GeothermalDatathon_Thermal Power.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# PIVOT 2022 Geothermal Datathon: SPE - Gulf Coast Section\n",
    "## Data Modeling: Extracted Thermal Power\n",
    "\n",
    "\n",
    "#### - Team: \n",
    "GeotherML    \n",
    "    \n",
    "#### - Team Members:\n",
    "- Antonio Dagnino\n",
    "- Andres Calvo\n",
    "- Oliver Rojas\n",
    "- Yusuf Falola\n",
    "- Chandryl Paterne    \n",
    "    \n",
    "#### - Contents:\n",
    "\n",
    "   - **Data Modeling: Extracted Thermal Power**\n",
    "       - Import inputs and outputs\n",
    "       - Train / Test split\n",
    "       - Scale Data\n",
    "       - Create Neural Network\n",
    "       - Fit the Model\n",
    "       \n",
    "\n",
    "   - **Evaluation Metrics**\n",
    "       - MAE\n",
    "       - MSE\n",
    "       - RMSE\n",
    "       - Explained variance score\n",
    "       - Compare mean values per year: Predictions Vs Real\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Modeling: Extracted Thermal Power"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Data Modeling Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "from pylab import rcParams\n",
    "\n",
    "# Ignore harmless warnings\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Activation\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.callbacks import EarlyStopping\n",
    "\n",
    "\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error,explained_variance_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Customize Plots\n",
    "rcParams['figure.figsize'] = 14,8\n",
    "rcParams['axes.titlesize'] = 15\n",
    "rcParams['font.size']=10\n",
    "sns.set_style('whitegrid')\n",
    "\n",
    "#show all columns in Jupyter Notebook\n",
    "pd.set_option(\"display.max_columns\", None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs = pd.read_csv('inputs.csv',index_col=0)\n",
    "df_Thermal = pd.read_csv('df_Thermal.csv',index_col=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Train / Test split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(inputs.values,df_Thermal.values,test_size=0.2,random_state=101)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Scale Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = scaler.fit_transform(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(36388, 37)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9097, 37)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Create Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add(Dense(37,activation='relu'))\n",
    "model.add(Dense(37,activation='relu'))\n",
    "model.add(Dense(37,activation='relu'))\n",
    "model.add(Dense(37,activation='relu'))\n",
    "model.add(Dense(20))\n",
    "\n",
    "\n",
    "model.compile(optimizer='adam',loss='mse')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Fit the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# - Hyperparameter Tunning:\n",
    "\n",
    "batchsize = 128\n",
    "epochs = 400"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "early_stop = EarlyStopping(monitor='val_loss', mode='min', verbose=1, patience=25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 2140218240.0000 - val_loss: 1975122304.0000\n",
      "Epoch 2/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 2075990016.0000 - val_loss: 1956270976.0000\n",
      "Epoch 3/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 2037509120.0000 - val_loss: 1886526208.0000\n",
      "Epoch 4/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1961463680.0000 - val_loss: 1834884480.0000\n",
      "Epoch 5/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1930111872.0000 - val_loss: 1828304640.0000\n",
      "Epoch 6/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1919724032.0000 - val_loss: 1814640000.0000\n",
      "Epoch 7/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1915166208.0000 - val_loss: 1813219200.0000\n",
      "Epoch 8/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1911165056.0000 - val_loss: 1812591616.0000\n",
      "Epoch 9/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1908210688.0000 - val_loss: 1804509056.0000\n",
      "Epoch 10/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1905755776.0000 - val_loss: 1801911936.0000\n",
      "Epoch 11/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1902731008.0000 - val_loss: 1803911936.0000\n",
      "Epoch 12/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1900427520.0000 - val_loss: 1799730432.0000\n",
      "Epoch 13/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1897449728.0000 - val_loss: 1795781120.0000\n",
      "Epoch 14/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1893122048.0000 - val_loss: 1788406016.0000\n",
      "Epoch 15/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1890847232.0000 - val_loss: 1787241344.0000\n",
      "Epoch 16/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1885774720.0000 - val_loss: 1780133888.0000\n",
      "Epoch 17/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1881640320.0000 - val_loss: 1778872064.0000\n",
      "Epoch 18/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1878676480.0000 - val_loss: 1772246272.0000\n",
      "Epoch 19/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1872784128.0000 - val_loss: 1761586304.0000\n",
      "Epoch 20/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1865615104.0000 - val_loss: 1755063296.0000\n",
      "Epoch 21/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1858770304.0000 - val_loss: 1750178176.0000\n",
      "Epoch 22/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1850559232.0000 - val_loss: 1737275648.0000\n",
      "Epoch 23/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1840865792.0000 - val_loss: 1720671872.0000\n",
      "Epoch 24/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1830032128.0000 - val_loss: 1704583680.0000\n",
      "Epoch 25/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1816815488.0000 - val_loss: 1690174848.0000\n",
      "Epoch 26/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1805723264.0000 - val_loss: 1677821056.0000\n",
      "Epoch 27/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1796930688.0000 - val_loss: 1667077504.0000\n",
      "Epoch 28/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1787776768.0000 - val_loss: 1660078592.0000\n",
      "Epoch 29/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1776300544.0000 - val_loss: 1667027456.0000\n",
      "Epoch 30/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1758211456.0000 - val_loss: 1619374336.0000\n",
      "Epoch 31/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1743377152.0000 - val_loss: 1600398464.0000\n",
      "Epoch 32/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1723572480.0000 - val_loss: 1581609088.0000\n",
      "Epoch 33/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1703706496.0000 - val_loss: 1558624768.0000\n",
      "Epoch 34/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1682115456.0000 - val_loss: 1540626432.0000\n",
      "Epoch 35/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1660114944.0000 - val_loss: 1513767424.0000\n",
      "Epoch 36/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1640828672.0000 - val_loss: 1494240256.0000\n",
      "Epoch 37/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1620241664.0000 - val_loss: 1476780288.0000\n",
      "Epoch 38/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1601751168.0000 - val_loss: 1462420480.0000\n",
      "Epoch 39/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1589292032.0000 - val_loss: 1447962368.0000\n",
      "Epoch 40/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1574374528.0000 - val_loss: 1439943424.0000\n",
      "Epoch 41/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1560083456.0000 - val_loss: 1424629504.0000\n",
      "Epoch 42/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1547979008.0000 - val_loss: 1417224832.0000\n",
      "Epoch 43/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1535361536.0000 - val_loss: 1423146240.0000\n",
      "Epoch 44/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1529198336.0000 - val_loss: 1442569984.0000\n",
      "Epoch 45/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1517136000.0000 - val_loss: 1393739008.0000\n",
      "Epoch 46/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1506173952.0000 - val_loss: 1389075328.0000\n",
      "Epoch 47/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1493120384.0000 - val_loss: 1371269120.0000\n",
      "Epoch 48/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1488029824.0000 - val_loss: 1382758016.0000\n",
      "Epoch 49/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1482955520.0000 - val_loss: 1352218368.0000\n",
      "Epoch 50/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1472105216.0000 - val_loss: 1346613120.0000\n",
      "Epoch 51/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1463089536.0000 - val_loss: 1352750592.0000\n",
      "Epoch 52/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1454063872.0000 - val_loss: 1385016448.0000\n",
      "Epoch 53/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1451168384.0000 - val_loss: 1327382656.0000\n",
      "Epoch 54/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1438440960.0000 - val_loss: 1319731200.0000\n",
      "Epoch 55/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1430046976.0000 - val_loss: 1324827392.0000\n",
      "Epoch 56/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1423861376.0000 - val_loss: 1311137024.0000\n",
      "Epoch 57/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1415458816.0000 - val_loss: 1295969280.0000\n",
      "Epoch 58/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1403203968.0000 - val_loss: 1288384000.0000\n",
      "Epoch 59/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1398401152.0000 - val_loss: 1303004032.0000\n",
      "Epoch 60/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1394407936.0000 - val_loss: 1278141184.0000\n",
      "Epoch 61/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1375960192.0000 - val_loss: 1275019392.0000\n",
      "Epoch 62/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1367157888.0000 - val_loss: 1262782848.0000\n",
      "Epoch 63/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1358175104.0000 - val_loss: 1251661440.0000\n",
      "Epoch 64/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1345452288.0000 - val_loss: 1246952064.0000\n",
      "Epoch 65/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1337250176.0000 - val_loss: 1242342016.0000\n",
      "Epoch 66/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1330007808.0000 - val_loss: 1229956096.0000\n",
      "Epoch 67/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1321307264.0000 - val_loss: 1227003136.0000\n",
      "Epoch 68/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1311110784.0000 - val_loss: 1219014144.0000\n",
      "Epoch 69/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "285/285 [==============================] - 1s 2ms/step - loss: 1305909760.0000 - val_loss: 1256925056.0000\n",
      "Epoch 70/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1292896000.0000 - val_loss: 1219465344.0000\n",
      "Epoch 71/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1286232576.0000 - val_loss: 1194689024.0000\n",
      "Epoch 72/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1278908416.0000 - val_loss: 1194305408.0000\n",
      "Epoch 73/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1267566336.0000 - val_loss: 1180721920.0000\n",
      "Epoch 74/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1260193920.0000 - val_loss: 1180947712.0000\n",
      "Epoch 75/400\n",
      "285/285 [==============================] - ETA: 0s - loss: 1257155584.00 - 0s 2ms/step - loss: 1252855424.0000 - val_loss: 1194780544.0000\n",
      "Epoch 76/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1245810816.0000 - val_loss: 1175817600.0000\n",
      "Epoch 77/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1239183616.0000 - val_loss: 1183485056.0000\n",
      "Epoch 78/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1236560256.0000 - val_loss: 1206519936.0000\n",
      "Epoch 79/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1228804096.0000 - val_loss: 1159416832.0000\n",
      "Epoch 80/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1218418560.0000 - val_loss: 1155544064.0000\n",
      "Epoch 81/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1224080384.0000 - val_loss: 1275629440.0000\n",
      "Epoch 82/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1215606144.0000 - val_loss: 1150292992.0000\n",
      "Epoch 83/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1208323456.0000 - val_loss: 1162663552.0000\n",
      "Epoch 84/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1208448128.0000 - val_loss: 1143511296.0000\n",
      "Epoch 85/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1207102592.0000 - val_loss: 1161411584.0000\n",
      "Epoch 86/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1196666112.0000 - val_loss: 1157396736.0000\n",
      "Epoch 87/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1196905856.0000 - val_loss: 1150867584.0000\n",
      "Epoch 88/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1186200448.0000 - val_loss: 1136854272.0000\n",
      "Epoch 89/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1189538176.0000 - val_loss: 1130561920.0000\n",
      "Epoch 90/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1187736192.0000 - val_loss: 1132010752.0000\n",
      "Epoch 91/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1184692864.0000 - val_loss: 1145654400.0000\n",
      "Epoch 92/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1184984192.0000 - val_loss: 1127427584.0000\n",
      "Epoch 93/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1175454592.0000 - val_loss: 1134113408.0000\n",
      "Epoch 94/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1172027392.0000 - val_loss: 1122885632.0000\n",
      "Epoch 95/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1172710656.0000 - val_loss: 1192640000.0000\n",
      "Epoch 96/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1175373568.0000 - val_loss: 1122936960.0000\n",
      "Epoch 97/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1165898624.0000 - val_loss: 1115846144.0000\n",
      "Epoch 98/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1164773120.0000 - val_loss: 1146329472.0000\n",
      "Epoch 99/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1161192576.0000 - val_loss: 1115578368.0000\n",
      "Epoch 100/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1158607872.0000 - val_loss: 1115462144.0000\n",
      "Epoch 101/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1160762752.0000 - val_loss: 1118130816.0000\n",
      "Epoch 102/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1152806656.0000 - val_loss: 1141316992.0000\n",
      "Epoch 103/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1163551360.0000 - val_loss: 1127300864.0000\n",
      "Epoch 104/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1153544192.0000 - val_loss: 1128822912.0000\n",
      "Epoch 105/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1151906432.0000 - val_loss: 1108681472.0000\n",
      "Epoch 106/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1149263872.0000 - val_loss: 1110781696.0000\n",
      "Epoch 107/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1156359680.0000 - val_loss: 1119172608.0000\n",
      "Epoch 108/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1147292672.0000 - val_loss: 1119767040.0000\n",
      "Epoch 109/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1141890816.0000 - val_loss: 1107280640.0000\n",
      "Epoch 110/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1143130368.0000 - val_loss: 1104617472.0000\n",
      "Epoch 111/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1138289536.0000 - val_loss: 1140572544.0000\n",
      "Epoch 112/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1140467584.0000 - val_loss: 1110115072.0000\n",
      "Epoch 113/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1139626624.0000 - val_loss: 1133891328.0000\n",
      "Epoch 114/400\n",
      "285/285 [==============================] - 0s 1ms/step - loss: 1146040192.0000 - val_loss: 1112892160.0000\n",
      "Epoch 115/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1135767936.0000 - val_loss: 1113640960.0000\n",
      "Epoch 116/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1142304256.0000 - val_loss: 1103143040.0000\n",
      "Epoch 117/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1138690944.0000 - val_loss: 1119647360.0000\n",
      "Epoch 118/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1131564416.0000 - val_loss: 1101852928.0000\n",
      "Epoch 119/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1135368320.0000 - val_loss: 1107151232.0000\n",
      "Epoch 120/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1132540288.0000 - val_loss: 1106838912.0000\n",
      "Epoch 121/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1129244800.0000 - val_loss: 1104138368.0000\n",
      "Epoch 122/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1134617216.0000 - val_loss: 1102594176.0000\n",
      "Epoch 123/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1126100096.0000 - val_loss: 1102146560.0000\n",
      "Epoch 124/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1134353536.0000 - val_loss: 1102054016.0000\n",
      "Epoch 125/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1123480832.0000 - val_loss: 1109392896.0000\n",
      "Epoch 126/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1130809472.0000 - val_loss: 1093357824.0000\n",
      "Epoch 127/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1125461632.0000 - val_loss: 1106017664.0000\n",
      "Epoch 128/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1121061248.0000 - val_loss: 1090430208.0000\n",
      "Epoch 129/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1121731072.0000 - val_loss: 1088611072.0000\n",
      "Epoch 130/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1118171904.0000 - val_loss: 1092081536.0000\n",
      "Epoch 131/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1118438272.0000 - val_loss: 1086390528.0000\n",
      "Epoch 132/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1116912896.0000 - val_loss: 1091091840.0000\n",
      "Epoch 133/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1115275264.0000 - val_loss: 1090862592.0000\n",
      "Epoch 134/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1122636032.0000 - val_loss: 1097834624.0000\n",
      "Epoch 135/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1114448000.0000 - val_loss: 1087944320.0000\n",
      "Epoch 136/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1112118528.0000 - val_loss: 1086463616.0000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 137/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1109698816.0000 - val_loss: 1082658304.0000\n",
      "Epoch 138/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1110233216.0000 - val_loss: 1079456768.0000\n",
      "Epoch 139/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1109368320.0000 - val_loss: 1099905920.0000\n",
      "Epoch 140/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1112467328.0000 - val_loss: 1078252672.0000\n",
      "Epoch 141/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1108533248.0000 - val_loss: 1116706560.0000\n",
      "Epoch 142/400\n",
      "285/285 [==============================] - ETA: 0s - loss: 1103186304.00 - 0s 2ms/step - loss: 1099404544.0000 - val_loss: 1083957248.0000\n",
      "Epoch 143/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1104008832.0000 - val_loss: 1090687488.0000\n",
      "Epoch 144/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1106150912.0000 - val_loss: 1080411136.0000\n",
      "Epoch 145/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1105672192.0000 - val_loss: 1077552000.0000\n",
      "Epoch 146/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1101860480.0000 - val_loss: 1119860096.0000\n",
      "Epoch 147/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1104028544.0000 - val_loss: 1086050944.0000\n",
      "Epoch 148/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1097294336.0000 - val_loss: 1075451904.0000\n",
      "Epoch 149/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1098023296.0000 - val_loss: 1194573056.0000\n",
      "Epoch 150/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1110111360.0000 - val_loss: 1074078592.0000\n",
      "Epoch 151/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1092881152.0000 - val_loss: 1075187840.0000\n",
      "Epoch 152/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1092361472.0000 - val_loss: 1080256768.0000\n",
      "Epoch 153/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1091145344.0000 - val_loss: 1115094528.0000\n",
      "Epoch 154/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1099907072.0000 - val_loss: 1078250752.0000\n",
      "Epoch 155/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1087334400.0000 - val_loss: 1072094336.0000\n",
      "Epoch 156/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1094828416.0000 - val_loss: 1075761920.0000\n",
      "Epoch 157/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1087529088.0000 - val_loss: 1084693632.0000\n",
      "Epoch 158/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1083227008.0000 - val_loss: 1066537088.0000\n",
      "Epoch 159/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1082967552.0000 - val_loss: 1069015360.0000\n",
      "Epoch 160/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1089890816.0000 - val_loss: 1066931904.0000\n",
      "Epoch 161/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1087832576.0000 - val_loss: 1083646720.0000\n",
      "Epoch 162/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1084686976.0000 - val_loss: 1065785344.0000\n",
      "Epoch 163/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1082123392.0000 - val_loss: 1062091648.0000\n",
      "Epoch 164/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1078966400.0000 - val_loss: 1064066176.0000\n",
      "Epoch 165/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1081340672.0000 - val_loss: 1065642880.0000\n",
      "Epoch 166/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1082235904.0000 - val_loss: 1063503168.0000\n",
      "Epoch 167/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1077253376.0000 - val_loss: 1067267648.0000\n",
      "Epoch 168/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1078160384.0000 - val_loss: 1078468992.0000\n",
      "Epoch 169/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1077568512.0000 - val_loss: 1087030656.0000\n",
      "Epoch 170/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1072187776.0000 - val_loss: 1076984448.0000\n",
      "Epoch 171/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1071571264.0000 - val_loss: 1070443968.0000\n",
      "Epoch 172/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1072580608.0000 - val_loss: 1077940096.0000\n",
      "Epoch 173/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1073681984.0000 - val_loss: 1071510720.0000\n",
      "Epoch 174/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1067571840.0000 - val_loss: 1078503168.0000\n",
      "Epoch 175/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1068540224.0000 - val_loss: 1061027968.0000\n",
      "Epoch 176/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1071135104.0000 - val_loss: 1057874432.0000\n",
      "Epoch 177/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1065679872.0000 - val_loss: 1098947456.0000\n",
      "Epoch 178/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1062526784.0000 - val_loss: 1070597824.0000\n",
      "Epoch 179/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1064025344.0000 - val_loss: 1058088704.0000\n",
      "Epoch 180/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1064133440.0000 - val_loss: 1061544576.0000\n",
      "Epoch 181/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1064282304.0000 - val_loss: 1062906112.0000\n",
      "Epoch 182/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1064053504.0000 - val_loss: 1062629824.0000\n",
      "Epoch 183/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1061690624.0000 - val_loss: 1078242560.0000\n",
      "Epoch 184/400\n",
      "285/285 [==============================] - ETA: 0s - loss: 1051809408.0000 ETA: 0s - loss: 107311449 - 1s 2ms/step - loss: 1063756416.0000 - val_loss: 1064092544.0000\n",
      "Epoch 185/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1061530176.0000 - val_loss: 1060908224.0000\n",
      "Epoch 186/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1061257472.0000 - val_loss: 1079195136.0000\n",
      "Epoch 187/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1056708160.0000 - val_loss: 1086702464.0000\n",
      "Epoch 188/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1064307840.0000 - val_loss: 1058546880.0000\n",
      "Epoch 189/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1054569728.0000 - val_loss: 1073017728.0000\n",
      "Epoch 190/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1060309632.0000 - val_loss: 1054577216.0000\n",
      "Epoch 191/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1055673152.0000 - val_loss: 1055461184.0000\n",
      "Epoch 192/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1052301056.0000 - val_loss: 1074446976.0000\n",
      "Epoch 193/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1059350592.0000 - val_loss: 1068937984.0000\n",
      "Epoch 194/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1046966528.0000 - val_loss: 1059326528.0000\n",
      "Epoch 195/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1049760256.0000 - val_loss: 1045018176.0000\n",
      "Epoch 196/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1049478656.0000 - val_loss: 1063126272.0000\n",
      "Epoch 197/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1053424320.0000 - val_loss: 1052220800.0000\n",
      "Epoch 198/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1048656576.0000 - val_loss: 1054130304.0000\n",
      "Epoch 199/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1048244352.0000 - val_loss: 1053737984.0000\n",
      "Epoch 200/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1040376448.0000 - val_loss: 1047457088.0000\n",
      "Epoch 201/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1052793216.0000 - val_loss: 1072991424.0000\n",
      "Epoch 202/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1043117312.0000 - val_loss: 1058075776.0000\n",
      "Epoch 203/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1043287232.0000 - val_loss: 1048922944.0000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 204/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1043126784.0000 - val_loss: 1043868864.0000\n",
      "Epoch 205/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1039690112.0000 - val_loss: 1046757696.0000\n",
      "Epoch 206/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1040483264.0000 - val_loss: 1037953280.0000\n",
      "Epoch 207/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1033305344.0000 - val_loss: 1055178816.0000\n",
      "Epoch 208/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1039718976.0000 - val_loss: 1042709760.0000\n",
      "Epoch 209/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1038628096.0000 - val_loss: 1036745920.0000\n",
      "Epoch 210/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1035179520.0000 - val_loss: 1068266688.0000\n",
      "Epoch 211/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1035643968.0000 - val_loss: 1061628736.0000\n",
      "Epoch 212/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1035017728.0000 - val_loss: 1056289600.0000\n",
      "Epoch 213/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1029792448.0000 - val_loss: 1038478336.0000\n",
      "Epoch 214/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1031129536.0000 - val_loss: 1069652736.0000\n",
      "Epoch 215/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1037589760.0000 - val_loss: 1045816640.0000\n",
      "Epoch 216/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1027412992.0000 - val_loss: 1038935360.0000\n",
      "Epoch 217/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1027000704.0000 - val_loss: 1083228160.0000\n",
      "Epoch 218/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1033894464.0000 - val_loss: 1032771968.0000\n",
      "Epoch 219/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1037527296.0000 - val_loss: 1039928128.0000\n",
      "Epoch 220/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1027858176.0000 - val_loss: 1031574336.0000\n",
      "Epoch 221/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1022710720.0000 - val_loss: 1043092416.0000\n",
      "Epoch 222/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1026175296.0000 - val_loss: 1068965440.0000\n",
      "Epoch 223/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1025873984.0000 - val_loss: 1039424192.0000\n",
      "Epoch 224/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1032697408.0000 - val_loss: 1036497024.0000\n",
      "Epoch 225/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1020730240.0000 - val_loss: 1029696448.0000\n",
      "Epoch 226/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1023778432.0000 - val_loss: 1027408512.0000\n",
      "Epoch 227/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1021396160.0000 - val_loss: 1044086528.0000\n",
      "Epoch 228/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1017612736.0000 - val_loss: 1027544512.0000\n",
      "Epoch 229/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1014635776.0000 - val_loss: 1033082368.0000\n",
      "Epoch 230/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1023189184.0000 - val_loss: 1083965056.0000\n",
      "Epoch 231/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1020449536.0000 - val_loss: 1048412416.0000\n",
      "Epoch 232/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1017061056.0000 - val_loss: 1038724288.0000\n",
      "Epoch 233/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1010294400.0000 - val_loss: 1144660608.0000\n",
      "Epoch 234/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1017673728.0000 - val_loss: 1056226560.0000\n",
      "Epoch 235/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1012816960.0000 - val_loss: 1054611776.0000\n",
      "Epoch 236/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1017648256.0000 - val_loss: 1034829376.0000\n",
      "Epoch 237/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1014096192.0000 - val_loss: 1034141888.0000\n",
      "Epoch 238/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1014651456.0000 - val_loss: 1028816960.0000 loss: 1028944192.00\n",
      "Epoch 239/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1003823360.0000 - val_loss: 1031832576.0000\n",
      "Epoch 240/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1026017728.0000 - val_loss: 1080331904.0000\n",
      "Epoch 241/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1019927040.0000 - val_loss: 1026336000.0000\n",
      "Epoch 242/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1014675200.0000 - val_loss: 1027062976.0000\n",
      "Epoch 243/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1004693696.0000 - val_loss: 1029680896.0000\n",
      "Epoch 244/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1009548864.0000 - val_loss: 1055480448.0000\n",
      "Epoch 245/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1012331136.0000 - val_loss: 1021111680.0000\n",
      "Epoch 246/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1009100608.0000 - val_loss: 1034685056.0000\n",
      "Epoch 247/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1004308992.0000 - val_loss: 1072973696.0000\n",
      "Epoch 248/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1004888512.0000 - val_loss: 1023979264.0000\n",
      "Epoch 249/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1003434880.0000 - val_loss: 1020015488.0000\n",
      "Epoch 250/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1010915456.0000 - val_loss: 1082281600.0000\n",
      "Epoch 251/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1001832960.0000 - val_loss: 1041273984.0000\n",
      "Epoch 252/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1004224000.0000 - val_loss: 1043787520.0000\n",
      "Epoch 253/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1000443392.0000 - val_loss: 1023620992.0000\n",
      "Epoch 254/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 1000103936.0000 - val_loss: 1045078912.0000\n",
      "Epoch 255/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 997354240.0000 - val_loss: 1029132352.0000\n",
      "Epoch 256/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 996748864.0000 - val_loss: 1059152960.0000\n",
      "Epoch 257/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 995078656.0000 - val_loss: 1034086848.0000\n",
      "Epoch 258/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 1006458560.0000 - val_loss: 1018704768.0000\n",
      "Epoch 259/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 991521344.0000 - val_loss: 1023710656.0000\n",
      "Epoch 260/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 993549312.0000 - val_loss: 1024708864.0000\n",
      "Epoch 261/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 994096256.0000 - val_loss: 1022873984.0000\n",
      "Epoch 262/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 997359680.0000 - val_loss: 1011800128.0000\n",
      "Epoch 263/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 997021120.0000 - val_loss: 1021667264.0000\n",
      "Epoch 264/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 991244800.0000 - val_loss: 1047964032.0000\n",
      "Epoch 265/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 992274816.0000 - val_loss: 1043536000.0000\n",
      "Epoch 266/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 987974976.0000 - val_loss: 1011667072.0000\n",
      "Epoch 267/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 985577088.0000 - val_loss: 1014745024.0000\n",
      "Epoch 268/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 983699968.0000 - val_loss: 1032859904.0000\n",
      "Epoch 269/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 987802176.0000 - val_loss: 1029132928.0000\n",
      "Epoch 270/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 982745408.0000 - val_loss: 1022482624.0000\n",
      "Epoch 271/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "285/285 [==============================] - 1s 2ms/step - loss: 981408832.0000 - val_loss: 1159162240.0000\n",
      "Epoch 272/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 998010944.0000 - val_loss: 1051348032.0000\n",
      "Epoch 273/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 984260608.0000 - val_loss: 1009750656.0000\n",
      "Epoch 274/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 981980992.0000 - val_loss: 1022232960.0000\n",
      "Epoch 275/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 983204864.0000 - val_loss: 1014326656.0000\n",
      "Epoch 276/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 985407872.0000 - val_loss: 1020999744.0000\n",
      "Epoch 277/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 980996480.0000 - val_loss: 1056073024.0000\n",
      "Epoch 278/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 973640320.0000 - val_loss: 1029635712.0000\n",
      "Epoch 279/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 974363200.0000 - val_loss: 1018743296.0000\n",
      "Epoch 280/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 987452352.0000 - val_loss: 1043103616.0000\n",
      "Epoch 281/400\n",
      "285/285 [==============================] - ETA: 0s - loss: 978500416.000 - 0s 2ms/step - loss: 978306624.0000 - val_loss: 1015626368.0000\n",
      "Epoch 282/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 980175232.0000 - val_loss: 1023220928.0000\n",
      "Epoch 283/400\n",
      "285/285 [==============================] - ETA: 0s - loss: 967411136.000 - 1s 2ms/step - loss: 974913088.0000 - val_loss: 1142728704.0000\n",
      "Epoch 284/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 974319936.0000 - val_loss: 1022576320.0000\n",
      "Epoch 285/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 968458176.0000 - val_loss: 1003913856.0000\n",
      "Epoch 286/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 989439488.0000 - val_loss: 1010902272.0000\n",
      "Epoch 287/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 969273216.0000 - val_loss: 1015990592.0000\n",
      "Epoch 288/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 974558336.0000 - val_loss: 1010916736.0000\n",
      "Epoch 289/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 969191168.0000 - val_loss: 1013447232.0000\n",
      "Epoch 290/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 968720896.0000 - val_loss: 1024379904.0000\n",
      "Epoch 291/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 976633664.0000 - val_loss: 1008665792.0000\n",
      "Epoch 292/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 968653568.0000 - val_loss: 1014733376.0000\n",
      "Epoch 293/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 965622848.0000 - val_loss: 1034853440.0000\n",
      "Epoch 294/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 973669696.0000 - val_loss: 1036522176.0000\n",
      "Epoch 295/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 964910336.0000 - val_loss: 1015896448.0000\n",
      "Epoch 296/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 969403136.0000 - val_loss: 1002411712.0000\n",
      "Epoch 297/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 971666496.0000 - val_loss: 1011574080.0000\n",
      "Epoch 298/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 964252928.0000 - val_loss: 1012692032.0000\n",
      "Epoch 299/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 973523008.0000 - val_loss: 1004946240.0000\n",
      "Epoch 300/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 970174720.0000 - val_loss: 1007575360.0000\n",
      "Epoch 301/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 966122432.0000 - val_loss: 1020065984.0000\n",
      "Epoch 302/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 969997696.0000 - val_loss: 995341376.0000\n",
      "Epoch 303/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 961042240.0000 - val_loss: 1005403776.0000\n",
      "Epoch 304/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 961250816.0000 - val_loss: 997216640.0000\n",
      "Epoch 305/400\n",
      "285/285 [==============================] - ETA: 0s - loss: 951669824.000 - 0s 2ms/step - loss: 956518848.0000 - val_loss: 1004811520.0000\n",
      "Epoch 306/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 958178880.0000 - val_loss: 1013728384.0000\n",
      "Epoch 307/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 954750528.0000 - val_loss: 1032700288.0000\n",
      "Epoch 308/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 955225600.0000 - val_loss: 1011090880.0000\n",
      "Epoch 309/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 958058048.0000 - val_loss: 997442240.0000\n",
      "Epoch 310/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 962359360.0000 - val_loss: 996107328.0000\n",
      "Epoch 311/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 959622208.0000 - val_loss: 1003076224.0000\n",
      "Epoch 312/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 952334336.0000 - val_loss: 994471488.0000\n",
      "Epoch 313/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 949815808.0000 - val_loss: 1001067712.0000\n",
      "Epoch 314/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 951431936.0000 - val_loss: 1094993408.0000\n",
      "Epoch 315/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 953965184.0000 - val_loss: 1001579328.0000\n",
      "Epoch 316/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 943905920.0000 - val_loss: 995951680.0000\n",
      "Epoch 317/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 953108800.0000 - val_loss: 1017344512.0000\n",
      "Epoch 318/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 949910208.0000 - val_loss: 1003607680.0000\n",
      "Epoch 319/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 947487168.0000 - val_loss: 1028848000.0000\n",
      "Epoch 320/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 951980864.0000 - val_loss: 1013714112.0000\n",
      "Epoch 321/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 955741952.0000 - val_loss: 1014690496.0000\n",
      "Epoch 322/400\n",
      "285/285 [==============================] - ETA: 0s - loss: 947021696.000 - 1s 2ms/step - loss: 949049536.0000 - val_loss: 1031263744.0000\n",
      "Epoch 323/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 942982720.0000 - val_loss: 1006925376.0000\n",
      "Epoch 324/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 950041984.0000 - val_loss: 1016586176.0000\n",
      "Epoch 325/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 947982016.0000 - val_loss: 996480000.0000\n",
      "Epoch 326/400\n",
      "285/285 [==============================] - 1s 2ms/step - loss: 941613952.0000 - val_loss: 1047131392.0000\n",
      "Epoch 327/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 941720000.0000 - val_loss: 999365184.0000\n",
      "Epoch 328/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 938658304.0000 - val_loss: 995527744.0000\n",
      "Epoch 329/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 946268928.0000 - val_loss: 991614592.0000 - ETA: 0s - loss: 959570112.0\n",
      "Epoch 330/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 942874304.0000 - val_loss: 988852096.0000\n",
      "Epoch 331/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 942397184.0000 - val_loss: 1013321984.0000\n",
      "Epoch 332/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 943363840.0000 - val_loss: 984647808.0000\n",
      "Epoch 333/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 946920000.0000 - val_loss: 1004915584.0000\n",
      "Epoch 334/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 932531008.0000 - val_loss: 1021486016.0000\n",
      "Epoch 335/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 933564288.0000 - val_loss: 992557376.0000\n",
      "Epoch 336/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 935196288.0000 - val_loss: 995579648.0000\n",
      "Epoch 337/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 931365184.0000 - val_loss: 997328000.0000\n",
      "Epoch 338/400\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "285/285 [==============================] - 0s 2ms/step - loss: 932890048.0000 - val_loss: 993576448.0000\n",
      "Epoch 339/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 928679104.0000 - val_loss: 1006130944.0000\n",
      "Epoch 340/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 945048128.0000 - val_loss: 992323968.0000 ETA: 0s - loss: 942458368.000\n",
      "Epoch 341/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 932806528.0000 - val_loss: 1004677248.0000\n",
      "Epoch 342/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 936312512.0000 - val_loss: 992839744.0000\n",
      "Epoch 343/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 929281344.0000 - val_loss: 989220288.0000\n",
      "Epoch 344/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 951090112.0000 - val_loss: 998221312.0000\n",
      "Epoch 345/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 931585664.0000 - val_loss: 1011673792.0000\n",
      "Epoch 346/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 933030912.0000 - val_loss: 998900928.0000\n",
      "Epoch 347/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 939861760.0000 - val_loss: 990118912.0000\n",
      "Epoch 348/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 932686528.0000 - val_loss: 1051336192.0000\n",
      "Epoch 349/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 938154816.0000 - val_loss: 1002549888.0000\n",
      "Epoch 350/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 933489856.0000 - val_loss: 1011036480.0000\n",
      "Epoch 351/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 928963904.0000 - val_loss: 1023456512.0000\n",
      "Epoch 352/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 926993920.0000 - val_loss: 993643392.0000\n",
      "Epoch 353/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 920602624.0000 - val_loss: 1033705856.0000\n",
      "Epoch 354/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 931417856.0000 - val_loss: 1013165760.0000\n",
      "Epoch 355/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 921947712.0000 - val_loss: 1002544256.0000\n",
      "Epoch 356/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 920432384.0000 - val_loss: 987805184.0000\n",
      "Epoch 357/400\n",
      "285/285 [==============================] - 0s 2ms/step - loss: 929471360.0000 - val_loss: 986192704.0000\n",
      "Epoch 00357: early stopping\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x27888abc730>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(x=X_train,\n",
    "        y=y_train,\n",
    "        validation_data=(X_test,y_test),\n",
    "        batch_size=batchsize,\n",
    "        epochs=epochs,\n",
    "        verbose=1,\n",
    "        callbacks=[early_stop])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "losses = pd.DataFrame(model.history.history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "losses.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluation Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def eval_metrics_timeseries(real_values,pred,name='Variable'):\n",
    "    print(f\"\"\"\n",
    "        ------- {name} ------\n",
    "        \n",
    "MAE: {mean_absolute_error(real_values,pred):.3f},\n",
    "MSE: {mean_squared_error(real_values,pred):.3f},\n",
    "RMSE: {np.sqrt(mean_squared_error(real_values,pred)):.3f},\n",
    "Explained Variance Score: {explained_variance_score(real_values,pred)*100:.3f}%,\n",
    "\n",
    "Compare Mean values per Year: Predictions VS Real:\n",
    "\n",
    "{pd.concat([pd.DataFrame(real_values).mean(),pd.DataFrame(pred).mean()],axis=1,keys=['Avg real',' Avg Pred'],names='Year')}\n",
    "    \"\"\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_Thermal = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "        ------- Extracted thermal Power (KJ/s) ------\n",
      "        \n",
      "MAE: 14447.514,\n",
      "MSE: 986192628.297,\n",
      "RMSE: 31403.704,\n",
      "Explained Variance Score: 50.256%,\n",
      "\n",
      "Compare Mean values per Year: Predictions VS Real:\n",
      "\n",
      "Year      Avg real      Avg Pred\n",
      "0     15869.373551  15595.935547\n",
      "1     13124.420794  13199.358398\n",
      "2     12011.464328  12245.181641\n",
      "3     11256.124639  11600.135742\n",
      "4     10681.397884  11105.348633\n",
      "5     10217.756699  10712.119141\n",
      "6      9828.400484  10366.557617\n",
      "7      9494.600307  10087.469727\n",
      "8      9200.976046   9833.971680\n",
      "9      8940.896700   9596.735352\n",
      "10     8705.766134   9391.377930\n",
      "11     8492.842743   9218.810547\n",
      "12     8297.599446   9035.027344\n",
      "13     8117.681475   8878.205078\n",
      "14     7950.126238   8738.871094\n",
      "15     7795.000547   8608.989258\n",
      "16     7650.036376   8481.620117\n",
      "17     7512.571181   8339.352539\n",
      "18     7383.389226   8236.674805\n",
      "19     7261.301960   8133.096191\n",
      "    \n"
     ]
    }
   ],
   "source": [
    "eval_metrics_timeseries(y_test,predictions_Thermal,name='Extracted thermal Power (KJ/s)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:title={'center':'Avg Thermal Power: Predicted Vs Real values'}>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.DataFrame(predictions_Thermal,columns=df_Thermal.columns).mean().plot(legend=True,label='Avg Thermal Power PRED',\n",
    "                                                                           title='Avg Thermal Power: Predicted Vs Real values')\n",
    "pd.DataFrame(y_test,columns=df_Thermal.columns).mean().plot(legend=True,label='Avg Thermal Power REAL')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### END"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}