      "20/20 [==============================] - 1s 41ms/step - loss: 0.0027\n",
      "Epoch 77/90\n",
      "20/20 [==============================] - 1s 35ms/step - loss: 0.0030\n",
      "Epoch 78/90\n",
      "20/20 [==============================] - 1s 27ms/step - loss: 0.0038\n",
      "Epoch 79/90\n",
      "20/20 [==============================] - 1s 27ms/step - loss: 0.0034\n",
      "Epoch 80/90\n",
      "20/20 [==============================] - 0s 23ms/step - loss: 0.0041\n",
      "Epoch 81/90\n",
      "20/20 [==============================] - 1s 27ms/step - loss: 0.0036\n",
      "Epoch 82/90\n",
      "20/20 [==============================] - 1s 27ms/step - loss: 0.0038\n",
      "Epoch 83/90\n",
      "20/20 [==============================] - 1s 26ms/step - loss: 0.0025\n",
      "Epoch 84/90\n",
      "20/20 [==============================] - 1s 33ms/step - loss: 0.0044\n",
      "Epoch 85/90\n",
      "20/20 [==============================] - 1s 43ms/step - loss: 0.0027\n",
      "Epoch 86/90\n",
      "20/20 [==============================] - 1s 37ms/step - loss: 0.0033\n",
      "Epoch 87/90\n",
      "20/20 [==============================] - 1s 35ms/step - loss: 0.0030\n",
      "Epoch 88/90\n",
      "20/20 [==============================] - 1s 26ms/step - loss: 0.0035\n",
      "Epoch 89/90\n",
      "20/20 [==============================] - 1s 31ms/step - loss: 0.0022\n",
      "Epoch 90/90\n",
      "20/20 [==============================] - 1s 28ms/step - loss: 0.0018\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1b22fde9da0>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit_generator(generator,epochs=90)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_list = []\n",
    "\n",
    "batch = train[-n_input:].reshape((1, n_input, n_features))\n",
    "\n",
    "for i in range(n_input):   \n",
    "    pred_list.append(model.predict(batch)[0]) \n",
    "    batch = np.append(batch[:,1:,:],[[pred_list[i]]],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_predict = pd.DataFrame(scaler.inverse_transform(pred_list),\n",
    "                          index=df[-n_input:].index, columns=['Prediction'])\n",
    "\n",
    "df_test = pd.concat([df,df_predict], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20, 5))\n",
    "plt.plot(df_test.index, df_test['AirPassengers'])\n",
    "plt.plot(df_test.index, df_test['Prediction'], color='r')\n",
    "plt.legend(loc='best', fontsize='xx-large')\n",
    "plt.xticks(fontsize=18)\n",
    "plt.yticks(fontsize=16)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rmse:  [21.85449131]\n"
     ]
    }
   ],
   "source": [
    "pred_actual_rmse = rmse(df_test.iloc[-n_input:, [0]], df_test.iloc[-n_input:, [1]])\n",
    "print(\"rmse: \", pred_actual_rmse)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler.fit(train)\n",
    "train = scaler.transform(train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_input = 12\n",
    "n_features = 1\n",
    "generator = TimeseriesGenerator(train, train, length=n_input, batch_size=6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/90\n",
      "22/22 [==============================] - 1s 23ms/step - loss: 0.0018\n",
      "Epoch 2/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0018\n",
      "Epoch 3/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0016\n",
      "Epoch 4/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0015\n",
      "Epoch 5/90\n",
      "22/22 [==============================] - 1s 33ms/step - loss: 0.0013\n",
      "Epoch 6/90\n",
      "22/22 [==============================] - 1s 32ms/step - loss: 0.0017\n",
      "Epoch 7/90\n",
      "22/22 [==============================] - 1s 27ms/step - loss: 0.0016\n",
      "Epoch 8/90\n",
      "22/22 [==============================] - 1s 33ms/step - loss: 0.0013\n",
      "Epoch 9/90\n",
      "22/22 [==============================] - 1s 31ms/step - loss: 0.0021\n",
      "Epoch 10/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0019\n",
      "Epoch 11/90\n",
      "22/22 [==============================] - 1s 31ms/step - loss: 0.0014\n",
      "Epoch 12/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0021\n",
      "Epoch 13/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0018\n",
      "Epoch 14/90\n",
      "22/22 [==============================] - 1s 32ms/step - loss: 0.0024\n",
      "Epoch 15/90\n",
      "22/22 [==============================] - 1s 27ms/step - loss: 0.0022\n",
      "Epoch 16/90\n",
      "22/22 [==============================] - 1s 40ms/step - loss: 0.0022\n",
      "Epoch 17/90\n",
      "22/22 [==============================] - 1s 37ms/step - loss: 0.0018\n",
      "Epoch 18/90\n",
      "22/22 [==============================] - 1s 35ms/step - loss: 0.0023\n",
      "Epoch 19/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0026\n",
      "Epoch 20/90\n",
      "22/22 [==============================] - 1s 31ms/step - loss: 0.0024\n",
      "Epoch 21/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0017\n",
      "Epoch 22/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0018\n",
      "Epoch 23/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0018\n",
      "Epoch 24/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0019\n",
      "Epoch 25/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0017\n",
      "Epoch 26/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0015\n",
      "Epoch 27/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0027\n",
      "Epoch 28/90\n",
      "22/22 [==============================] - 1s 30ms/step - loss: 0.0025\n",
      "Epoch 29/90\n",
      "22/22 [==============================] - 1s 31ms/step - loss: 0.0022\n",
      "Epoch 30/90\n",
      "22/22 [==============================] - 1s 32ms/step - loss: 0.0022\n",
      "Epoch 31/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0018\n",
      "Epoch 32/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0019\n",
      "Epoch 33/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0017\n",
      "Epoch 34/90\n",
      "22/22 [==============================] - 1s 27ms/step - loss: 0.0019\n",
      "Epoch 35/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0019\n",
      "Epoch 36/90\n",
      "22/22 [==============================] - 1s 30ms/step - loss: 0.0017\n",
      "Epoch 37/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0018\n",
      "Epoch 38/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0018\n",
      "Epoch 39/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0017\n",
      "Epoch 40/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0018\n",
      "Epoch 41/90\n",
      "22/22 [==============================] - 1s 39ms/step - loss: 0.0014\n",
      "Epoch 42/90\n",
      "22/22 [==============================] - 1s 35ms/step - loss: 0.0017\n",
      "Epoch 43/90\n",
      "22/22 [==============================] - 1s 38ms/step - loss: 0.0026\n",
      "Epoch 44/90\n",
      "22/22 [==============================] - 1s 32ms/step - loss: 0.0020\n",
      "Epoch 45/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0016\n",
      "Epoch 46/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0013\n",
      "Epoch 47/90\n",
      "22/22 [==============================] - 1s 27ms/step - loss: 0.0018: 0s - loss: 0.0\n",
      "Epoch 48/90\n",
      "22/22 [==============================] - 1s 31ms/step - loss: 0.0014: 0s - loss: 0.\n",
      "Epoch 49/90\n",
      "22/22 [==============================] - 1s 27ms/step - loss: 0.0018\n",
      "Epoch 50/90\n",
      "22/22 [==============================] - 1s 31ms/step - loss: 0.0014\n",
      "Epoch 51/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0014\n",
      "Epoch 52/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0015\n",
      "Epoch 53/90\n",
      "22/22 [==============================] - 1s 30ms/step - loss: 0.0011\n",
      "Epoch 54/90\n",
      "22/22 [==============================] - 1s 34ms/step - loss: 0.0020\n",
      "Epoch 55/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0024\n",
      "Epoch 56/90\n",
      "22/22 [==============================] - 1s 31ms/step - loss: 0.0023\n",
      "Epoch 57/90\n",
      "22/22 [==============================] - 1s 27ms/step - loss: 0.0019\n",
      "Epoch 58/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0016\n",
      "Epoch 59/90\n",
      "22/22 [==============================] - 1s 30ms/step - loss: 0.0015\n",
      "Epoch 60/90\n",
      "22/22 [==============================] - 1s 24ms/step - loss: 0.0013\n",
      "Epoch 61/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0020\n",
      "Epoch 62/90\n",
      "22/22 [==============================] - 1s 27ms/step - loss: 0.0017\n",
      "Epoch 63/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0020\n",
      "Epoch 64/90\n",
      "22/22 [==============================] - 1s 27ms/step - loss: 0.0017\n",
      "Epoch 65/90\n",
      "22/22 [==============================] - 1s 39ms/step - loss: 0.0011\n",
      "Epoch 66/90\n",
      "22/22 [==============================] - 1s 39ms/step - loss: 0.0022\n",
      "Epoch 67/90\n",
      "22/22 [==============================] - 1s 32ms/step - loss: 0.0025\n",
      "Epoch 68/90\n",
      "22/22 [==============================] - 1s 38ms/step - loss: 0.0025\n",
      "Epoch 69/90\n",
      "22/22 [==============================] - 1s 24ms/step - loss: 0.0014\n",
      "Epoch 70/90\n",
      "22/22 [==============================] - 1s 28ms/step - loss: 0.0022\n",
      "Epoch 71/90\n",
      "22/22 [==============================] - 1s 31ms/step - loss: 0.0018\n",
      "Epoch 72/90\n",
      "22/22 [==============================] - 1s 49ms/step - loss: 0.0030\n",
      "Epoch 73/90\n",
      "22/22 [==============================] - 1s 32ms/step - loss: 0.0013\n",
      "Epoch 74/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0013\n",
      "Epoch 75/90\n",
      "22/22 [==============================] - 1s 34ms/step - loss: 0.0017\n",
      "Epoch 76/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0019\n",
      "Epoch 77/90\n",
      "22/22 [==============================] - 1s 40ms/step - loss: 0.0030\n",
      "Epoch 78/90\n",
      "22/22 [==============================] - 1s 41ms/step - loss: 0.0015\n",
      "Epoch 79/90\n",
      "22/22 [==============================] - 1s 44ms/step - loss: 0.0016: 0s - loss: 0.00\n",
      "Epoch 80/90\n",
      "22/22 [==============================] - 1s 41ms/step - loss: 0.0016\n",
      "Epoch 81/90\n",
      "22/22 [==============================] - 1s 39ms/step - loss: 0.0020: 1s - lo\n",
      "Epoch 82/90\n",
      "22/22 [==============================] - 1s 40ms/step - loss: 0.0018\n",
      "Epoch 83/90\n",
      "22/22 [==============================] - 1s 31ms/step - loss: 0.0015\n",
      "Epoch 84/90\n",
      "22/22 [==============================] - 1s 36ms/step - loss: 0.0014: 0s - loss: 0.0\n",
      "Epoch 85/90\n",
      "22/22 [==============================] - 1s 33ms/step - loss: 0.0016\n",
      "Epoch 86/90\n",
      "22/22 [==============================] - 1s 35ms/step - loss: 0.0014\n",
      "Epoch 87/90\n",
      "22/22 [==============================] - 1s 36ms/step - loss: 0.0011\n",
      "Epoch 88/90\n",
      "22/22 [==============================] - 1s 38ms/step - loss: 0.0022\n",
      "Epoch 89/90\n",
      "22/22 [==============================] - 1s 30ms/step - loss: 0.0016\n",
      "Epoch 90/90\n",
      "22/22 [==============================] - 1s 29ms/step - loss: 0.0018\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1b23e39a0f0>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit_generator(generator,epochs=90)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_list = []\n",
    "\n",
    "batch = train[-n_input:].reshape((1, n_input, n_features))\n",
    "\n",
    "for i in range(n_input):   \n",
    "    pred_list.append(model.predict(batch)[0]) \n",
    "    batch = np.append(batch[:,1:,:],[[pred_list[i]]],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pandas.tseries.offsets import DateOffset\n",
    "add_dates = [df.index[-1] + DateOffset(months=x) for x in range(0,13) ]\n",
    "future_dates = pd.DataFrame(index=add_dates[1:],columns=df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_predict = pd.DataFrame(scaler.inverse_transform(pred_list),\n",
    "                          index=future_dates[-n_input:].index, columns=['Prediction'])\n",
    "\n",
    "df_proj = pd.concat([df,df_predict], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20, 5))\n",
    "plt.plot(df_proj.index, df_proj['AirPassengers'])\n",
    "plt.plot(df_proj.index, df_proj['Prediction'], color='r')\n",
    "plt.legend(loc='best', fontsize='xx-large')\n",
    "plt.xticks(fontsize=18)\n",
    "plt.yticks(fontsize=16)\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}