n4_model_RNN.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Project5: Niraj \n",
    "## Modeling Accident time series - Recurrent Neural Network (RNN) model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error, mean_absolute_percentage_error\n",
    "\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Dropout, GRU, LSTM\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.preprocessing.sequence import TimeseriesGenerator\n",
    "\n",
    "import tensorflow as tf\n",
    "\n",
    "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras.regularizers import l2\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/sarniraj/opt/anaconda3/lib/python3.8/site-packages/IPython/core/interactiveshell.py:3172: DtypeWarning: Columns (3,6,8,11,13,21,25,27,29,31,33,35,37,39,41,43,45,49,51,53) have mixed types.Specify dtype option on import or set low_memory=False.\n",
      "  has_raised = await self.run_ast_nodes(code_ast.body, cell_name,\n"
     ]
    }
   ],
   "source": [
    "# Read the saved csv from the earlier notebook and set the index to the date_time column\n",
    "\n",
    "df_accident_ts = pd.read_csv(\"./data/accident_ts_10years.csv\", parse_dates= ['date_time'], index_col=\"date_time\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_accident_ts.reset_index()\n",
    "df_accident_ts.sort_index(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_accident_ts.index.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create new df for modeling which has the columns we need, by day \n",
    "# Can do by day: 365*11 = 4000 records or by 3 hour windows = 32k records. Weekly and monthly will have too little data. \n",
    "# For now, only do daily.\n",
    "df_model_day = df_accident_ts.resample(\"D\").sum()[[\"fatals\", \"drunk_dr\", \"peds\", \"ve_total\", \"persons\" ]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>fatals</th>\n",
       "      <th>drunk_dr</th>\n",
       "      <th>peds</th>\n",
       "      <th>ve_total</th>\n",
       "      <th>persons</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date_time</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>2010-01-01</th>\n",
       "      <td>134</td>\n",
       "      <td>74</td>\n",
       "      <td>19</td>\n",
       "      <td>181</td>\n",
       "      <td>321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02</th>\n",
       "      <td>61</td>\n",
       "      <td>25</td>\n",
       "      <td>9</td>\n",
       "      <td>74</td>\n",
       "      <td>117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-03</th>\n",
       "      <td>54</td>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>78</td>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-04</th>\n",
       "      <td>70</td>\n",
       "      <td>13</td>\n",
       "      <td>9</td>\n",
       "      <td>104</td>\n",
       "      <td>136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-05</th>\n",
       "      <td>68</td>\n",
       "      <td>12</td>\n",
       "      <td>19</td>\n",
       "      <td>115</td>\n",
       "      <td>154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-12-27</th>\n",
       "      <td>98</td>\n",
       "      <td>27</td>\n",
       "      <td>21</td>\n",
       "      <td>128</td>\n",
       "      <td>201</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-12-28</th>\n",
       "      <td>98</td>\n",
       "      <td>12</td>\n",
       "      <td>17</td>\n",
       "      <td>140</td>\n",
       "      <td>198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-12-29</th>\n",
       "      <td>88</td>\n",
       "      <td>16</td>\n",
       "      <td>22</td>\n",
       "      <td>126</td>\n",
       "      <td>194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-12-30</th>\n",
       "      <td>96</td>\n",
       "      <td>26</td>\n",
       "      <td>20</td>\n",
       "      <td>129</td>\n",
       "      <td>164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-12-31</th>\n",
       "      <td>105</td>\n",
       "      <td>21</td>\n",
       "      <td>30</td>\n",
       "      <td>158</td>\n",
       "      <td>241</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4018 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            fatals  drunk_dr  peds  ve_total  persons\n",
       "date_time                                            \n",
       "2010-01-01     134        74    19       181      321\n",
       "2010-01-02      61        25     9        74      117\n",
       "2010-01-03      54        27     3        78      132\n",
       "2010-01-04      70        13     9       104      136\n",
       "2010-01-05      68        12    19       115      154\n",
       "...            ...       ...   ...       ...      ...\n",
       "2020-12-27      98        27    21       128      201\n",
       "2020-12-28      98        12    17       140      198\n",
       "2020-12-29      88        16    22       126      194\n",
       "2020-12-30      96        26    20       129      164\n",
       "2020-12-31     105        21    30       158      241\n",
       "\n",
       "[4018 rows x 5 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_model_day"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Let's Model! \n",
    "\n",
    "Recurrent Neural Networks are deep learning models used to solve problems with sequential input data. I will use the same five variables in the RNN model, as all the inputs are related to one another, to enable easier comparison between the two models. \n",
    "\n",
    "Similar to linear regression models, and unlike the time series, these variables are like the X variables used for predictions and the target variable is just the fatals. \n",
    "\n",
    "For RNN, we will use 4 variables - drunk_dr, peds, ve_total and persons to predict fatals.  \n",
    "\n",
    "I will use a small test size of 2% during the train test split and the Keras TimeseriesGenerator to feed 16 sequences (batchsize) of length of 3 (and 5) through our model. Our validation data set is small (only 81 rows) so choose a small batch size of 16. The length is the number of lag observations to be used in the input portion of each sample.   \n",
    "\n",
    "LSTM (long short-term memory) and GRU (Gated Recurrent Units) can alleviate the vanishing gradient problem which is a known shortcoming of RNNS especially with long time series extending over hundreds of periods. Along with Regularization using Dropout layers and EarlyStopping for optimization, we will attempt to get to a reasonably performing predictive model. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create X and y\n",
    "# maybe add other highly correlated cols like construction zone \n",
    "X = df_model_day.drop(columns = ['fatals'])\n",
    "y = df_model_day['fatals']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Train/Test Split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.02, shuffle = False) # keep in chronological order \n",
    "# Difficult to predict far out in the future, keep only last 2% as our test\n",
    "# The daily time periods spanning 11 years has a total of 4018 observations, of which the validation data of 2% is 81 observations. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((3937, 4), (81, 4), (3937,), (81,))"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, X_test.shape, y_train.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scale training and testing data\n",
    "sc =StandardScaler()\n",
    "X_train_sc =sc.fit_transform(X_train)\n",
    "X_test_sc = sc.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Keras TimeSeriesGenerator takes in a sequence of data-points gathered at equal intervals, along with time series parameters such as length of the sequences/windows, spacing between two sequence/windows, etc., to produce batches of timeseries inputs and targets.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create training sequences\n",
    "\n",
    "# 64 sequences of len=3 fed through our model at a time\n",
    "train_seq = TimeseriesGenerator( X_train_sc, y_train, length = 3, batch_size=16) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>drunk_dr</th>\n",
       "      <th>peds</th>\n",
       "      <th>ve_total</th>\n",
       "      <th>persons</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date_time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2010-01-01</th>\n",
       "      <td>74</td>\n",
       "      <td>19</td>\n",
       "      <td>181</td>\n",
       "      <td>321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02</th>\n",
       "      <td>25</td>\n",
       "      <td>9</td>\n",
       "      <td>74</td>\n",
       "      <td>117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-03</th>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>78</td>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-04</th>\n",
       "      <td>13</td>\n",
       "      <td>9</td>\n",
       "      <td>104</td>\n",
       "      <td>136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-05</th>\n",
       "      <td>12</td>\n",
       "      <td>19</td>\n",
       "      <td>115</td>\n",
       "      <td>154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-07</th>\n",
       "      <td>22</td>\n",
       "      <td>26</td>\n",
       "      <td>167</td>\n",
       "      <td>218</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-08</th>\n",
       "      <td>21</td>\n",
       "      <td>22</td>\n",
       "      <td>169</td>\n",
       "      <td>216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-09</th>\n",
       "      <td>34</td>\n",
       "      <td>31</td>\n",
       "      <td>229</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-10</th>\n",
       "      <td>53</td>\n",
       "      <td>23</td>\n",
       "      <td>233</td>\n",
       "      <td>353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-11</th>\n",
       "      <td>48</td>\n",
       "      <td>24</td>\n",
       "      <td>185</td>\n",
       "      <td>262</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3937 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            drunk_dr  peds  ve_total  persons\n",
       "date_time                                    \n",
       "2010-01-01        74    19       181      321\n",
       "2010-01-02        25     9        74      117\n",
       "2010-01-03        27     3        78      132\n",
       "2010-01-04        13     9       104      136\n",
       "2010-01-05        12    19       115      154\n",
       "...              ...   ...       ...      ...\n",
       "2020-10-07        22    26       167      218\n",
       "2020-10-08        21    22       169      216\n",
       "2020-10-09        34    31       229      303\n",
       "2020-10-10        53    23       233      353\n",
       "2020-10-11        48    24       185      262\n",
       "\n",
       "[3937 rows x 4 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[ 3.78561972,  0.04294179,  1.51767441,  2.4346476 ],\n",
       "        [-0.04499528, -1.50808429, -2.10584635, -1.62303778],\n",
       "        [ 0.11135636, -2.43869993, -1.97038763, -1.32467856]],\n",
       "\n",
       "       [[-0.04499528, -1.50808429, -2.10584635, -1.62303778],\n",
       "        [ 0.11135636, -2.43869993, -1.97038763, -1.32467856],\n",
       "        [-0.98310507, -1.50808429, -1.08990595, -1.2451161 ]],\n",
       "\n",
       "       [[ 0.11135636, -2.43869993, -1.97038763, -1.32467856],\n",
       "        [-0.98310507, -1.50808429, -1.08990595, -1.2451161 ],\n",
       "        [-1.06128089,  0.04294179, -0.71739447, -0.88708504]],\n",
       "\n",
       "       [[-0.98310507, -1.50808429, -1.08990595, -1.2451161 ],\n",
       "        [-1.06128089,  0.04294179, -0.71739447, -0.88708504],\n",
       "        [-1.53033579, -1.50808429, -2.03811699, -1.88161577]],\n",
       "\n",
       "       [[-1.06128089,  0.04294179, -0.71739447, -0.88708504],\n",
       "        [-1.53033579, -1.50808429, -2.03811699, -1.88161577],\n",
       "        [-0.82675344, -0.57746864, -0.34488299, -0.13124168]],\n",
       "\n",
       "       [[-1.53033579, -1.50808429, -2.03811699, -1.88161577],\n",
       "        [-0.82675344, -0.57746864, -0.34488299, -0.13124168],\n",
       "        [-0.27952272, -1.04277646, -1.73333487, -1.76227208]],\n",
       "\n",
       "       [[-0.82675344, -0.57746864, -0.34488299, -0.13124168],\n",
       "        [-0.27952272, -1.04277646, -1.73333487, -1.76227208],\n",
       "        [ 0.65858707, -0.73257125, -0.85285319, -0.19091352]],\n",
       "\n",
       "       [[-0.27952272, -1.04277646, -1.73333487, -1.76227208],\n",
       "        [ 0.65858707, -0.73257125, -0.85285319, -0.19091352],\n",
       "        [ 0.42405962, -1.97339211, -1.56401147, -1.18544425]],\n",
       "\n",
       "       [[ 0.65858707, -0.73257125, -0.85285319, -0.19091352],\n",
       "        [ 0.42405962, -1.97339211, -1.56401147, -1.18544425],\n",
       "        [-0.74857762, -0.42236603, -1.49628211, -1.48380347]],\n",
       "\n",
       "       [[ 0.42405962, -1.97339211, -1.56401147, -1.18544425],\n",
       "        [-0.74857762, -0.42236603, -1.49628211, -1.48380347],\n",
       "        [-0.43587436, -0.42236603, -1.22536467, -1.12577241]],\n",
       "\n",
       "       [[-0.74857762, -0.42236603, -1.49628211, -1.48380347],\n",
       "        [-0.43587436, -0.42236603, -1.22536467, -1.12577241],\n",
       "        [-0.90492926, -1.19787907, -0.81898851, -0.80752258]],\n",
       "\n",
       "       [[-0.43587436, -0.42236603, -1.22536467, -1.12577241],\n",
       "        [-0.90492926, -1.19787907, -0.81898851, -0.80752258],\n",
       "        [-0.43587436, -0.73257125,  0.40013997,  0.28646123]],\n",
       "\n",
       "       [[-0.90492926, -1.19787907, -0.81898851, -0.80752258],\n",
       "        [-0.43587436, -0.73257125,  0.40013997,  0.28646123],\n",
       "        [-0.35769854, -0.26726342, -1.32695871, -0.84730381]],\n",
       "\n",
       "       [[-0.43587436, -0.73257125,  0.40013997,  0.28646123],\n",
       "        [-0.35769854, -0.26726342, -1.32695871, -0.84730381],\n",
       "        [-0.12317109,  0.50824962, -0.92058255, -0.50916336]],\n",
       "\n",
       "       [[-0.35769854, -0.26726342, -1.32695871, -0.84730381],\n",
       "        [-0.12317109,  0.50824962, -0.92058255, -0.50916336],\n",
       "        [ 0.42405962, -0.26726342, -1.46241743, -0.5688352 ]],\n",
       "\n",
       "       [[-0.12317109,  0.50824962, -0.92058255, -0.50916336],\n",
       "        [ 0.42405962, -0.26726342, -1.46241743, -0.5688352 ],\n",
       "        [-0.82675344, -0.42236603, -1.02217659, -1.08599118]]])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create test sequences\n",
    "# one bacth of data (batch of sequences)\n",
    "train_seq[0][0]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([70, 68, 55, 84, 61, 84, 60, 68, 76, 70, 97, 75, 81, 85, 75, 77])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# First batch of target values \n",
    "train_seq[0][1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date_time\n",
       "2010-01-01    134\n",
       "2010-01-02     61\n",
       "2010-01-03     54\n",
       "2010-01-04     70\n",
       "2010-01-05     68\n",
       "             ... \n",
       "2020-10-07    114\n",
       "2020-10-08    115\n",
       "2020-10-09    147\n",
       "2020-10-10    168\n",
       "2020-10-11    126\n",
       "Freq: D, Name: fatals, Length: 3937, dtype: int64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "246"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# how many batches in the whole training set\n",
    "\n",
    "len(train_seq)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>drunk_dr</th>\n",
       "      <th>peds</th>\n",
       "      <th>ve_total</th>\n",
       "      <th>persons</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date_time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2010-01-01</th>\n",
       "      <td>74</td>\n",
       "      <td>19</td>\n",
       "      <td>181</td>\n",
       "      <td>321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-02</th>\n",
       "      <td>25</td>\n",
       "      <td>9</td>\n",
       "      <td>74</td>\n",
       "      <td>117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-03</th>\n",
       "      <td>27</td>\n",
       "      <td>3</td>\n",
       "      <td>78</td>\n",
       "      <td>132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-04</th>\n",
       "      <td>13</td>\n",
       "      <td>9</td>\n",
       "      <td>104</td>\n",
       "      <td>136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-05</th>\n",
       "      <td>12</td>\n",
       "      <td>19</td>\n",
       "      <td>115</td>\n",
       "      <td>154</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            drunk_dr  peds  ve_total  persons\n",
       "date_time                                    \n",
       "2010-01-01        74    19       181      321\n",
       "2010-01-02        25     9        74      117\n",
       "2010-01-03        27     3        78      132\n",
       "2010-01-04        13     9       104      136\n",
       "2010-01-05        12    19       115      154"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.head() # 5 columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>drunk_dr</th>\n",
       "      <th>peds</th>\n",
       "      <th>ve_total</th>\n",
       "      <th>persons</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date_time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-10-12</th>\n",
       "      <td>18</td>\n",
       "      <td>21</td>\n",
       "      <td>173</td>\n",
       "      <td>229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-13</th>\n",
       "      <td>19</td>\n",
       "      <td>21</td>\n",
       "      <td>155</td>\n",
       "      <td>194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-14</th>\n",
       "      <td>15</td>\n",
       "      <td>17</td>\n",
       "      <td>175</td>\n",
       "      <td>219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-15</th>\n",
       "      <td>18</td>\n",
       "      <td>25</td>\n",
       "      <td>161</td>\n",
       "      <td>201</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-10-16</th>\n",
       "      <td>27</td>\n",
       "      <td>39</td>\n",
       "      <td>190</td>\n",
       "      <td>265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-12-27</th>\n",
       "      <td>27</td>\n",
       "      <td>21</td>\n",
       "      <td>128</td>\n",
       "      <td>201</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-12-28</th>\n",
       "      <td>12</td>\n",
       "      <td>17</td>\n",
       "      <td>140</td>\n",
       "      <td>198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-12-29</th>\n",
       "      <td>16</td>\n",
       "      <td>22</td>\n",
       "      <td>126</td>\n",
       "      <td>194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-12-30</th>\n",
       "      <td>26</td>\n",
       "      <td>20</td>\n",
       "      <td>129</td>\n",
       "      <td>164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-12-31</th>\n",
       "      <td>21</td>\n",
       "      <td>30</td>\n",
       "      <td>158</td>\n",
       "      <td>241</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>81 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            drunk_dr  peds  ve_total  persons\n",
       "date_time                                    \n",
       "2020-10-12        18    21       173      229\n",
       "2020-10-13        19    21       155      194\n",
       "2020-10-14        15    17       175      219\n",
       "2020-10-15        18    25       161      201\n",
       "2020-10-16        27    39       190      265\n",
       "...              ...   ...       ...      ...\n",
       "2020-12-27        27    21       128      201\n",
       "2020-12-28        12    17       140      198\n",
       "2020-12-29        16    22       126      194\n",
       "2020-12-30        26    20       129      164\n",
       "2020-12-31        21    30       158      241\n",
       "\n",
       "[81 rows x 4 columns]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test # small percentage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Now do for test data, same len and batch_size as train data\n",
    "test_seq = TimeseriesGenerator( X_test_sc, y_test, length = 3, batch_size=16) \n",
    "len(test_seq)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Modeling time!\n",
    "---\n",
    "\n",
    "Now it's time to design our RNN model. We'll have:\n",
    "\n",
    "- 2 [`GRU`](https://www.tensorflow.org/api_docs/python/tf/keras/layers/GRU) layers\n",
    "- 1 `Dense` output layer\n",
    "\n",
    "We're doing a linear regression problem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3, 4)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Check shape of a single sequence of data (this will be the input shape for our model)\n",
    "\n",
    "train_seq[0][0][0].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-04-16 14:29:05.081520: I tensorflow/core/platform/cpu_feature_guard.cc:151] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n"
     ]
    }
   ],
   "source": [
    "# Design RNN\n",
    "model = Sequential()\n",
    "\n",
    "## First GRU layer returns hidden state, passes back into self\n",
    "model.add(GRU(16, input_shape = train_seq[0][0][0].shape, \n",
    "             return_sequences=True))\n",
    "## Second GRU layer passes outputs forward to next layer\n",
    "model.add(GRU(8, return_sequences=False))\n",
    "\n",
    "model.add(Dense(8, activation = 'relu'))\n",
    "\n",
    "model.add(Dense(1)) # output layer\n",
    "\n",
    "model.compile(\n",
    "    optimizer=\"adam\", ## Adam is the most commonly used optimized\n",
    "    loss=\"mean_squared_error\", # or 'mse' is the most common loss function for regression\n",
    "    metrics= [tf.keras.metrics.RootMeanSquaredError(),'mae', 'mape'], # Additional metrics to track - no 'rmse' so instead we can use tensorflow's RMSE function\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " gru (GRU)                   (None, 3, 16)             1056      \n",
      "                                                                 \n",
      " gru_1 (GRU)                 (None, 8)                 624       \n",
      "                                                                 \n",
      " dense (Dense)               (None, 8)                 72        \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 1)                 9         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 1,761\n",
      "Trainable params: 1,761\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "246/246 [==============================] - 10s 9ms/step - loss: 8369.0781 - root_mean_squared_error: 91.4827 - mae: 88.7819 - mape: 92.6434 - val_loss: 9315.3877 - val_root_mean_squared_error: 96.5163 - val_mae: 94.8270 - val_mape: 85.2812\n",
      "Epoch 2/50\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 5285.7168 - root_mean_squared_error: 72.7029 - mae: 69.1948 - mape: 71.0701 - val_loss: 5642.4922 - val_root_mean_squared_error: 75.1165 - val_mae: 72.9335 - val_mape: 64.9865\n",
      "Epoch 3/50\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 2590.0620 - root_mean_squared_error: 50.8927 - mae: 45.4572 - mape: 44.8516 - val_loss: 2712.2673 - val_root_mean_squared_error: 52.0794 - val_mae: 48.8782 - val_mape: 42.6880\n",
      "Epoch 4/50\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 1053.1005 - root_mean_squared_error: 32.4515 - mae: 25.9179 - mape: 24.6111 - val_loss: 1225.4331 - val_root_mean_squared_error: 35.0062 - val_mae: 30.3281 - val_mape: 25.6146\n",
      "Epoch 5/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 556.3615 - root_mean_squared_error: 23.5873 - mae: 18.2153 - mape: 18.3041 - val_loss: 716.7839 - val_root_mean_squared_error: 26.7728 - val_mae: 21.4753 - val_mape: 17.8253\n",
      "Epoch 6/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 472.0024 - root_mean_squared_error: 21.7256 - mae: 17.2916 - mape: 18.6190 - val_loss: 592.0918 - val_root_mean_squared_error: 24.3329 - val_mae: 18.9111 - val_mape: 15.6760\n",
      "Epoch 7/50\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 465.0101 - root_mean_squared_error: 21.5641 - mae: 17.3447 - mape: 19.0785 - val_loss: 563.4494 - val_root_mean_squared_error: 23.7371 - val_mae: 18.3178 - val_mape: 15.1961\n",
      "Epoch 8/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 464.3332 - root_mean_squared_error: 21.5484 - mae: 17.3475 - mape: 19.1159 - val_loss: 559.1777 - val_root_mean_squared_error: 23.6469 - val_mae: 18.2291 - val_mape: 15.1251\n",
      "Epoch 9/50\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 464.6140 - root_mean_squared_error: 21.5549 - mae: 17.3853 - mape: 19.2234 - val_loss: 560.5414 - val_root_mean_squared_error: 23.6758 - val_mae: 18.2576 - val_mape: 15.1479\n",
      "Epoch 10/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 464.2557 - root_mean_squared_error: 21.5466 - mae: 17.3484 - mape: 19.1403 - val_loss: 541.1282 - val_root_mean_squared_error: 23.2622 - val_mae: 17.8454 - val_mape: 14.8176\n",
      "Epoch 11/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 425.8065 - root_mean_squared_error: 20.6351 - mae: 16.4130 - mape: 17.7998 - val_loss: 477.4350 - val_root_mean_squared_error: 21.8503 - val_mae: 16.7964 - val_mape: 14.0587\n",
      "Epoch 12/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 317.1026 - root_mean_squared_error: 17.8074 - mae: 13.7870 - mape: 14.5539 - val_loss: 387.0381 - val_root_mean_squared_error: 19.6733 - val_mae: 15.3303 - val_mape: 13.1384\n",
      "Epoch 13/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 279.3535 - root_mean_squared_error: 16.7139 - mae: 12.9786 - mape: 13.8492 - val_loss: 308.2929 - val_root_mean_squared_error: 17.5583 - val_mae: 13.7507 - val_mape: 12.0947\n",
      "Epoch 14/50\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 260.3676 - root_mean_squared_error: 16.1359 - mae: 12.5761 - mape: 13.5611 - val_loss: 283.8689 - val_root_mean_squared_error: 16.8484 - val_mae: 13.0995 - val_mape: 11.6852\n",
      "Epoch 15/50\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 250.5849 - root_mean_squared_error: 15.8299 - mae: 12.3087 - mape: 13.2936 - val_loss: 266.1931 - val_root_mean_squared_error: 16.3154 - val_mae: 12.6367 - val_mape: 11.4128\n",
      "Epoch 16/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 246.5835 - root_mean_squared_error: 15.7030 - mae: 12.2316 - mape: 13.2775 - val_loss: 270.3901 - val_root_mean_squared_error: 16.4435 - val_mae: 12.7677 - val_mape: 11.5821\n",
      "Epoch 17/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 241.5857 - root_mean_squared_error: 15.5430 - mae: 12.0980 - mape: 13.1462 - val_loss: 259.1178 - val_root_mean_squared_error: 16.0971 - val_mae: 12.2551 - val_mape: 11.2463\n",
      "Epoch 18/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 238.7905 - root_mean_squared_error: 15.4528 - mae: 12.0270 - mape: 13.0840 - val_loss: 263.5323 - val_root_mean_squared_error: 16.2337 - val_mae: 12.3997 - val_mape: 11.3194\n",
      "Epoch 19/50\n",
      "246/246 [==============================] - 2s 9ms/step - loss: 236.0567 - root_mean_squared_error: 15.3641 - mae: 11.9869 - mape: 13.0401 - val_loss: 263.3778 - val_root_mean_squared_error: 16.2289 - val_mae: 12.4009 - val_mape: 11.2984\n",
      "Epoch 20/50\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 234.1566 - root_mean_squared_error: 15.3022 - mae: 11.9096 - mape: 12.9337 - val_loss: 267.1142 - val_root_mean_squared_error: 16.3436 - val_mae: 12.5651 - val_mape: 11.4072\n",
      "Epoch 21/50\n",
      "246/246 [==============================] - 2s 9ms/step - loss: 233.1711 - root_mean_squared_error: 15.2699 - mae: 11.8884 - mape: 12.9564 - val_loss: 257.0743 - val_root_mean_squared_error: 16.0335 - val_mae: 12.1932 - val_mape: 11.2389\n",
      "Epoch 22/50\n",
      "246/246 [==============================] - 2s 9ms/step - loss: 231.4704 - root_mean_squared_error: 15.2142 - mae: 11.8643 - mape: 12.9094 - val_loss: 255.0753 - val_root_mean_squared_error: 15.9711 - val_mae: 12.1659 - val_mape: 11.2089\n",
      "Epoch 23/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 230.1993 - root_mean_squared_error: 15.1723 - mae: 11.8328 - mape: 12.8959 - val_loss: 264.2562 - val_root_mean_squared_error: 16.2560 - val_mae: 12.4509 - val_mape: 11.2842\n",
      "Epoch 24/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 228.2363 - root_mean_squared_error: 15.1075 - mae: 11.7582 - mape: 12.8006 - val_loss: 263.0103 - val_root_mean_squared_error: 16.2176 - val_mae: 12.3491 - val_mape: 11.4064\n",
      "Epoch 25/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 226.1003 - root_mean_squared_error: 15.0366 - mae: 11.7363 - mape: 12.7852 - val_loss: 255.6410 - val_root_mean_squared_error: 15.9888 - val_mae: 12.2038 - val_mape: 11.1728\n",
      "Epoch 26/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 226.2336 - root_mean_squared_error: 15.0411 - mae: 11.7273 - mape: 12.7914 - val_loss: 255.4354 - val_root_mean_squared_error: 15.9823 - val_mae: 12.1754 - val_mape: 11.1022\n",
      "Epoch 27/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 226.2872 - root_mean_squared_error: 15.0428 - mae: 11.7239 - mape: 12.7871 - val_loss: 255.4592 - val_root_mean_squared_error: 15.9831 - val_mae: 12.1411 - val_mape: 11.1287\n",
      "Epoch 28/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 224.4857 - root_mean_squared_error: 14.9828 - mae: 11.6609 - mape: 12.7154 - val_loss: 257.3898 - val_root_mean_squared_error: 16.0434 - val_mae: 12.0779 - val_mape: 11.0862\n",
      "Epoch 29/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 222.6176 - root_mean_squared_error: 14.9204 - mae: 11.6475 - mape: 12.7149 - val_loss: 269.1460 - val_root_mean_squared_error: 16.4057 - val_mae: 12.6765 - val_mape: 11.4292\n",
      "Epoch 30/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 223.2134 - root_mean_squared_error: 14.9403 - mae: 11.6557 - mape: 12.7196 - val_loss: 266.5277 - val_root_mean_squared_error: 16.3257 - val_mae: 12.5971 - val_mape: 11.3336\n",
      "Epoch 31/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 223.0449 - root_mean_squared_error: 14.9347 - mae: 11.6455 - mape: 12.6921 - val_loss: 249.1574 - val_root_mean_squared_error: 15.7847 - val_mae: 11.9025 - val_mape: 11.0345\n",
      "Epoch 32/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 221.7947 - root_mean_squared_error: 14.8928 - mae: 11.6106 - mape: 12.6789 - val_loss: 251.8612 - val_root_mean_squared_error: 15.8701 - val_mae: 11.9335 - val_mape: 11.0830\n",
      "Epoch 33/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 220.4992 - root_mean_squared_error: 14.8492 - mae: 11.5614 - mape: 12.5986 - val_loss: 250.1981 - val_root_mean_squared_error: 15.8177 - val_mae: 11.8153 - val_mape: 10.9358\n",
      "Epoch 34/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 220.7000 - root_mean_squared_error: 14.8560 - mae: 11.6103 - mape: 12.6811 - val_loss: 263.9018 - val_root_mean_squared_error: 16.2451 - val_mae: 12.5131 - val_mape: 11.2820\n",
      "Epoch 35/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 219.5224 - root_mean_squared_error: 14.8163 - mae: 11.5619 - mape: 12.5951 - val_loss: 258.4959 - val_root_mean_squared_error: 16.0778 - val_mae: 12.2418 - val_mape: 11.2937\n",
      "Epoch 36/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 219.3269 - root_mean_squared_error: 14.8097 - mae: 11.5761 - mape: 12.6180 - val_loss: 248.2090 - val_root_mean_squared_error: 15.7547 - val_mae: 11.9497 - val_mape: 11.0792\n",
      "Epoch 37/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 218.9114 - root_mean_squared_error: 14.7957 - mae: 11.5518 - mape: 12.6190 - val_loss: 251.6205 - val_root_mean_squared_error: 15.8626 - val_mae: 11.9909 - val_mape: 11.1341\n",
      "Epoch 38/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 218.4695 - root_mean_squared_error: 14.7807 - mae: 11.5339 - mape: 12.5816 - val_loss: 260.4797 - val_root_mean_squared_error: 16.1394 - val_mae: 12.3536 - val_mape: 11.1851\n",
      "Epoch 39/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 219.2048 - root_mean_squared_error: 14.8056 - mae: 11.5788 - mape: 12.6634 - val_loss: 260.6173 - val_root_mean_squared_error: 16.1436 - val_mae: 12.4208 - val_mape: 11.3227\n",
      "Epoch 40/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 218.4792 - root_mean_squared_error: 14.7810 - mae: 11.5528 - mape: 12.6133 - val_loss: 256.9432 - val_root_mean_squared_error: 16.0294 - val_mae: 12.2167 - val_mape: 11.2854\n",
      "Epoch 41/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 217.6410 - root_mean_squared_error: 14.7527 - mae: 11.5386 - mape: 12.5916 - val_loss: 272.1413 - val_root_mean_squared_error: 16.4967 - val_mae: 12.8805 - val_mape: 11.5134\n",
      "Epoch 42/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 216.8663 - root_mean_squared_error: 14.7264 - mae: 11.5189 - mape: 12.5776 - val_loss: 288.4165 - val_root_mean_squared_error: 16.9828 - val_mae: 13.4265 - val_mape: 11.9242\n",
      "Epoch 43/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 216.8081 - root_mean_squared_error: 14.7244 - mae: 11.5365 - mape: 12.5827 - val_loss: 266.7534 - val_root_mean_squared_error: 16.3326 - val_mae: 12.5154 - val_mape: 11.3795\n",
      "Epoch 44/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 215.7960 - root_mean_squared_error: 14.6900 - mae: 11.4865 - mape: 12.5304 - val_loss: 253.1481 - val_root_mean_squared_error: 15.9106 - val_mae: 12.1807 - val_mape: 11.1576\n",
      "Epoch 45/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 215.9951 - root_mean_squared_error: 14.6968 - mae: 11.5184 - mape: 12.5884 - val_loss: 256.7208 - val_root_mean_squared_error: 16.0225 - val_mae: 12.2505 - val_mape: 11.1344\n",
      "Epoch 46/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 215.5933 - root_mean_squared_error: 14.6831 - mae: 11.5103 - mape: 12.5478 - val_loss: 263.6308 - val_root_mean_squared_error: 16.2367 - val_mae: 12.4682 - val_mape: 11.2202\n",
      "Epoch 47/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 215.4762 - root_mean_squared_error: 14.6791 - mae: 11.5348 - mape: 12.6043 - val_loss: 255.1558 - val_root_mean_squared_error: 15.9736 - val_mae: 12.1902 - val_mape: 11.1110\n",
      "Epoch 48/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 215.1405 - root_mean_squared_error: 14.6677 - mae: 11.4947 - mape: 12.5453 - val_loss: 253.9248 - val_root_mean_squared_error: 15.9350 - val_mae: 12.1679 - val_mape: 11.1811\n",
      "Epoch 49/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 214.5052 - root_mean_squared_error: 14.6460 - mae: 11.4642 - mape: 12.5095 - val_loss: 264.2709 - val_root_mean_squared_error: 16.2564 - val_mae: 12.5532 - val_mape: 11.3272\n",
      "Epoch 50/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 214.2165 - root_mean_squared_error: 14.6361 - mae: 11.4724 - mape: 12.5211 - val_loss: 254.7378 - val_root_mean_squared_error: 15.9605 - val_mae: 12.2441 - val_mape: 11.2645\n"
     ]
    }
   ],
   "source": [
    "# Fit to training sequences, validate with test sequences\n",
    "\n",
    "history = model.fit(train_seq, epochs = 50, verbose = 1, validation_data=test_seq)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y_test_baseline: 95.47320294640589, y_train_baseline_mean: 95.47320294640589, baseline_RMSE: 23.24519113999734, baseline MAPE: 0.148302794152187, baseline MAE: 0.148302794152187 \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1152x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Evaluate the model performance with various metrics\n",
    "\n",
    "y_train_preds = model.predict(train_seq)\n",
    "y_test_preds = model.predict(test_seq)\n",
    "\n",
    "# baseline\n",
    "y_test_baseline = [y_train.mean()]*len(y_test)\n",
    "\n",
    "baseline_RMSE = mean_squared_error(y_test, y_test_baseline)**0.5\n",
    "baseline_mean = y_train.mean()\n",
    "y_train_baseline = [y_train.mean()]*len(y_train)\n",
    "\n",
    "baseline_MAPE = mean_absolute_percentage_error(y_test, y_test_baseline)\n",
    "baseline_MAE = mean_absolute_error (y_test, y_test_baseline)\n",
    "\n",
    "print( f\"y_test_baseline: {y_test_baseline[0]}, y_train_baseline_mean: {baseline_mean}, baseline_RMSE: {baseline_RMSE}, baseline MAPE: {baseline_MAPE}, baseline MAE: {baseline_MAPE} \")\n",
    "\n",
    "#print( f\"Baseline R2 score: {metrics.r2_score(y_train, y_train_baseline)}\" )\n",
    "#print( f\"Baseline Training R2 score: {metrics.r2_score(y_test, y_test_baseline)}\" )\n",
    "\n",
    "#print( f\"Training R2 score: {metrics.r2_score( y_train[3:], y_train_preds)}\") # R2 scores are not relevant in RNN, but print anyway\n",
    "#print( f\"Test R2 score: {metrics.r2_score( y_test[3:], y_test_preds[:])}\") \n",
    "\n",
    "y_test_preds_df = pd.DataFrame(data = y_test_preds, index = y_test[3:].index, columns = [\"y_test_preds\"])\n",
    "\n",
    "fig, (ax1, ax2) = plt.subplots(1,2, figsize = (16,8))\n",
    "labels = ['y_test', 'preds']\n",
    "\n",
    "ax1.plot( y_test[3:18], color = 'blue', label=\"y_test\")\n",
    "ax1.plot( y_test_preds_df[:15], color = 'red', label=\"Predicted\")\n",
    "ax1.set_title(\"Predicted vs actual - first 15 days\")\n",
    "ax1.xaxis.set_ticks_position('none')\n",
    "ax1.set_ylabel(\"Count, per day\")\n",
    "ax1.tick_params(labelsize=8)\n",
    "ax1.legend()\n",
    "\n",
    "ax2.plot( y_test[3:], color = 'blue', label=\"y_test\")\n",
    "ax2.plot( y_test_preds_df[:], color = 'red', label=\"Predicted\")\n",
    "ax2.set_title(\"Predicted vs actual - All 81 days\")\n",
    "ax2.set_ylabel(\"Count, per day\")\n",
    "ax2.tick_params(labelsize=8)\n",
    "ax2.legend();\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Interpretation:**  \n",
    "The predictions are quite good, more so in the first 15 days. Over the total 81 days of predictions, the predictions don't quite match the peaks. In fact, towards the end they overshoot the data considerably. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 576x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, (ax1, ax2) = plt.subplots(1,2, figsize = (8,4))\n",
    "\n",
    "## Plot Training & Validation Loss\n",
    "ax1.plot(history.history['loss'], label=\"Training Loss\")\n",
    "ax1.plot(history.history['val_loss'], label=\"Testing Loss\")\n",
    "ax1.legend();\n",
    "\n",
    "## Plot Training & Validation Accuracy\n",
    "ax2.plot(history.history['root_mean_squared_error'], label=\"Training RMSE\")\n",
    "ax2.plot(history.history['val_root_mean_squared_error'], label=\"Testing RMSE\")\n",
    "ax2.legend();\n",
    "\n",
    "plt.suptitle(\"Evaluation of RNN model\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Interpretation:**  \n",
    "The graphs are almost perfect! The testing RMSE is marginally higher than the Training RMSE, showing the model is not overfit and should do a good job on predicting new unseen data. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Now try additional layers\n",
    "\n",
    "LSTM in this model, rest everything is the same"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# change GRU to LSTM\n",
    "\n",
    "model_2 = Sequential()\n",
    "\n",
    "## First LSTM layer returns hidden state, passes back into self\n",
    "model_2.add(LSTM(16, input_shape = train_seq[0][0][0].shape, \n",
    "             return_sequences=True))\n",
    "## Second LSTM layer passes outputs forward to next layer\n",
    "model_2.add(LSTM(8, return_sequences=False))\n",
    "\n",
    "model_2.add(Dense(8, activation = 'relu'))\n",
    "\n",
    "model_2.add(Dense(1))\n",
    "\n",
    "model_2.compile(\n",
    "    optimizer=\"adam\",\n",
    "    loss=\"mean_squared_error\", \n",
    "    metrics= [tf.keras.metrics.RootMeanSquaredError(),'mae']\n",
    ")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm (LSTM)                 (None, 3, 16)             1344      \n",
      "                                                                 \n",
      " lstm_1 (LSTM)               (None, 8)                 800       \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 8)                 72        \n",
      "                                                                 \n",
      " dense_3 (Dense)             (None, 1)                 9         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 2,225\n",
      "Trainable params: 2,225\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model_2.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "246/246 [==============================] - 7s 10ms/step - loss: 9286.3633 - root_mean_squared_error: 96.3658 - mae: 93.9061 - val_loss: 11534.5908 - val_root_mean_squared_error: 107.3992 - val_mae: 105.8827\n",
      "Epoch 2/50\n",
      "246/246 [==============================] - 2s 6ms/step - loss: 8103.2368 - root_mean_squared_error: 90.0180 - mae: 87.3865 - val_loss: 10155.9561 - val_root_mean_squared_error: 100.7768 - val_mae: 99.1595\n",
      "Epoch 3/50\n",
      "246/246 [==============================] - 2s 9ms/step - loss: 6884.1240 - root_mean_squared_error: 82.9706 - mae: 80.0853 - val_loss: 8639.2412 - val_root_mean_squared_error: 92.9475 - val_mae: 91.1917\n",
      "Epoch 4/50\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 5578.3008 - root_mean_squared_error: 74.6880 - mae: 71.4576 - val_loss: 7037.9502 - val_root_mean_squared_error: 83.8925 - val_mae: 81.9432\n",
      "Epoch 5/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 4287.4106 - root_mean_squared_error: 65.4783 - mae: 61.7120 - val_loss: 5487.0513 - val_root_mean_squared_error: 74.0746 - val_mae: 71.8598\n",
      "Epoch 6/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 3117.5439 - root_mean_squared_error: 55.8350 - mae: 51.5338 - val_loss: 4089.9832 - val_root_mean_squared_error: 63.9530 - val_mae: 61.3741\n",
      "Epoch 7/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 2155.4780 - root_mean_squared_error: 46.4271 - mae: 41.1453 - val_loss: 2939.5759 - val_root_mean_squared_error: 54.2179 - val_mae: 51.1507\n",
      "Epoch 8/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 1440.9459 - root_mean_squared_error: 37.9598 - mae: 31.6799 - val_loss: 2063.7385 - val_root_mean_squared_error: 45.4284 - val_mae: 41.7202\n",
      "Epoch 9/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 968.3132 - root_mean_squared_error: 31.1177 - mae: 24.5480 - val_loss: 1449.5231 - val_root_mean_squared_error: 38.0726 - val_mae: 33.6714\n",
      "Epoch 10/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 692.0643 - root_mean_squared_error: 26.3071 - mae: 20.2282 - val_loss: 1063.6564 - val_root_mean_squared_error: 32.6137 - val_mae: 27.6863\n",
      "Epoch 11/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 553.5169 - root_mean_squared_error: 23.5269 - mae: 18.1767 - val_loss: 824.8807 - val_root_mean_squared_error: 28.7207 - val_mae: 23.4448\n",
      "Epoch 12/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 494.4365 - root_mean_squared_error: 22.2359 - mae: 17.4515 - val_loss: 693.4742 - val_root_mean_squared_error: 26.3339 - val_mae: 21.0179\n",
      "Epoch 13/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 472.6543 - root_mean_squared_error: 21.7406 - mae: 17.2455 - val_loss: 617.5804 - val_root_mean_squared_error: 24.8512 - val_mae: 19.4424\n",
      "Epoch 14/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 463.5076 - root_mean_squared_error: 21.5292 - mae: 17.1904 - val_loss: 578.9110 - val_root_mean_squared_error: 24.0606 - val_mae: 18.6452\n",
      "Epoch 15/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 457.1216 - root_mean_squared_error: 21.3804 - mae: 17.1173 - val_loss: 559.2377 - val_root_mean_squared_error: 23.6482 - val_mae: 18.2397\n",
      "Epoch 16/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 449.7478 - root_mean_squared_error: 21.2073 - mae: 16.9566 - val_loss: 547.1532 - val_root_mean_squared_error: 23.3913 - val_mae: 17.9933\n",
      "Epoch 17/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 428.1470 - root_mean_squared_error: 20.6917 - mae: 16.3910 - val_loss: 516.5466 - val_root_mean_squared_error: 22.7277 - val_mae: 17.4662\n",
      "Epoch 18/50\n",
      "246/246 [==============================] - 2s 6ms/step - loss: 359.8273 - root_mean_squared_error: 18.9691 - mae: 14.7332 - val_loss: 450.0673 - val_root_mean_squared_error: 21.2148 - val_mae: 16.4253\n",
      "Epoch 19/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 313.1766 - root_mean_squared_error: 17.6968 - mae: 13.7373 - val_loss: 401.0764 - val_root_mean_squared_error: 20.0269 - val_mae: 15.6750\n",
      "Epoch 20/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 290.4501 - root_mean_squared_error: 17.0426 - mae: 13.2557 - val_loss: 370.0262 - val_root_mean_squared_error: 19.2361 - val_mae: 15.0359\n",
      "Epoch 21/50\n",
      "246/246 [==============================] - 2s 6ms/step - loss: 275.5373 - root_mean_squared_error: 16.5993 - mae: 12.9011 - val_loss: 326.0882 - val_root_mean_squared_error: 18.0579 - val_mae: 14.2131\n",
      "Epoch 22/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 265.0611 - root_mean_squared_error: 16.2807 - mae: 12.6512 - val_loss: 315.5058 - val_root_mean_squared_error: 17.7625 - val_mae: 13.9367\n",
      "Epoch 23/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 257.3703 - root_mean_squared_error: 16.0428 - mae: 12.4948 - val_loss: 305.1971 - val_root_mean_squared_error: 17.4699 - val_mae: 13.7956\n",
      "Epoch 24/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 251.9986 - root_mean_squared_error: 15.8745 - mae: 12.3564 - val_loss: 290.0126 - val_root_mean_squared_error: 17.0298 - val_mae: 13.3886\n",
      "Epoch 25/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 246.4041 - root_mean_squared_error: 15.6973 - mae: 12.2309 - val_loss: 277.6556 - val_root_mean_squared_error: 16.6630 - val_mae: 12.9924\n",
      "Epoch 26/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 242.8381 - root_mean_squared_error: 15.5833 - mae: 12.1609 - val_loss: 269.6580 - val_root_mean_squared_error: 16.4213 - val_mae: 12.7868\n",
      "Epoch 27/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 239.4113 - root_mean_squared_error: 15.4729 - mae: 12.0633 - val_loss: 274.7551 - val_root_mean_squared_error: 16.5757 - val_mae: 12.9273\n",
      "Epoch 28/50\n",
      "246/246 [==============================] - 2s 6ms/step - loss: 236.2505 - root_mean_squared_error: 15.3704 - mae: 11.9849 - val_loss: 284.9084 - val_root_mean_squared_error: 16.8792 - val_mae: 13.3586\n",
      "Epoch 29/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 234.9105 - root_mean_squared_error: 15.3268 - mae: 11.9513 - val_loss: 268.7167 - val_root_mean_squared_error: 16.3926 - val_mae: 12.7487\n",
      "Epoch 30/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 231.8360 - root_mean_squared_error: 15.2262 - mae: 11.9018 - val_loss: 260.2539 - val_root_mean_squared_error: 16.1324 - val_mae: 12.4968\n",
      "Epoch 31/50\n",
      "246/246 [==============================] - 2s 6ms/step - loss: 230.5888 - root_mean_squared_error: 15.1851 - mae: 11.8612 - val_loss: 257.4217 - val_root_mean_squared_error: 16.0444 - val_mae: 12.4471\n",
      "Epoch 32/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 229.0054 - root_mean_squared_error: 15.1329 - mae: 11.8310 - val_loss: 252.2287 - val_root_mean_squared_error: 15.8817 - val_mae: 12.2814\n",
      "Epoch 33/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 228.2167 - root_mean_squared_error: 15.1068 - mae: 11.7958 - val_loss: 252.4967 - val_root_mean_squared_error: 15.8901 - val_mae: 12.2746\n",
      "Epoch 34/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 226.0506 - root_mean_squared_error: 15.0350 - mae: 11.7527 - val_loss: 248.1655 - val_root_mean_squared_error: 15.7533 - val_mae: 12.2017\n",
      "Epoch 35/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 224.6570 - root_mean_squared_error: 14.9886 - mae: 11.7065 - val_loss: 254.7187 - val_root_mean_squared_error: 15.9599 - val_mae: 12.4396\n",
      "Epoch 36/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 223.5784 - root_mean_squared_error: 14.9525 - mae: 11.6864 - val_loss: 254.7160 - val_root_mean_squared_error: 15.9598 - val_mae: 12.4257\n",
      "Epoch 37/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 223.0143 - root_mean_squared_error: 14.9337 - mae: 11.6701 - val_loss: 253.7452 - val_root_mean_squared_error: 15.9294 - val_mae: 12.3867\n",
      "Epoch 38/50\n",
      "246/246 [==============================] - 2s 6ms/step - loss: 221.9980 - root_mean_squared_error: 14.8996 - mae: 11.6547 - val_loss: 249.0181 - val_root_mean_squared_error: 15.7803 - val_mae: 12.2335\n",
      "Epoch 39/50\n",
      "246/246 [==============================] - 2s 6ms/step - loss: 220.5522 - root_mean_squared_error: 14.8510 - mae: 11.6180 - val_loss: 258.2337 - val_root_mean_squared_error: 16.0697 - val_mae: 12.5383\n",
      "Epoch 40/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 219.8501 - root_mean_squared_error: 14.8273 - mae: 11.5927 - val_loss: 251.4008 - val_root_mean_squared_error: 15.8556 - val_mae: 12.3102\n",
      "Epoch 41/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 219.3266 - root_mean_squared_error: 14.8097 - mae: 11.5929 - val_loss: 250.8876 - val_root_mean_squared_error: 15.8394 - val_mae: 12.2367\n",
      "Epoch 42/50\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 218.3388 - root_mean_squared_error: 14.7763 - mae: 11.5610 - val_loss: 251.0320 - val_root_mean_squared_error: 15.8440 - val_mae: 12.2358\n",
      "Epoch 43/50\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 218.3914 - root_mean_squared_error: 14.7781 - mae: 11.5577 - val_loss: 248.8025 - val_root_mean_squared_error: 15.7735 - val_mae: 12.1266\n",
      "Epoch 44/50\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 217.7266 - root_mean_squared_error: 14.7556 - mae: 11.5600 - val_loss: 259.8417 - val_root_mean_squared_error: 16.1196 - val_mae: 12.5810\n",
      "Epoch 45/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 217.4232 - root_mean_squared_error: 14.7453 - mae: 11.5428 - val_loss: 253.3537 - val_root_mean_squared_error: 15.9171 - val_mae: 12.3305\n",
      "Epoch 46/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 217.0948 - root_mean_squared_error: 14.7341 - mae: 11.5453 - val_loss: 245.4448 - val_root_mean_squared_error: 15.6667 - val_mae: 12.0893\n",
      "Epoch 47/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 216.3080 - root_mean_squared_error: 14.7074 - mae: 11.5330 - val_loss: 251.3011 - val_root_mean_squared_error: 15.8525 - val_mae: 12.3767\n",
      "Epoch 48/50\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 215.7940 - root_mean_squared_error: 14.6899 - mae: 11.5198 - val_loss: 250.8377 - val_root_mean_squared_error: 15.8379 - val_mae: 12.2685\n",
      "Epoch 49/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 215.8764 - root_mean_squared_error: 14.6927 - mae: 11.5253 - val_loss: 251.2655 - val_root_mean_squared_error: 15.8514 - val_mae: 12.2295\n",
      "Epoch 50/50\n",
      "246/246 [==============================] - 2s 7ms/step - loss: 214.7767 - root_mean_squared_error: 14.6553 - mae: 11.4684 - val_loss: 266.8436 - val_root_mean_squared_error: 16.3353 - val_mae: 12.7492\n"
     ]
    }
   ],
   "source": [
    "history_2 = model_2.fit( train_seq, epochs = 50, verbose = 1, validation_data = test_seq)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Plot Training & Validation Loss\n",
    "plt.plot(history_2.history['loss'], label=\"Training Loss\")\n",
    "plt.plot(history_2.history['val_loss'], label=\"Testing Loss\")\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Plot Training & Validation Accuracy\n",
    "plt.plot(history_2.history['root_mean_squared_error'], label=\"Training RMSE\")\n",
    "plt.plot(history_2.history['val_root_mean_squared_error'], label=\"Testing RMSE\")\n",
    "plt.legend();\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y_test_baseline: 95.47320294640589, y_train_baseline_mean: 95.47320294640589, baseline_RMSE: 23.24519113999734 \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1008x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_train_preds = model_2.predict(train_seq)\n",
    "y_test_preds = model_2.predict(test_seq)\n",
    "\n",
    "# baseline\n",
    "y_test_baseline = [y_train.mean()]*len(y_test)\n",
    "\n",
    "baseline_RMSE = mean_squared_error(y_test, y_test_baseline)**0.5\n",
    "baseline_mean = y_train.mean()\n",
    "y_train_baseline = [y_train.mean()]*len(y_train)\n",
    "\n",
    "print( f\"y_test_baseline: {y_test_baseline[0]}, y_train_baseline_mean: {baseline_mean}, baseline_RMSE: {baseline_RMSE} \")\n",
    "\n",
    "\n",
    "#print( metrics.r2_score(y_train, y_train_baseline) )\n",
    "#print( metrics.r2_score(y_test, y_test_baseline) )\n",
    "\n",
    "#print( f\"Training R2 score: {metrics.r2_score( y_train[3:], y_train_preds)}\")\n",
    "#print( f\"Test R2 score: {metrics.r2_score( y_test[-15:], y_test_preds[-15:])}\") # last few days \n",
    "\n",
    "y_test_preds_df = pd.DataFrame(data = y_test_preds, index = y_test[3:].index, columns = [\"y_test_preds\"])\n",
    "\n",
    "fig, (ax1, ax2) = plt.subplots(1,2, figsize = (14,8))\n",
    "\n",
    "ax1.plot( y_test[3:18], color = 'blue')\n",
    "ax1.plot( y_test_preds_df[:15], color = 'red')\n",
    "ax1.set_title(\"Predicted vs actual - first 15 days\")\n",
    "ax1.xaxis.set_ticks_position('none')\n",
    "ax1.tick_params(labelsize=8)\n",
    "\n",
    "ax2.plot( y_test[3:], color = 'blue')\n",
    "ax2.plot( y_test_preds_df[:], color = 'red')\n",
    "ax2.set_title(\"Predicted vs actual - All 81 days\")\n",
    "ax2.tick_params(labelsize=8)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Next variation "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 3 - RNN with LSTM and Regularization\n",
    "\n",
    "model_3 = Sequential()\n",
    "\n",
    "## First LSTM layer returns hidden state, passes back into self\n",
    "model_3.add(LSTM(16, input_shape = train_seq[0][0][0].shape, \n",
    "             return_sequences=True))\n",
    "## Second LSTM layer passes outputs forward to next layer\n",
    "model_3.add(LSTM(8, return_sequences=False))\n",
    "\n",
    "model_3.add(Dense(8, activation = 'relu', kernel_regularizer=l2(0.001)))\n",
    "model_3.add(Dropout(0.2))\n",
    "\n",
    "model_3.add(Dense(1))\n",
    "\n",
    "model_3.compile(\n",
    "    optimizer=\"adam\",\n",
    "    loss=\"mean_squared_error\", \n",
    "    metrics= [tf.keras.metrics.RootMeanSquaredError(),'mae']\n",
    ")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm_2 (LSTM)               (None, 3, 16)             1344      \n",
      "                                                                 \n",
      " lstm_3 (LSTM)               (None, 8)                 800       \n",
      "                                                                 \n",
      " dense_4 (Dense)             (None, 8)                 72        \n",
      "                                                                 \n",
      " dropout (Dropout)           (None, 8)                 0         \n",
      "                                                                 \n",
      " dense_5 (Dense)             (None, 1)                 9         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 2,225\n",
      "Trainable params: 2,225\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model_3.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/40\n",
      "246/246 [==============================] - 6s 9ms/step - loss: 9053.1182 - root_mean_squared_error: 95.1478 - mae: 92.6208 - val_loss: 10347.7490 - val_root_mean_squared_error: 101.7238 - val_mae: 100.1216\n",
      "Epoch 2/40\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 6320.2705 - root_mean_squared_error: 79.5000 - mae: 76.2633 - val_loss: 7118.1265 - val_root_mean_squared_error: 84.3689 - val_mae: 82.4299\n",
      "Epoch 3/40\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 3961.6013 - root_mean_squared_error: 62.9410 - mae: 57.8470 - val_loss: 4411.8574 - val_root_mean_squared_error: 66.4216 - val_mae: 63.9414\n",
      "Epoch 4/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 2336.3318 - root_mean_squared_error: 48.3352 - mae: 40.7116 - val_loss: 2502.4678 - val_root_mean_squared_error: 50.0242 - val_mae: 46.6818\n",
      "Epoch 5/40\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 1548.2565 - root_mean_squared_error: 39.3472 - mae: 30.4329 - val_loss: 1514.4110 - val_root_mean_squared_error: 38.9147 - val_mae: 34.5965\n",
      "Epoch 6/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 1239.6449 - root_mean_squared_error: 35.2078 - mae: 27.5586 - val_loss: 1067.2778 - val_root_mean_squared_error: 32.6683 - val_mae: 27.7488\n",
      "Epoch 7/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 1150.4390 - root_mean_squared_error: 33.9172 - mae: 26.3726 - val_loss: 900.6294 - val_root_mean_squared_error: 30.0095 - val_mae: 24.8405\n",
      "Epoch 8/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 1179.7162 - root_mean_squared_error: 34.3461 - mae: 27.0724 - val_loss: 813.2036 - val_root_mean_squared_error: 28.5156 - val_mae: 23.2581\n",
      "Epoch 9/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 1151.5140 - root_mean_squared_error: 33.9330 - mae: 26.6478 - val_loss: 788.1612 - val_root_mean_squared_error: 28.0731 - val_mae: 22.8530\n",
      "Epoch 10/40\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 1115.7042 - root_mean_squared_error: 33.4012 - mae: 26.3800 - val_loss: 740.2656 - val_root_mean_squared_error: 27.2066 - val_mae: 22.0476\n",
      "Epoch 11/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 1035.9548 - root_mean_squared_error: 32.1852 - mae: 25.0942 - val_loss: 642.4836 - val_root_mean_squared_error: 25.3459 - val_mae: 20.4008\n",
      "Epoch 12/40\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 1016.2499 - root_mean_squared_error: 31.8776 - mae: 24.8368 - val_loss: 556.0931 - val_root_mean_squared_error: 23.5801 - val_mae: 18.7864\n",
      "Epoch 13/40\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 969.3789 - root_mean_squared_error: 31.1337 - mae: 24.1763 - val_loss: 507.1505 - val_root_mean_squared_error: 22.5184 - val_mae: 17.8576\n",
      "Epoch 14/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 962.2518 - root_mean_squared_error: 31.0190 - mae: 24.1051 - val_loss: 442.6395 - val_root_mean_squared_error: 21.0372 - val_mae: 16.5714\n",
      "Epoch 15/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 998.7044 - root_mean_squared_error: 31.6011 - mae: 24.5561 - val_loss: 475.8976 - val_root_mean_squared_error: 21.8134 - val_mae: 17.6192\n",
      "Epoch 16/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 970.9427 - root_mean_squared_error: 31.1587 - mae: 24.3885 - val_loss: 449.6775 - val_root_mean_squared_error: 21.2038 - val_mae: 17.0212\n",
      "Epoch 17/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 950.5838 - root_mean_squared_error: 30.8303 - mae: 23.9631 - val_loss: 406.0697 - val_root_mean_squared_error: 20.1493 - val_mae: 16.1018\n",
      "Epoch 18/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 910.2868 - root_mean_squared_error: 30.1697 - mae: 23.7031 - val_loss: 398.4794 - val_root_mean_squared_error: 19.9600 - val_mae: 15.9944\n",
      "Epoch 19/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 958.0744 - root_mean_squared_error: 30.9515 - mae: 24.3690 - val_loss: 395.0321 - val_root_mean_squared_error: 19.8734 - val_mae: 16.0363\n",
      "Epoch 20/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 962.2720 - root_mean_squared_error: 31.0192 - mae: 23.9742 - val_loss: 418.0826 - val_root_mean_squared_error: 20.4451 - val_mae: 16.7406\n",
      "Epoch 21/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 931.0156 - root_mean_squared_error: 30.5112 - mae: 24.0321 - val_loss: 397.0414 - val_root_mean_squared_error: 19.9239 - val_mae: 16.3231\n",
      "Epoch 22/40\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 911.8487 - root_mean_squared_error: 30.1955 - mae: 23.8473 - val_loss: 412.8573 - val_root_mean_squared_error: 20.3169 - val_mae: 16.7532\n",
      "Epoch 23/40\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 889.2200 - root_mean_squared_error: 29.8184 - mae: 23.4440 - val_loss: 341.2505 - val_root_mean_squared_error: 18.4708 - val_mae: 14.7816\n",
      "Epoch 24/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 909.2462 - root_mean_squared_error: 30.1524 - mae: 23.7255 - val_loss: 386.3600 - val_root_mean_squared_error: 19.6540 - val_mae: 16.1414\n",
      "Epoch 25/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 901.0096 - root_mean_squared_error: 30.0155 - mae: 23.7571 - val_loss: 381.5182 - val_root_mean_squared_error: 19.5304 - val_mae: 16.0459\n",
      "Epoch 26/40\n",
      "246/246 [==============================] - 2s 6ms/step - loss: 897.2289 - root_mean_squared_error: 29.9524 - mae: 23.2212 - val_loss: 454.3583 - val_root_mean_squared_error: 21.3138 - val_mae: 18.0089\n",
      "Epoch 27/40\n",
      "246/246 [==============================] - 2s 6ms/step - loss: 901.0145 - root_mean_squared_error: 30.0155 - mae: 23.4977 - val_loss: 323.3560 - val_root_mean_squared_error: 17.9798 - val_mae: 14.5088\n",
      "Epoch 28/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 911.5669 - root_mean_squared_error: 30.1908 - mae: 23.6504 - val_loss: 388.5270 - val_root_mean_squared_error: 19.7090 - val_mae: 16.3612\n",
      "Epoch 29/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 898.8906 - root_mean_squared_error: 29.9801 - mae: 23.5654 - val_loss: 351.9127 - val_root_mean_squared_error: 18.7571 - val_mae: 15.4145\n",
      "Epoch 30/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 904.7462 - root_mean_squared_error: 30.0776 - mae: 23.4401 - val_loss: 401.3820 - val_root_mean_squared_error: 20.0325 - val_mae: 16.6573\n",
      "Epoch 31/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 874.0992 - root_mean_squared_error: 29.5638 - mae: 23.0889 - val_loss: 372.5162 - val_root_mean_squared_error: 19.2985 - val_mae: 15.9765\n",
      "Epoch 32/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 899.4989 - root_mean_squared_error: 29.9903 - mae: 23.2109 - val_loss: 367.9268 - val_root_mean_squared_error: 19.1792 - val_mae: 15.8368\n",
      "Epoch 33/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 902.6745 - root_mean_squared_error: 30.0432 - mae: 23.6638 - val_loss: 366.1108 - val_root_mean_squared_error: 19.1318 - val_mae: 15.7965\n",
      "Epoch 34/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 877.3630 - root_mean_squared_error: 29.6189 - mae: 23.2455 - val_loss: 361.5483 - val_root_mean_squared_error: 19.0122 - val_mae: 15.7889\n",
      "Epoch 35/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 912.6794 - root_mean_squared_error: 30.2092 - mae: 23.6085 - val_loss: 328.2949 - val_root_mean_squared_error: 18.1166 - val_mae: 14.7839\n",
      "Epoch 36/40\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 868.8535 - root_mean_squared_error: 29.4749 - mae: 23.0496 - val_loss: 351.0257 - val_root_mean_squared_error: 18.7334 - val_mae: 15.5330\n",
      "Epoch 37/40\n",
      "246/246 [==============================] - 1s 5ms/step - loss: 869.8473 - root_mean_squared_error: 29.4917 - mae: 23.2937 - val_loss: 417.8434 - val_root_mean_squared_error: 20.4391 - val_mae: 17.1903\n",
      "Epoch 38/40\n",
      "246/246 [==============================] - 1s 6ms/step - loss: 887.6589 - root_mean_squared_error: 29.7922 - mae: 23.3630 - val_loss: 356.0038 - val_root_mean_squared_error: 18.8658 - val_mae: 15.7319\n",
      "Epoch 39/40\n",
      "246/246 [==============================] - 2s 6ms/step - loss: 867.6439 - root_mean_squared_error: 29.4544 - mae: 23.0930 - val_loss: 373.8032 - val_root_mean_squared_error: 19.3318 - val_mae: 16.0342\n",
      "Epoch 40/40\n",
      "246/246 [==============================] - 2s 8ms/step - loss: 884.6341 - root_mean_squared_error: 29.7414 - mae: 23.3040 - val_loss: 339.3452 - val_root_mean_squared_error: 18.4190 - val_mae: 15.1307\n"
     ]
    }
   ],
   "source": [
    "history_3 = model_3.fit(train_seq, epochs = 40, verbose = 1, validation_data = test_seq)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Plot Training & Validation Loss\n",
    "plt.plot(history_3.history['loss'], label=\"Training Loss\")\n",
    "plt.plot(history_3.history['val_loss'], label=\"Testing Loss\")\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Plot Training & Validation Accuracy\n",
    "plt.plot(history_3.history['root_mean_squared_error'], label=\"Training RMSE\")\n",
    "plt.plot(history_3.history['val_root_mean_squared_error'], label=\"Testing RMSE\")\n",
    "plt.legend();\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Interpretation:**  \n",
    "\n",
    "The graphs are not as good as the previous ones. \n",
    "After epoch 5, the Testing RMSE diverges significantly from the Training RMSE, and so does the loss\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5/5 [==============================] - 0s 4ms/step - loss: 339.3452 - root_mean_squared_error: 18.4190 - mae: 15.1307\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[339.34521484375, 18.418989181518555, 15.13070011138916]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_3.evaluate(test_seq)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y_test_baseline: 95.47320294640589, y_train_baseline_mean: 95.47320294640589, baseline_RMSE: 23.24519113999734 \n"
     ]
    }
   ],
   "source": [
    "# baseline\n",
    "y_test_baseline = [y_train.mean()]*len(y_test)\n",
    "\n",
    "baseline_RMSE = mean_squared_error(y_test, y_test_baseline)**0.5\n",
    "baseline_mean = y_train.mean()\n",
    "\n",
    "print( f\"y_test_baseline: {y_test_baseline[0]}, y_train_baseline_mean: {baseline_mean}, baseline_RMSE: {baseline_RMSE} \")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### For the 4th iteration, change the length (number of lags). We could potentially try with a smaller batchsize in the future.  \n",
    "\n",
    "Do EarlyStopping too"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 16 sequences of len=5 fed through our model at a time\n",
    "train_seq2 = TimeseriesGenerator( X_train_sc, y_train, length = 5, batch_size=16) \n",
    "test_seq2 = TimeseriesGenerator( X_test_sc, y_test, length = 5, batch_size=16) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_3\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm_4 (LSTM)               (None, 5, 16)             1344      \n",
      "                                                                 \n",
      " lstm_5 (LSTM)               (None, 8)                 800       \n",
      "                                                                 \n",
      " dense_6 (Dense)             (None, 8)                 72        \n",
      "                                                                 \n",
      " dropout_1 (Dropout)         (None, 8)                 0         \n",
      "                                                                 \n",
      " dense_7 (Dense)             (None, 1)                 9         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 2,225\n",
      "Trainable params: 2,225\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Design RNN - Using LSTM layers & Regularization\n",
    "model_4 = Sequential()\n",
    "\n",
    "model_4.add(LSTM(16, input_shape = train_seq2[0][0][0].shape,\n",
    "              return_sequences=True)) ## First LSTM layer returns hidden state, passes back into self\n",
    "model_4.add(LSTM(8, return_sequences=False)) ## Second LSTM layer passes outputs forward to next layer\n",
    "\n",
    "model_4.add(Dense(8, activation = 'relu', kernel_regularizer=l2(0.001)))\n",
    "model_4.add(Dropout(0.2))\n",
    "model_4.add(Dense(1))\n",
    "\n",
    "model_4.compile(\n",
    "    optimizer=\"adam\",\n",
    "    loss=\"mean_squared_error\", \n",
    "    metrics= [tf.keras.metrics.RootMeanSquaredError(),'mae']\n",
    ")\n",
    "\n",
    "model_4.summary()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.EarlyStopping at 0x7fd1f92d0f40>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "EarlyStopping(monitor = \"val_root_mean_squared_error\", mode = \"min\", verbose = 1, patience = 20, restore_best_weights=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/40\n",
      "246/246 [==============================] - 11s 18ms/step - loss: 8649.1680 - root_mean_squared_error: 93.0008 - mae: 90.2984 - val_loss: 9186.4775 - val_root_mean_squared_error: 95.8460 - val_mae: 94.1305\n",
      "Epoch 2/40\n",
      "246/246 [==============================] - 2s 9ms/step - loss: 4854.8350 - root_mean_squared_error: 69.6765 - mae: 65.4745 - val_loss: 4778.2632 - val_root_mean_squared_error: 69.1247 - val_mae: 66.7257\n",
      "Epoch 3/40\n",
      "246/246 [==============================] - 3s 10ms/step - loss: 2171.2419 - root_mean_squared_error: 46.5961 - mae: 39.4826 - val_loss: 2056.7351 - val_root_mean_squared_error: 45.3507 - val_mae: 41.6026\n",
      "Epoch 4/40\n",
      "246/246 [==============================] - 3s 11ms/step - loss: 1073.7402 - root_mean_squared_error: 32.7671 - mae: 25.7174 - val_loss: 1013.0956 - val_root_mean_squared_error: 31.8282 - val_mae: 26.7745\n",
      "Epoch 5/40\n",
      "246/246 [==============================] - 3s 11ms/step - loss: 840.8230 - root_mean_squared_error: 28.9958 - mae: 23.0280 - val_loss: 740.9884 - val_root_mean_squared_error: 27.2199 - val_mae: 21.8876\n",
      "Epoch 6/40\n",
      "246/246 [==============================] - 3s 10ms/step - loss: 822.9259 - root_mean_squared_error: 28.6855 - mae: 22.8027 - val_loss: 676.3163 - val_root_mean_squared_error: 26.0048 - val_mae: 20.6251\n",
      "Epoch 7/40\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 819.6959 - root_mean_squared_error: 28.6291 - mae: 22.7410 - val_loss: 667.8302 - val_root_mean_squared_error: 25.8411 - val_mae: 20.4512\n",
      "Epoch 8/40\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 843.2596 - root_mean_squared_error: 29.0377 - mae: 23.0232 - val_loss: 667.9878 - val_root_mean_squared_error: 25.8441 - val_mae: 20.4547\n",
      "Epoch 9/40\n",
      "246/246 [==============================] - 2s 9ms/step - loss: 807.8516 - root_mean_squared_error: 28.4215 - mae: 22.6790 - val_loss: 651.0999 - val_root_mean_squared_error: 25.5153 - val_mae: 20.1043\n",
      "Epoch 10/40\n",
      "246/246 [==============================] - 2s 9ms/step - loss: 794.4141 - root_mean_squared_error: 28.1841 - mae: 22.0114 - val_loss: 533.7678 - val_root_mean_squared_error: 23.1019 - val_mae: 18.3328\n",
      "Epoch 11/40\n",
      "246/246 [==============================] - 3s 11ms/step - loss: 695.7576 - root_mean_squared_error: 26.3758 - mae: 20.5952 - val_loss: 428.8517 - val_root_mean_squared_error: 20.7069 - val_mae: 16.4073\n",
      "Epoch 12/40\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 656.4328 - root_mean_squared_error: 25.6194 - mae: 19.9208 - val_loss: 324.6202 - val_root_mean_squared_error: 18.0150 - val_mae: 14.1700\n",
      "Epoch 13/40\n",
      "246/246 [==============================] - 3s 10ms/step - loss: 624.5194 - root_mean_squared_error: 24.9888 - mae: 19.4991 - val_loss: 320.5880 - val_root_mean_squared_error: 17.9027 - val_mae: 14.1867\n",
      "Epoch 14/40\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 618.8934 - root_mean_squared_error: 24.8759 - mae: 19.3049 - val_loss: 279.8095 - val_root_mean_squared_error: 16.7250 - val_mae: 13.0671\n",
      "Epoch 15/40\n",
      "246/246 [==============================] - 3s 11ms/step - loss: 564.1467 - root_mean_squared_error: 23.7500 - mae: 18.7757 - val_loss: 265.4604 - val_root_mean_squared_error: 16.2903 - val_mae: 12.8298\n",
      "Epoch 16/40\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 595.1446 - root_mean_squared_error: 24.3938 - mae: 19.1521 - val_loss: 260.0985 - val_root_mean_squared_error: 16.1249 - val_mae: 12.6541\n",
      "Epoch 17/40\n",
      "246/246 [==============================] - 3s 10ms/step - loss: 562.5787 - root_mean_squared_error: 23.7169 - mae: 18.7220 - val_loss: 260.8197 - val_root_mean_squared_error: 16.1472 - val_mae: 12.8572\n",
      "Epoch 18/40\n",
      "246/246 [==============================] - 3s 10ms/step - loss: 581.0689 - root_mean_squared_error: 24.1035 - mae: 18.8810 - val_loss: 262.0343 - val_root_mean_squared_error: 16.1847 - val_mae: 12.8731\n",
      "Epoch 19/40\n",
      "246/246 [==============================] - 3s 11ms/step - loss: 570.8930 - root_mean_squared_error: 23.8915 - mae: 18.7422 - val_loss: 243.9708 - val_root_mean_squared_error: 15.6167 - val_mae: 12.1533\n",
      "Epoch 20/40\n",
      "246/246 [==============================] - 3s 13ms/step - loss: 579.8486 - root_mean_squared_error: 24.0782 - mae: 18.8469 - val_loss: 244.8764 - val_root_mean_squared_error: 15.6456 - val_mae: 12.3646\n",
      "Epoch 21/40\n",
      "246/246 [==============================] - 3s 12ms/step - loss: 544.0391 - root_mean_squared_error: 23.3227 - mae: 18.4067 - val_loss: 260.3614 - val_root_mean_squared_error: 16.1329 - val_mae: 12.8717\n",
      "Epoch 22/40\n",
      "246/246 [==============================] - 3s 13ms/step - loss: 573.9816 - root_mean_squared_error: 23.9560 - mae: 18.6478 - val_loss: 262.6966 - val_root_mean_squared_error: 16.2051 - val_mae: 13.0456\n",
      "Epoch 23/40\n",
      "246/246 [==============================] - 3s 12ms/step - loss: 561.2424 - root_mean_squared_error: 23.6886 - mae: 18.4604 - val_loss: 245.6437 - val_root_mean_squared_error: 15.6701 - val_mae: 12.5510\n",
      "Epoch 24/40\n",
      "246/246 [==============================] - 3s 13ms/step - loss: 571.3496 - root_mean_squared_error: 23.9010 - mae: 18.8028 - val_loss: 237.5697 - val_root_mean_squared_error: 15.4103 - val_mae: 12.2087\n",
      "Epoch 25/40\n",
      "246/246 [==============================] - 3s 13ms/step - loss: 548.6438 - root_mean_squared_error: 23.4212 - mae: 18.1999 - val_loss: 242.8579 - val_root_mean_squared_error: 15.5809 - val_mae: 12.2285\n",
      "Epoch 26/40\n",
      "246/246 [==============================] - 3s 12ms/step - loss: 589.7563 - root_mean_squared_error: 24.2830 - mae: 18.9751 - val_loss: 229.4660 - val_root_mean_squared_error: 15.1451 - val_mae: 11.6717\n",
      "Epoch 27/40\n",
      "246/246 [==============================] - 4s 15ms/step - loss: 564.0298 - root_mean_squared_error: 23.7474 - mae: 18.6331 - val_loss: 239.9147 - val_root_mean_squared_error: 15.4862 - val_mae: 12.0743\n",
      "Epoch 28/40\n",
      "246/246 [==============================] - 3s 13ms/step - loss: 564.6121 - root_mean_squared_error: 23.7596 - mae: 18.5559 - val_loss: 226.4373 - val_root_mean_squared_error: 15.0447 - val_mae: 11.7393\n",
      "Epoch 29/40\n",
      "246/246 [==============================] - 3s 12ms/step - loss: 565.0258 - root_mean_squared_error: 23.7683 - mae: 18.6410 - val_loss: 257.0822 - val_root_mean_squared_error: 16.0308 - val_mae: 12.8099\n",
      "Epoch 30/40\n",
      "246/246 [==============================] - 3s 12ms/step - loss: 572.7675 - root_mean_squared_error: 23.9306 - mae: 18.7785 - val_loss: 241.3345 - val_root_mean_squared_error: 15.5319 - val_mae: 12.2876\n",
      "Epoch 31/40\n",
      "246/246 [==============================] - 3s 10ms/step - loss: 555.3548 - root_mean_squared_error: 23.5640 - mae: 18.4136 - val_loss: 231.2030 - val_root_mean_squared_error: 15.2022 - val_mae: 12.0151\n",
      "Epoch 32/40\n",
      "246/246 [==============================] - 3s 11ms/step - loss: 542.7933 - root_mean_squared_error: 23.2959 - mae: 18.2219 - val_loss: 249.8341 - val_root_mean_squared_error: 15.8031 - val_mae: 12.5983\n",
      "Epoch 33/40\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 541.8129 - root_mean_squared_error: 23.2748 - mae: 18.1335 - val_loss: 228.6591 - val_root_mean_squared_error: 15.1183 - val_mae: 11.9525\n",
      "Epoch 34/40\n",
      "246/246 [==============================] - 3s 10ms/step - loss: 561.2361 - root_mean_squared_error: 23.6884 - mae: 18.5280 - val_loss: 229.6176 - val_root_mean_squared_error: 15.1500 - val_mae: 11.9203\n",
      "Epoch 35/40\n",
      "246/246 [==============================] - 2s 10ms/step - loss: 558.2868 - root_mean_squared_error: 23.6261 - mae: 18.2946 - val_loss: 224.7111 - val_root_mean_squared_error: 14.9872 - val_mae: 11.8239\n",
      "Epoch 36/40\n",
      "246/246 [==============================] - 3s 12ms/step - loss: 548.6974 - root_mean_squared_error: 23.4222 - mae: 18.3129 - val_loss: 253.2678 - val_root_mean_squared_error: 15.9114 - val_mae: 12.8359\n",
      "Epoch 37/40\n",
      "246/246 [==============================] - 3s 13ms/step - loss: 554.4311 - root_mean_squared_error: 23.5443 - mae: 18.2825 - val_loss: 225.8764 - val_root_mean_squared_error: 15.0260 - val_mae: 12.0067\n",
      "Epoch 38/40\n",
      "246/246 [==============================] - 3s 11ms/step - loss: 544.7841 - root_mean_squared_error: 23.3386 - mae: 18.2146 - val_loss: 252.3398 - val_root_mean_squared_error: 15.8822 - val_mae: 12.7997\n",
      "Epoch 39/40\n",
      "246/246 [==============================] - 3s 12ms/step - loss: 558.0847 - root_mean_squared_error: 23.6218 - mae: 18.3563 - val_loss: 245.4121 - val_root_mean_squared_error: 15.6625 - val_mae: 12.7166\n",
      "Epoch 40/40\n",
      "246/246 [==============================] - 3s 11ms/step - loss: 535.8838 - root_mean_squared_error: 23.1471 - mae: 18.0518 - val_loss: 224.4345 - val_root_mean_squared_error: 14.9779 - val_mae: 11.8985\n"
     ]
    }
   ],
   "source": [
    "history_4 = model_4.fit(train_seq2, epochs = 40, verbose = 1, validation_data = test_seq2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5/5 [==============================] - 0s 3ms/step - loss: 224.4345 - root_mean_squared_error: 14.9779 - mae: 11.8985\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[224.4345245361328, 14.977896690368652, 11.898500442504883]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_4.evaluate(test_seq2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Plot Training & Validation Loss\n",
    "plt.plot(history_4.history['loss'], label=\"Training Loss\")\n",
    "plt.plot(history_4.history['val_loss'], label=\"Testing Loss\")\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Plot Training & Validation Accuracy\n",
    "plt.plot(history_4.history['root_mean_squared_error'], label=\"Training RMSE\")\n",
    "plt.plot(history_4.history['val_root_mean_squared_error'], label=\"Testing RMSE\")\n",
    "plt.legend();\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Interprettion**: \n",
    "\n",
    "Not very good at all. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Final model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_4\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " gru_2 (GRU)                 (None, 5, 128)            51456     \n",
      "                                                                 \n",
      " gru_3 (GRU)                 (None, 5, 64)             37248     \n",
      "                                                                 \n",
      " gru_4 (GRU)                 (None, 32)                9408      \n",
      "                                                                 \n",
      " dense_8 (Dense)             (None, 32)                1056      \n",
      "                                                                 \n",
      " dropout_2 (Dropout)         (None, 32)                0         \n",
      "                                                                 \n",
      " dense_9 (Dense)             (None, 16)                528       \n",
      "                                                                 \n",
      " dropout_3 (Dropout)         (None, 16)                0         \n",
      "                                                                 \n",
      " dense_10 (Dense)            (None, 8)                 136       \n",
      "                                                                 \n",
      " dense_11 (Dense)            (None, 1)                 9         \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 99,841\n",
      "Trainable params: 99,841\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model_5 = Sequential()\n",
    "\n",
    "model_5.add(GRU(128, input_shape = train_seq2[0][0][0].shape, return_sequences=True))\n",
    "model_5.add(GRU(64, return_sequences=True))\n",
    "model_5.add(GRU(32, return_sequences=False))\n",
    "\n",
    "model_5.add(Dense(32, activation = \"relu\"))\n",
    "model_5.add(Dropout(0.2))\n",
    "model_5.add(Dense(16, activation = \"relu\"))\n",
    "model_5.add(Dropout(0.2))\n",
    "model_5.add(Dense(8,  activation = \"relu\"))\n",
    "model_5.add(Dense(1))\n",
    "\n",
    "model_5.compile(\n",
    "    optimizer=\"adam\",\n",
    "    loss=\"mean_squared_error\", \n",
    "    metrics= [tf.keras.metrics.RootMeanSquaredError(),'mae']\n",
    ")\n",
    "\n",
    "model_5.summary()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "246/246 [==============================] - 13s 24ms/step - loss: 3461.6848 - root_mean_squared_error: 58.8361 - mae: 46.3198 - val_loss: 656.7572 - val_root_mean_squared_error: 25.6273 - val_mae: 20.2217\n",
      "Epoch 2/100\n",
      "246/246 [==============================] - 5s 20ms/step - loss: 864.8810 - root_mean_squared_error: 29.4089 - mae: 23.3431 - val_loss: 601.9168 - val_root_mean_squared_error: 24.5340 - val_mae: 19.0686\n",
      "Epoch 3/100\n",
      "246/246 [==============================] - 5s 19ms/step - loss: 841.7609 - root_mean_squared_error: 29.0131 - mae: 23.2114 - val_loss: 667.5823 - val_root_mean_squared_error: 25.8376 - val_mae: 20.4471\n",
      "Epoch 4/100\n",
      "246/246 [==============================] - 5s 19ms/step - loss: 815.6233 - root_mean_squared_error: 28.5591 - mae: 22.6691 - val_loss: 608.3774 - val_root_mean_squared_error: 24.6653 - val_mae: 19.2136\n",
      "Epoch 5/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 700.9817 - root_mean_squared_error: 26.4761 - mae: 20.8348 - val_loss: 375.5449 - val_root_mean_squared_error: 19.3790 - val_mae: 16.1077\n",
      "Epoch 6/100\n",
      "246/246 [==============================] - 4s 16ms/step - loss: 594.3010 - root_mean_squared_error: 24.3783 - mae: 19.2439 - val_loss: 287.4317 - val_root_mean_squared_error: 16.9538 - val_mae: 13.6029\n",
      "Epoch 7/100\n",
      "246/246 [==============================] - 4s 16ms/step - loss: 547.1525 - root_mean_squared_error: 23.3913 - mae: 18.4712 - val_loss: 240.5502 - val_root_mean_squared_error: 15.5097 - val_mae: 11.6534\n",
      "Epoch 8/100\n",
      "246/246 [==============================] - 4s 16ms/step - loss: 524.6032 - root_mean_squared_error: 22.9042 - mae: 18.0392 - val_loss: 268.2011 - val_root_mean_squared_error: 16.3768 - val_mae: 13.3051\n",
      "Epoch 9/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 513.4994 - root_mean_squared_error: 22.6605 - mae: 17.8534 - val_loss: 233.4414 - val_root_mean_squared_error: 15.2788 - val_mae: 11.6932\n",
      "Epoch 10/100\n",
      "246/246 [==============================] - 4s 18ms/step - loss: 513.8199 - root_mean_squared_error: 22.6676 - mae: 17.8620 - val_loss: 255.4059 - val_root_mean_squared_error: 15.9814 - val_mae: 12.9682\n",
      "Epoch 11/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 490.9555 - root_mean_squared_error: 22.1575 - mae: 17.4249 - val_loss: 248.6015 - val_root_mean_squared_error: 15.7671 - val_mae: 12.6586\n",
      "Epoch 12/100\n",
      "246/246 [==============================] - 4s 16ms/step - loss: 458.4687 - root_mean_squared_error: 21.4119 - mae: 16.8503 - val_loss: 317.9152 - val_root_mean_squared_error: 17.8302 - val_mae: 14.6084\n",
      "Epoch 13/100\n",
      "246/246 [==============================] - 4s 16ms/step - loss: 437.6779 - root_mean_squared_error: 20.9208 - mae: 16.3557 - val_loss: 230.1010 - val_root_mean_squared_error: 15.1691 - val_mae: 11.8648\n",
      "Epoch 14/100\n",
      "246/246 [==============================] - 4s 16ms/step - loss: 454.8521 - root_mean_squared_error: 21.3273 - mae: 16.7301 - val_loss: 359.2763 - val_root_mean_squared_error: 18.9546 - val_mae: 15.7259\n",
      "Epoch 15/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 409.6078 - root_mean_squared_error: 20.2388 - mae: 15.9554 - val_loss: 290.5123 - val_root_mean_squared_error: 17.0444 - val_mae: 14.1163\n",
      "Epoch 16/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 410.0795 - root_mean_squared_error: 20.2504 - mae: 15.8758 - val_loss: 374.8029 - val_root_mean_squared_error: 19.3598 - val_mae: 16.1033\n",
      "Epoch 17/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 402.0530 - root_mean_squared_error: 20.0513 - mae: 15.8272 - val_loss: 561.0876 - val_root_mean_squared_error: 23.6873 - val_mae: 20.6538\n",
      "Epoch 18/100\n",
      "246/246 [==============================] - 4s 16ms/step - loss: 381.2250 - root_mean_squared_error: 19.5250 - mae: 15.3379 - val_loss: 547.2593 - val_root_mean_squared_error: 23.3936 - val_mae: 20.2673\n",
      "Epoch 19/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 376.6199 - root_mean_squared_error: 19.4067 - mae: 15.1923 - val_loss: 486.9158 - val_root_mean_squared_error: 22.0662 - val_mae: 18.9121\n",
      "Epoch 20/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 357.0340 - root_mean_squared_error: 18.8953 - mae: 14.7670 - val_loss: 522.1536 - val_root_mean_squared_error: 22.8507 - val_mae: 19.6476\n",
      "Epoch 21/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 354.0307 - root_mean_squared_error: 18.8157 - mae: 14.7383 - val_loss: 499.3592 - val_root_mean_squared_error: 22.3463 - val_mae: 19.0095\n",
      "Epoch 22/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 346.7385 - root_mean_squared_error: 18.6209 - mae: 14.5677 - val_loss: 642.3856 - val_root_mean_squared_error: 25.3453 - val_mae: 22.3267\n",
      "Epoch 23/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 338.9709 - root_mean_squared_error: 18.4112 - mae: 14.4830 - val_loss: 587.3255 - val_root_mean_squared_error: 24.2348 - val_mae: 21.2439\n",
      "Epoch 24/100\n",
      "246/246 [==============================] - 4s 16ms/step - loss: 327.4203 - root_mean_squared_error: 18.0948 - mae: 14.1670 - val_loss: 661.9304 - val_root_mean_squared_error: 25.7280 - val_mae: 22.6361\n",
      "Epoch 25/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 330.5961 - root_mean_squared_error: 18.1823 - mae: 14.2634 - val_loss: 503.0297 - val_root_mean_squared_error: 22.4283 - val_mae: 19.1508\n",
      "Epoch 26/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 320.0026 - root_mean_squared_error: 17.8886 - mae: 13.9844 - val_loss: 603.1254 - val_root_mean_squared_error: 24.5586 - val_mae: 21.1826\n",
      "Epoch 27/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 317.0082 - root_mean_squared_error: 17.8047 - mae: 13.8513 - val_loss: 466.7253 - val_root_mean_squared_error: 21.6038 - val_mae: 18.3480\n",
      "Epoch 28/100\n",
      "246/246 [==============================] - 5s 19ms/step - loss: 313.7868 - root_mean_squared_error: 17.7140 - mae: 13.7453 - val_loss: 766.2665 - val_root_mean_squared_error: 27.6815 - val_mae: 24.5194\n",
      "Epoch 29/100\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 309.9115 - root_mean_squared_error: 17.6043 - mae: 13.8537 - val_loss: 565.6262 - val_root_mean_squared_error: 23.7829 - val_mae: 20.3265\n",
      "Epoch 30/100\n",
      "246/246 [==============================] - 5s 18ms/step - loss: 305.0364 - root_mean_squared_error: 17.4653 - mae: 13.7268 - val_loss: 659.2844 - val_root_mean_squared_error: 25.6765 - val_mae: 22.2690\n",
      "Epoch 31/100\n",
      "246/246 [==============================] - 4s 18ms/step - loss: 303.0757 - root_mean_squared_error: 17.4091 - mae: 13.6232 - val_loss: 639.0826 - val_root_mean_squared_error: 25.2801 - val_mae: 21.9389\n",
      "Epoch 32/100\n",
      "246/246 [==============================] - 4s 18ms/step - loss: 290.0278 - root_mean_squared_error: 17.0302 - mae: 13.3602 - val_loss: 692.6238 - val_root_mean_squared_error: 26.3177 - val_mae: 23.0278\n",
      "Epoch 33/100\n",
      "243/246 [============================>.] - ETA: 0s - loss: 302.8845 - root_mean_squared_error: 17.4036 - mae: 13.6125Restoring model weights from the end of the best epoch: 13.\n",
      "246/246 [==============================] - 4s 17ms/step - loss: 301.2241 - root_mean_squared_error: 17.3558 - mae: 13.5618 - val_loss: 772.4328 - val_root_mean_squared_error: 27.7927 - val_mae: 24.5597\n",
      "Epoch 33: early stopping\n"
     ]
    }
   ],
   "source": [
    "callbacks = [\n",
    "    EarlyStopping(monitor = \"val_root_mean_squared_error\", mode = \"min\", verbose = 1, patience = 20, restore_best_weights=True),\n",
    "    ModelCheckpoint(\"model_checkpoint.keras\", save_best_only=True, monitor = \"val_root_mean_squared_error\", mode = \"min\")\n",
    "]\n",
    "\n",
    "history_5 = model_5.fit(train_seq2, epochs = 100, verbose = 1, validation_data = test_seq2,\n",
    "                        callbacks = callbacks)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Interpretation**: \n",
    "\n",
    "Epoch 16 has the best score: \n",
    "root_mean_squared_error: 19.9212; val_root_mean_squared_error: 19.4927 \n",
    "\n",
    "The val_root_mean_squared_error is almost the same as the root_mean_squared_error\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Plot Training & Validation Loss\n",
    "plt.plot(history_5.history['loss'], label=\"Training Loss\")\n",
    "plt.plot(history_5.history['val_loss'], label=\"Testing Loss\")\n",
    "plt.legend();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Plot Training & Validation Accuracy\n",
    "plt.plot(history_5.history['root_mean_squared_error'], label=\"Training RMSE\")\n",
    "plt.plot(history_5.history['val_root_mean_squared_error'], label=\"Testing RMSE\")\n",
    "plt.legend();\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([55, 84, 61, 84, 60, 68, 76, 70, 97, 75, 81, 85, 75, 77, 61, 71])"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_seq2[0][1] # y_train - batch of 32 - doesn't have first 5 from y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1008x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_train_preds = model_5.predict(train_seq2)\n",
    "y_test_preds = model_5.predict(test_seq2)\n",
    "\n",
    "# baseline\n",
    "y_test_baseline = [y_train.mean()]*len(y_test)\n",
    "\n",
    "baseline_RMSE = mean_squared_error(y_test, y_test_baseline)**0.5\n",
    "baseline_mean = y_train.mean()\n",
    "y_train_baseline = [y_train.mean()]*len(y_train)\n",
    "\n",
    "#print( metrics.r2_score(y_train, y_train_baseline) ) \n",
    "#print( metrics.r2_score(y_test, y_test_baseline) )\n",
    "\n",
    "#print( f\"Training R2 score: {metrics.r2_score( y_train[5:], y_train_preds)}\")\n",
    "#print( f\"Test R2 score: {metrics.r2_score( y_test[-15:], y_test_preds[-15:])}\") # last few days \n",
    "\n",
    "y_test_preds_df = pd.DataFrame(data = y_test_preds, index = y_test[5:].index, columns = [\"y_test_preds\"])\n",
    "\n",
    "fig, (ax1, ax2) = plt.subplots(1,2, figsize = (14,8))\n",
    "\n",
    "ax1.plot( y_test[-15:], color = 'blue')\n",
    "ax1.plot( y_test_preds_df[-15:], color = 'red')\n",
    "ax1.set_title(\"Predicted vs actual - last 15 days\")\n",
    "ax1.xaxis.set_ticks_position('none')\n",
    "ax1.tick_params(labelsize=8)\n",
    "\n",
    "ax2.plot( y_test[5:], color = 'blue')\n",
    "ax2.plot( y_test_preds_df[:], color = 'red')\n",
    "ax2.set_title(\"Predicted vs actual - All 81 days\")\n",
    "ax2.tick_params(labelsize=8)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5/5 [==============================] - 0s 4ms/step - loss: 230.1010 - root_mean_squared_error: 15.1691 - mae: 11.8648\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[230.1009521484375, 15.169078826904297, 11.864835739135742]"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_5.evaluate(test_seq2) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1008x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, (ax1, ax2) = plt.subplots(1,2, figsize = (14,8))\n",
    "\n",
    "ax1.plot( y_test[-15:], color = 'blue')\n",
    "ax1.plot( y_test_preds_df[-15:], color = 'red')\n",
    "ax1.set_title(\"Predicted vs actual - last 15 days\")\n",
    "ax1.xaxis.set_ticks_position('none')\n",
    "ax1.tick_params(labelsize=8)\n",
    "\n",
    "ax2.plot( y_test[5:], color = 'blue')\n",
    "ax2.plot( y_test_preds_df[:], color = 'red')\n",
    "ax2.set_title(\"Predicted vs actual - All 81 days\")\n",
    "ax2.tick_params(labelsize=8)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### **Interpretation** : \n",
    "The predictions are accurate only for a short interval "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Summary of RNN modeling:**  \n",
    "\n",
    "Of the 5 models, the first one performed the best. \n",
    "\n",
    "The predictions are quite good, more so in the first few days. Over the total 81 days of predictions, the predictions don't quite match the peaks. In fact, towards the end they overshoot the data considerably.\n",
    "\n",
    "We used RMSE and MAPE to see how far off are predictions are in absolute numbers (in the same units as our target) and in percent terms. \n",
    "\n",
    "The RMSE and MAPE confirm the graphs. The Training RMSE of 14.69 is substantially better than baseline (23.25) and the Test RMSE (15.32) is only marginally higher than the Training RMSE. Our Test MAPE is 10% which indicates that our model’s forecast for fatalities is off by 10% on an average day or 15 fatalities. \n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "RNN model evaluation \n",
    "<img src=\"./images/RNN_model_evaluation.png\"  width=\"500\" height=\"100\">  \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### We're done with RNN! \n",
    "\n",
    "### Model Comparison \n",
    "**VAR:** Testing RMSE: 503/30=16, Testing MAPE: Fatals: 0.14, Peds: 0.08  \n",
    "**RNN:** Testing RMSE: 15.32, Testing MAPE Fatals: 0.10  \n",
    "\n",
    "Both did quite well, though RNN is lightly better on fatalities.   \n",
    "\n",
    "VAR gets to forecast on more than one variable and did a really good on forecasting pedestrian fatalities.   \n",
    "\n",
    "**Best Params:**  \n",
    "RNN:  \n",
    "2 GRU hidden layers of 16 and 8 nodes  \n",
    "Dense hidden layer with 8 nodes   \n",
    "Output layer (1 node, regression, relu activation)  \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### **Recommendations:**  \n",
    "Model is good to go!   \n",
    "Reduce the 90% drunk driver involved fatalities  \n",
    "- Sobriety checkpoints and roving patrols between 9pm-3am on weekends and holidays \n",
    "- Alcohol Ignition locks\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusions<a class=\"anchor\" id=\"Conclusions\"></a>\n",
    "RNN did a little bit better, though VAR gets to forecast on more than one variable. \n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Next Steps<a class=\"anchor\" id=\"Next-Steps\"></a>\n",
    "\n",
    "- Need to tune the models to improve performance (time consuming and CPU intensive)\n",
    "- Room for improvement: \n",
    "   - Try Linear Regression with exogenous variables – weather, distracted, construction zone\n",
    "   - VAR: investigate how to improve weekly and daily forecasts \n",
    "   - RNN: Can try weekly and monthly resampling to compare with VAR. Could try further hyper-parameter tuning\n",
    " - Clean up UserWarnings\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The End! "
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "992d1ff092356cfc279c6c1b3b4d471c27a08f3699f455fb2c398ff39d3fb7fe"
  },
  "kernelspec": {
   "display_name": "Python 3.8.8 64-bit ('base': conda)",
   "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": "3.8.8"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}