lstm_example_with_multivariate_time_series.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "EPkQkTqhjabZ"
   },
   "source": [
    "# LSTM example with multivariate time series\n",
    "- the feature matrix consists of lags of the target signal and lags of other features\n",
    "- keras LSTM model is used to forecast the target signal\n",
    "- for info on the dataset used and the methodology see: https://machinelearningmastery.com/multivariate-time-series-forecasting-lstms-keras/\n",
    "- it uses a data generator for train/valid/test data\n",
    "    - this will help for large time-series datasets with many features that can cause memory problems\n",
    "- don't need to add lags to dataframe; the generator will take care of it\n",
    "- if you want lagged aggregation, just use add columns with pandas' rolling() and mean() or other agg method\n",
    "- the model was trained on a free GPU from https://colab.research.google.com\n",
    "    - LSTMs don't get the same kind of training acceleration from GPUs as CNNs (and LSTMs can even run slower on GPUs)\n",
    "    - However this notebook uses CuDNNLSTM layers which provide accelerated preformance on GPUs\n",
    "* `TODO`: replace generator function with new  **keras.preprocessing.sequence.TimeseriesGenerator**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "DBL1nFtWjaba"
   },
   "source": [
    "# Import Dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "FLzogEmbjabc"
   },
   "outputs": [],
   "source": [
    "#from keras.preprocessing.sequence import TimeseriesGenerator\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from keras.models import Sequential\n",
    "from keras.layers import LSTM, CuDNNLSTM\n",
    "from keras.layers import Dropout\n",
    "from keras.models import Model\n",
    "from keras.layers import Input, Dense\n",
    "from keras.models import load_model\n",
    "from keras.callbacks import EarlyStopping, ModelCheckpoint\n",
    "import keras.backend as K\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from scipy import stats\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "s33SswXUjabh"
   },
   "source": [
    "## Read pollution.csv into pandas df\n",
    "- set date as index\n",
    "- display first 5 rows\n",
    "- the Air Pollution data is discussed here: https://machinelearningmastery.com/multivariate-time-series-forecasting-lstms-keras/\n",
    "- this is a dataset that reports on the weather and the level of pollution each hour for five years at the US embassy in Beijing, China.\n",
    "- we can use this data and frame a forecasting problem where, given the weather conditions and pollution for prior hours, we forecast the pollution at the next hour."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 225
    },
    "colab_type": "code",
    "id": "C8Nlpd_0jabi",
    "outputId": "1123c536-e4a7-43d5-c9e3-8b06eb631cc3"
   },
   "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>pollution</th>\n",
       "      <th>dew</th>\n",
       "      <th>temp</th>\n",
       "      <th>press</th>\n",
       "      <th>wnd_dir</th>\n",
       "      <th>wnd_spd</th>\n",
       "      <th>snow</th>\n",
       "      <th>rain</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2010-01-02 00:00:00</th>\n",
       "      <td>129.0</td>\n",
       "      <td>-16</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>1020.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>1.79</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02 01:00:00</th>\n",
       "      <td>148.0</td>\n",
       "      <td>-15</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>1020.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>2.68</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02 02:00:00</th>\n",
       "      <td>159.0</td>\n",
       "      <td>-11</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>1021.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>3.57</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02 03:00:00</th>\n",
       "      <td>181.0</td>\n",
       "      <td>-7</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>1022.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>5.36</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02 04:00:00</th>\n",
       "      <td>138.0</td>\n",
       "      <td>-7</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>1022.0</td>\n",
       "      <td>SE</td>\n",
       "      <td>6.25</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     pollution  dew  temp   press wnd_dir  wnd_spd  snow  rain\n",
       "date                                                                          \n",
       "2010-01-02 00:00:00      129.0  -16  -4.0  1020.0      SE     1.79     0     0\n",
       "2010-01-02 01:00:00      148.0  -15  -4.0  1020.0      SE     2.68     0     0\n",
       "2010-01-02 02:00:00      159.0  -11  -5.0  1021.0      SE     3.57     0     0\n",
       "2010-01-02 03:00:00      181.0   -7  -5.0  1022.0      SE     5.36     1     0\n",
       "2010-01-02 04:00:00      138.0   -7  -5.0  1022.0      SE     6.25     2     0"
      ]
     },
     "execution_count": 81,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('https://raw.githubusercontent.com/sagarmk/Forecasting-on-Air-pollution-with-RNN-LSTM/master/pollution.csv',index_col=0)\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "C2AnH2fjjabm",
    "outputId": "e8a1ca24-8849-4ab9-e5ff-90ffa4bd91c3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of rows =  43800\n"
     ]
    }
   ],
   "source": [
    "print('number of rows = ', df.shape[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot the data\n",
    "- for this step, we'll use the same code used here: https://machinelearningmastery.com/multivariate-time-series-forecasting-lstms-keras/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 349
    },
    "colab_type": "code",
    "id": "BrALNU0Sjabp",
    "outputId": "23eaefce-bf52-4e34-9327-336ba5878878"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeoAAAFMCAYAAAAEKP/JAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f55c23414a8>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# specify columns to plot\n",
    "values = df.values\n",
    "groups = [0, 1, 2, 3, 5, 6, 7]\n",
    "i = 1\n",
    "# plot each column\n",
    "plt.figure()\n",
    "for group in groups:\n",
    "    plt.subplot(len(groups), 1, i)\n",
    "    plt.plot(values[:, group])\n",
    "    plt.title(df.columns[group], y=0.5, loc='right')\n",
    "    i += 1\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "9W--5Nfzjabs"
   },
   "source": [
    "## One-hot-encode ['wnd_dir'] column\n",
    "- wind_dir column is categorical\n",
    "- use pd.get_dummies to preform this on pandas dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "0BGlX37Yjabt",
    "outputId": "aa5d83e0-6379-43be-f150-27190ec59a9a"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['SE', 'cv', 'NW', 'NE']"
      ]
     },
     "execution_count": 84,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(df['wnd_dir'].unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 118
    },
    "colab_type": "code",
    "id": "WFfphlGwjabw",
    "outputId": "ce081e6a-596a-4bfc-c22e-ca27960c8b66"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "wnd_dir\n",
       "NE     4996\n",
       "NW    14130\n",
       "SE    15290\n",
       "cv     9384\n",
       "Name: wnd_dir, dtype: int64"
      ]
     },
     "execution_count": 85,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby('wnd_dir')['wnd_dir'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "94pPvyDMjabz"
   },
   "outputs": [],
   "source": [
    "df = pd.concat([df,pd.get_dummies(df['wnd_dir'], prefix='wnd_dir')],axis=1)\n",
    "df.drop(['wnd_dir'],axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 225
    },
    "colab_type": "code",
    "id": "tuOHEVIYjab3",
    "outputId": "8f4d4bf7-ae63-4fa9-ba4b-1670ee70fd75"
   },
   "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>pollution</th>\n",
       "      <th>dew</th>\n",
       "      <th>temp</th>\n",
       "      <th>press</th>\n",
       "      <th>wnd_spd</th>\n",
       "      <th>snow</th>\n",
       "      <th>rain</th>\n",
       "      <th>wnd_dir_NE</th>\n",
       "      <th>wnd_dir_NW</th>\n",
       "      <th>wnd_dir_SE</th>\n",
       "      <th>wnd_dir_cv</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2010-01-02 00:00:00</th>\n",
       "      <td>129.0</td>\n",
       "      <td>-16</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>1020.0</td>\n",
       "      <td>1.79</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02 01:00:00</th>\n",
       "      <td>148.0</td>\n",
       "      <td>-15</td>\n",
       "      <td>-4.0</td>\n",
       "      <td>1020.0</td>\n",
       "      <td>2.68</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02 02:00:00</th>\n",
       "      <td>159.0</td>\n",
       "      <td>-11</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>1021.0</td>\n",
       "      <td>3.57</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02 03:00:00</th>\n",
       "      <td>181.0</td>\n",
       "      <td>-7</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>1022.0</td>\n",
       "      <td>5.36</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02 04:00:00</th>\n",
       "      <td>138.0</td>\n",
       "      <td>-7</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>1022.0</td>\n",
       "      <td>6.25</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     pollution  dew  temp   press  wnd_spd  snow  rain  \\\n",
       "date                                                                     \n",
       "2010-01-02 00:00:00      129.0  -16  -4.0  1020.0     1.79     0     0   \n",
       "2010-01-02 01:00:00      148.0  -15  -4.0  1020.0     2.68     0     0   \n",
       "2010-01-02 02:00:00      159.0  -11  -5.0  1021.0     3.57     0     0   \n",
       "2010-01-02 03:00:00      181.0   -7  -5.0  1022.0     5.36     1     0   \n",
       "2010-01-02 04:00:00      138.0   -7  -5.0  1022.0     6.25     2     0   \n",
       "\n",
       "                     wnd_dir_NE  wnd_dir_NW  wnd_dir_SE  wnd_dir_cv  \n",
       "date                                                                 \n",
       "2010-01-02 00:00:00           0           0           1           0  \n",
       "2010-01-02 01:00:00           0           0           1           0  \n",
       "2010-01-02 02:00:00           0           0           1           0  \n",
       "2010-01-02 03:00:00           0           0           1           0  \n",
       "2010-01-02 04:00:00           0           0           1           0  "
      ]
     },
     "execution_count": 87,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "9gEhTYz_nTuq"
   },
   "source": [
    "## we'll drop the date index\n",
    "- there's no missing timestamps, so we won't be needing it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "n0fce_CPnSDs"
   },
   "outputs": [],
   "source": [
    "df = df.reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "r2HSGqZ5jab6"
   },
   "source": [
    "## Split the Data\n",
    "- split the data sequentially mainting chronological order\n",
    "- split 80/20 training/test\n",
    "    - we will split-off a validation set from the training set later"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "T7BQqFjcjab7"
   },
   "outputs": [],
   "source": [
    "df_train = df.loc[:int(df.shape[0]*0.8),:]\n",
    "df_test = df.loc[int(df.shape[0]*0.8):,:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "zhK1_lGWktTu"
   },
   "source": [
    "## Plot the train/test split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 347
    },
    "colab_type": "code",
    "id": "JO0JTArRksHV",
    "outputId": "f0ef557e-c094-4854-cef5-e29b30597cb4"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAf8AAAFKCAYAAAAT9lZ4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nREREBsPyJyIiMpjveFTC4a5jgcUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f55c233d438>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(df_train.index,df_train['pollution'])\n",
    "plt.plot(df_test.index,df_test['pollution'])\n",
    "plt.ylabel('pollution',fontsize=18)\n",
    "plt.legend(['train','test'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Yy1c-Q0pjab-"
   },
   "source": [
    "## Normalize the Data\n",
    "- for available scalers see: http://scikit-learn.org/stable/auto_examples/preprocessing/plot_all_scaling.html#sphx-glr-auto-examples-preprocessing-plot-all-scaling-py\n",
    "- becareful of outliers when scaling!\n",
    "- perform the normalization on the pandas dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "colab_type": "code",
    "id": "pRVR4nPpjab-",
    "outputId": "8034ff8e-fe74-473b-f067-56d6a4d07194"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.6/dist-packages/sklearn/preprocessing/data.py:323: DataConversionWarning: Data with input dtype int64, float64 were all converted to float64 by MinMaxScaler.\n",
      "  return self.partial_fit(X, y)\n",
      "/usr/local/lib/python3.6/dist-packages/sklearn/preprocessing/data.py:323: DataConversionWarning: Data with input dtype int64, float64 were all converted to float64 by MinMaxScaler.\n",
      "  return self.partial_fit(X, y)\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>pollution</th>\n",
       "      <th>dew</th>\n",
       "      <th>temp</th>\n",
       "      <th>press</th>\n",
       "      <th>wnd_spd</th>\n",
       "      <th>snow</th>\n",
       "      <th>rain</th>\n",
       "      <th>wnd_dir_NE</th>\n",
       "      <th>wnd_dir_NW</th>\n",
       "      <th>wnd_dir_SE</th>\n",
       "      <th>wnd_dir_cv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.129779</td>\n",
       "      <td>0.278689</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.527273</td>\n",
       "      <td>0.002290</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.148893</td>\n",
       "      <td>0.295082</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.527273</td>\n",
       "      <td>0.003811</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.159960</td>\n",
       "      <td>0.360656</td>\n",
       "      <td>0.233333</td>\n",
       "      <td>0.545455</td>\n",
       "      <td>0.005332</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.182093</td>\n",
       "      <td>0.426230</td>\n",
       "      <td>0.233333</td>\n",
       "      <td>0.563636</td>\n",
       "      <td>0.008391</td>\n",
       "      <td>0.037037</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.138833</td>\n",
       "      <td>0.426230</td>\n",
       "      <td>0.233333</td>\n",
       "      <td>0.563636</td>\n",
       "      <td>0.009912</td>\n",
       "      <td>0.074074</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   pollution       dew      temp     press   wnd_spd      snow  rain  \\\n",
       "0   0.129779  0.278689  0.250000  0.527273  0.002290  0.000000   0.0   \n",
       "1   0.148893  0.295082  0.250000  0.527273  0.003811  0.000000   0.0   \n",
       "2   0.159960  0.360656  0.233333  0.545455  0.005332  0.000000   0.0   \n",
       "3   0.182093  0.426230  0.233333  0.563636  0.008391  0.037037   0.0   \n",
       "4   0.138833  0.426230  0.233333  0.563636  0.009912  0.074074   0.0   \n",
       "\n",
       "   wnd_dir_NE  wnd_dir_NW  wnd_dir_SE  wnd_dir_cv  \n",
       "0           0           0           1           0  \n",
       "1           0           0           1           0  \n",
       "2           0           0           1           0  \n",
       "3           0           0           1           0  \n",
       "4           0           0           1           0  "
      ]
     },
     "execution_count": 92,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train_norm = df_train.copy()\n",
    "df_test_norm = df_test.copy()\n",
    "\n",
    "scaler = MinMaxScaler(feature_range=(0, 1))\n",
    "df_train_norm.loc[:,:'rain']  = scaler.fit_transform(df_train.loc[:,:'rain'])\n",
    "df_test_norm.loc[:,:'rain']  = scaler.fit_transform(df_test.loc[:,:'rain'])\n",
    "\n",
    "df_train_norm.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "r9Wi9R-njacI"
   },
   "source": [
    "## Use a data generator\n",
    "- creates batches on the fly\n",
    "- good for data with large number of rows and features\n",
    "- don't need to perform lags on dataframe\n",
    "- the function below was adapted from https://github.com/fchollet/deep-learning-with-python-notebooks/blob/master/6.3-advanced-usage-of-recurrent-neural-networks.ipynb\n",
    "- note: Keras added a TimeseriesGenerator class in version 2.1.5 \n",
    "    - https://github.com/keras-team/keras/releases/tag/2.1.5\n",
    "    - keras.preprocessing.sequence.TimeseriesGenerator\n",
    "keras.preprocessing.sequence.TimeseriesGenerator\n",
    "It yields a tuple `(samples, targets)` where `samples` is one batch of input data and \n",
    "`targets` is the corresponding array of target temperatures. It takes the following arguments:\n",
    "\n",
    "* `dataframe`: pandas dataframe with feature matrix and target column\n",
    "* `target_name`: string of column with target values\n",
    "* `lookback`: How many timesteps back should our input data go.\n",
    "* `delay`: How many timesteps in the future should our target be.\n",
    "* `min_index` and `max_index`: Indices in the `data` array that delimit which timesteps to draw from. This is useful for keeping a segment \n",
    "of the data for validation and another one for testing.\n",
    "* `shuffle`: Whether to shuffle our samples or draw them in chronological order.\n",
    "* `batch_size`: The number of samples per batch.\n",
    "* `step`: The period, in timesteps, at which we sample data. We will set it 6 in order to draw one data point every hour."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Q289g6TtjacI"
   },
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "LS8F2IHBjacJ"
   },
   "outputs": [],
   "source": [
    "def generator(dataframe, target_name, lookback, delay, min_index, max_index,shuffle=False, batch_size=128, step=6):\n",
    "    \n",
    "    data = dataframe.values\n",
    "    data = data.astype(np.float32)\n",
    "    target_indx = df.columns.get_loc(target_name)\n",
    "    \n",
    "    if max_index is None:\n",
    "        max_index = len(data) - delay - 1\n",
    "    i = min_index + lookback\n",
    "    while 1:\n",
    "        if shuffle:\n",
    "            rows = np.random.randint(\n",
    "                min_index + lookback, max_index, size=batch_size)\n",
    "        else:\n",
    "            if i + batch_size >= max_index:\n",
    "                i = min_index + lookback\n",
    "            rows = np.arange(i, min(i + batch_size, max_index))\n",
    "            i += len(rows)\n",
    "\n",
    "        samples = np.zeros((len(rows),\n",
    "                           lookback // step,\n",
    "                           data.shape[-1]))\n",
    "        targets = np.zeros((len(rows),))\n",
    "        for j, row in enumerate(rows):\n",
    "            indices = range(rows[j] - lookback, rows[j], step)\n",
    "            samples[j] = data[indices]\n",
    "            targets[j] = data[rows[j] + delay][target_indx]\n",
    "        yield samples, targets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "k3V-ssIMjacN"
   },
   "source": [
    "## create train, validation, test data generators\n",
    "- we will 20% of the df_train to get the validation generator\n",
    "    - use the min_index,max_index arguments for slices from generator function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "YifTrpIhjacP"
   },
   "outputs": [],
   "source": [
    "#use previous 450 samples to predict next target ('pollution') samples\n",
    "lookback = 450 # window used for prediction\n",
    "delay = 1 # predict target one step ahead\n",
    "batch_size = 256\n",
    "\n",
    "train_gen = generator(dataframe = df_train_norm, \n",
    "                      target_name = 'pollution', \n",
    "                      lookback = lookback,\n",
    "                      delay = 1,\n",
    "                      min_index = 0,\n",
    "                      max_index = int(df_train.shape[0]*0.8),\n",
    "                      step = 1,\n",
    "                      shuffle = True,\n",
    "                      batch_size = batch_size)\n",
    "\n",
    "valid_gen = generator(dataframe = df_train_norm, \n",
    "                      target_name = 'pollution', \n",
    "                      lookback = lookback,\n",
    "                      delay = 1,\n",
    "                      min_index = int(df_train.shape[0]*0.8),\n",
    "                      max_index = None,\n",
    "                      step = 1,\n",
    "                      shuffle = False,\n",
    "                      batch_size = batch_size)\n",
    "\n",
    "# test_gen = generator(dataframe = df_test, \n",
    "#                      target_name = 'pollution', \n",
    "#                      lookback = lookback,\n",
    "#                      delay = 1,\n",
    "#                      min_index = 0,\n",
    "#                      max_index = None,\n",
    "#                      step = 1,\n",
    "#                      shuffle = False,\n",
    "#                      batch_size = batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 67
    },
    "colab_type": "code",
    "id": "R9krAVuMjacV",
    "outputId": "794dab90-3943-43b7-814f-3e351d3462cf"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "validation_steps =  11\n",
      "train_steps =  121\n",
      "test_steps =  32\n"
     ]
    }
   ],
   "source": [
    "# This is how many steps to draw from `val_gen`\n",
    "# in order to see the whole validation set:\n",
    "val_steps = int((df_train_norm.shape[0]*.1 - lookback) // batch_size)\n",
    "print('validation_steps = ', val_steps)\n",
    "# This is how many steps to draw from `train_gen`\n",
    "# in order to see the whole train set:\n",
    "train_steps = int((df_train_norm.shape[0]*.9 - lookback) // batch_size)\n",
    "print('train_steps = ', train_steps)\n",
    "# This is how many steps to draw from `train_gen`\n",
    "# in order to see the whole train set:\n",
    "test_steps = int((df_test_norm.shape[0] - lookback) // batch_size)\n",
    "print('test_steps = ', test_steps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "7shIrJr2jacZ"
   },
   "outputs": [],
   "source": [
    "def r2_keras(y_true, y_pred):\n",
    "    \"\"\"Coefficient of Determination \n",
    "    \"\"\"\n",
    "    SS_res =  K.sum(K.square( y_true - y_pred ))\n",
    "    SS_tot = K.sum(K.square( y_true - K.mean(y_true) ) )\n",
    "    return ( 1 - SS_res/(SS_tot + K.epsilon()) )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "w9FLFXkvjacb"
   },
   "source": [
    "## Create LSTM model\n",
    "- to increase the modeling speed, replace the LSTM model with CuDNNLSTM and train on a GPU\n",
    "- can also use lightwieght GRU layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "L3Y8_mFhjacc"
   },
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "# model.add(LSTM(50,\n",
    "#                activation='relu',\n",
    "#                return_sequences=True,\n",
    "#                stateful=False,\n",
    "#                input_shape=(None, df_train.shape[1])))\n",
    "# model.add(LSTM(50,\n",
    "#              activation='relu',\n",
    "#              return_sequences=False,\n",
    "#              stateful=False))\n",
    "\n",
    "model.add(CuDNNLSTM(64,\n",
    "               return_sequences=True,\n",
    "               stateful=False,\n",
    "               input_shape=(None, df_train_norm.shape[1])))\n",
    "model.add(CuDNNLSTM(32,\n",
    "               return_sequences=True,\n",
    "               stateful=False))\n",
    "model.add(CuDNNLSTM(16,\n",
    "             return_sequences=False,\n",
    "             stateful=False))\n",
    "#model.add(Dropout(0.4))\n",
    "\n",
    "model.add(Dense(1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "iZknUNitjach"
   },
   "source": [
    "## Fit the Model\n",
    "- to increase the modeling speed, replace the LSTM model with CuDNNLSTM and train on a GPU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 608
    },
    "colab_type": "code",
    "id": "oHO1dP-Ejach",
    "outputId": "df57b1bb-5a8b-4904-b33b-195ce8b7f27d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "121/121 [==============================] - 22s 180ms/step - loss: 0.0453 - r2_keras: 0.4901 - val_loss: 0.0298 - val_r2_keras: 0.4302\n",
      "Epoch 2/100\n",
      "121/121 [==============================] - 20s 166ms/step - loss: 0.0357 - r2_keras: 0.6518 - val_loss: 0.0267 - val_r2_keras: 0.6231\n",
      "Epoch 3/100\n",
      "121/121 [==============================] - 20s 167ms/step - loss: 0.0289 - r2_keras: 0.7603 - val_loss: 0.0240 - val_r2_keras: 0.7037\n",
      "Epoch 4/100\n",
      "121/121 [==============================] - 20s 167ms/step - loss: 0.0259 - r2_keras: 0.8019 - val_loss: 0.0199 - val_r2_keras: 0.6603\n",
      "Epoch 5/100\n",
      "121/121 [==============================] - 20s 166ms/step - loss: 0.0229 - r2_keras: 0.8267 - val_loss: 0.0221 - val_r2_keras: 0.7786\n",
      "Epoch 6/100\n",
      "121/121 [==============================] - 20s 166ms/step - loss: 0.0217 - r2_keras: 0.8357 - val_loss: 0.0186 - val_r2_keras: 0.6464\n",
      "Epoch 7/100\n",
      "121/121 [==============================] - 20s 167ms/step - loss: 0.0218 - r2_keras: 0.8269 - val_loss: 0.0202 - val_r2_keras: 0.8528\n",
      "Epoch 8/100\n",
      "121/121 [==============================] - 20s 166ms/step - loss: 0.0214 - r2_keras: 0.8306 - val_loss: 0.0176 - val_r2_keras: 0.7226\n",
      "Epoch 9/100\n",
      "121/121 [==============================] - 20s 166ms/step - loss: 0.0207 - r2_keras: 0.8353 - val_loss: 0.0197 - val_r2_keras: 0.7849\n",
      "Epoch 10/100\n",
      "121/121 [==============================] - 20s 166ms/step - loss: 0.0204 - r2_keras: 0.8434 - val_loss: 0.0181 - val_r2_keras: 0.7733\n",
      "Epoch 11/100\n",
      "121/121 [==============================] - 20s 165ms/step - loss: 0.0209 - r2_keras: 0.8302 - val_loss: 0.0166 - val_r2_keras: 0.7623\n",
      "Epoch 12/100\n",
      "121/121 [==============================] - 20s 166ms/step - loss: 0.0205 - r2_keras: 0.8398 - val_loss: 0.0200 - val_r2_keras: 0.7793\n",
      "Epoch 13/100\n",
      "121/121 [==============================] - 20s 166ms/step - loss: 0.0209 - r2_keras: 0.8284 - val_loss: 0.0165 - val_r2_keras: 0.7235\n",
      "Epoch 14/100\n",
      "121/121 [==============================] - 20s 166ms/step - loss: 0.0206 - r2_keras: 0.8390 - val_loss: 0.0195 - val_r2_keras: 0.8255\n",
      "Epoch 15/100\n",
      "121/121 [==============================] - 20s 167ms/step - loss: 0.0206 - r2_keras: 0.8381 - val_loss: 0.0167 - val_r2_keras: 0.6744\n",
      "Epoch 16/100\n",
      "121/121 [==============================] - 20s 167ms/step - loss: 0.0204 - r2_keras: 0.8382 - val_loss: 0.0205 - val_r2_keras: 0.8323\n",
      "Epoch 17/100\n",
      "121/121 [==============================] - 20s 166ms/step - loss: 0.0204 - r2_keras: 0.8381 - val_loss: 0.0168 - val_r2_keras: 0.7421\n"
     ]
    }
   ],
   "source": [
    "callbacks = [EarlyStopping(monitor='val_loss', min_delta = 0, patience=4),\n",
    "             ModelCheckpoint(filepath='lstm_model.h5', monitor='val_loss', save_best_only=True)]\n",
    "\n",
    "model.compile(loss='mae', optimizer='adam',metrics=[r2_keras])\n",
    "history = model.fit_generator(train_gen,\n",
    "                              steps_per_epoch=train_steps,\n",
    "                              epochs=100,\n",
    "                              shuffle=False,\n",
    "                              validation_data=valid_gen,\n",
    "                              validation_steps=val_steps,\n",
    "                              callbacks = callbacks)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Q_Rt3pdsjacl"
   },
   "source": [
    "## Get model predictions on df_test\n",
    "- get (X,y_truth) by calling train_gen using next\n",
    "- set the batch_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "rSGvlzaBjacm",
    "outputId": "31ee4fb5-0d9a-488d-958e-6eb95f49e93b"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8760"
      ]
     },
     "execution_count": 121,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test_norm.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "J-6pPDV0jacq"
   },
   "outputs": [],
   "source": [
    "test_gen = generator(dataframe = df_test_norm, \n",
    "                     target_name = 'pollution', \n",
    "                     lookback = lookback,\n",
    "                     delay = 1,\n",
    "                     min_index = 0,\n",
    "                     max_index = None,\n",
    "                     step = 1,\n",
    "                     shuffle = False,\n",
    "                     batch_size = df_test.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 67
    },
    "colab_type": "code",
    "id": "Wj5zSkm_jacs",
    "outputId": "41912754-626a-4ae5-a2fe-1c6e45a90aa4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_shape =  (8308, 450, 11)\n",
      "y_truth =  (8308,)\n",
      "y_pred =  (8308, 1)\n"
     ]
    }
   ],
   "source": [
    "(X, y_truth) = next(test_gen)\n",
    "\n",
    "y_pred = model.predict(X)\n",
    "\n",
    "\n",
    "print('X_shape = ',X.shape)\n",
    "print('y_truth = ',y_truth.shape)\n",
    "print('y_pred = ',y_pred.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "BouEZfKzjacv",
    "outputId": "7bdb0800-2cdd-438a-9f20-4d6608339890"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8308, 1)"
      ]
     },
     "execution_count": 124,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.expand_dims(y_truth,axis=1).shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ISUUZtfMjacz"
   },
   "source": [
    "### Need to remove normalization\n",
    "- since we preformed normlization transform on 7 cols of dataframe, we need to add dummy colums to preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "f80SdQaNjac0"
   },
   "outputs": [],
   "source": [
    "y_pred = np.concatenate((y_pred,np.zeros((y_pred.shape[0],6))),axis=1)\n",
    "y_pred = scaler.inverse_transform(y_pred)\n",
    "y_pred = y_pred[:,0]\n",
    "\n",
    "y_truth = np.concatenate((np.expand_dims(y_truth,axis=1),np.zeros((y_truth.shape[0],6))),axis=1)\n",
    "y_truth = scaler.inverse_transform(y_truth)\n",
    "y_truth = y_truth[:,0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "IsWb1_Hojac3"
   },
   "source": [
    "### Get R2 for this prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "lp5fVzFXjac4",
    "outputId": "3dfd642c-068c-44c2-cb44-baa29681ff50"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R2 =  0.8771529854688266\n"
     ]
    }
   ],
   "source": [
    "slope, intercept, r_value, p_value, std_err = stats.linregress(x=y_pred,y=y_truth)\n",
    "print('R2 = ',r_value*r_value)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 364
    },
    "colab_type": "code",
    "id": "dRn6j-5Vjac9",
    "outputId": "973429fa-d13b-4616-c6bc-dd4c8062bfad"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7f55b5cf84a8>"
      ]
     },
     "execution_count": 127,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAecAAAFKCAYAAAAnj5dkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEmKociaEEEJCDFXOhBBCSIj5P0rQJRgQdFVgAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f55b7975da0>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(x=y_pred,y=y_truth)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 347
    },
    "colab_type": "code",
    "id": "gi4-BFd4jadC",
    "outputId": "a6b08730-8e2c-40f3-a222-7e591ef516e9"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeoAAAFKCAYAAADScRzUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f55b5d51eb8>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(y_truth)\n",
    "plt.plot(y_pred)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "f0h37CyIjadH"
   },
   "source": [
    "## Let's try the evaluate_generator in keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "pQsfYFgcjadH",
    "outputId": "6bcc04f0-c1f7-4dd5-ce41-bd69f9534416"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R2 =  0.8759715557098389\n"
     ]
    }
   ],
   "source": [
    "(_,r2) = model.evaluate_generator(generator=test_gen,steps=1,workers=1)\n",
    "print('R2 = ',r2)"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "name": "LSTM_test_with_multivariate_time_series.ipynb",
   "provenance": [],
   "version": "0.3.2"
  },
  "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": 1
}