Bosch Connected Experience(BCX) 2020.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bosch Connected Experience (BCX) 2020"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Implementation of [City Tree](https://greencitysolutions.de/en/) concept under the category of Grow Clean Air in BCX20 by: "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$Mehryar{\\,}Majd$       ${\\,}{\\,}{\\,}{\\,}{\\,}{\\,}{\\,}$           __Gmail: mehryar.majd@gmail.com__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__17-18 Feb, 2020, Berlin__ "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![img](https://i.imgur.com/D7rQ05K.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![img](https://i.imgur.com/mZA5Rcu.jpg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!/usr/bin/env python 2. \n",
    "# -*- coding: utf-8 -*- \n",
    "__author__ = \"Mehryar Majd\" \n",
    "__copyright__ = [\"Copyright 2020, Green City Solutions GmbH \"] \n",
    "__credits__ = [\"Mehryar Majd\",\"Bosch Connected Experience(BCX) 2020\"] \n",
    "__license__ = \"Hackathon\" \n",
    "__version__ = \"Python 3.6\" \n",
    "__maintainer__ = \"Mehryar Majd\" \n",
    "__email__ = \"mehryar.majd@gmail.com\" \n",
    "__status__ = \"Prototype\" \n",
    "\n",
    "'''  Provides prediction of humidity, temperature, fan speed & dif pressure developments on eTree based on time series. \n",
    "       Preprocessing data phase: Sorting, clean data, identification and plotting missing data, detect outliers \n",
    "       Missing data analysis, missing data frequency visualization  --->  missingno module. \n",
    "       ****************************************************\n",
    "       ML & DL models implementation phase using: \n",
    "       Splitting data, Creating ANNs/DNNs models: LSTM. \n",
    "       Design, fit, compile models, train & test data, \n",
    "       \n",
    "       **************************************************** \n",
    "       Implementing evaluation: Plot loss plots, Print results: MAE, MSE \n",
    "       \n",
    "''' "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import and call the needed libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import and call the needed libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "import csv\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, LSTM, Dropout, Activation , BatchNormalization, GRU, Flatten\n",
    "from keras.callbacks import EarlyStopping\n",
    "\n",
    "import numpy\n",
    "from numpy import array, hstack\n",
    "from numpy.random import uniform\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.model_selection import train_test_split, KFold, cross_val_score\n",
    "from sklearn.metrics import mean_squared_error, confusion_matrix, auc, roc_curve, mean_absolute_error\n",
    "from sklearn import preprocessing, tree, linear_model\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.neural_network import MLPRegressor\n",
    "from sklearn.pipeline import Pipeline, make_pipeline\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.linear_model import Ridge, LogisticRegression\n",
    "from sklearn.preprocessing import PolynomialFeatures, MinMaxScaler, StandardScaler\n",
    "from sklearn import ensemble\n",
    "from sklearn import datasets\n",
    "\n",
    "import seaborn as sns\n",
    "\n",
    "import sys\n",
    "import time\n",
    "import datetime\n",
    "from math import sqrt\n",
    "import math\n",
    "from itertools import cycle\n",
    "\n",
    "from scipy.special import boxcox1p\n",
    "from scipy import stats\n",
    "from scipy import interp\n",
    "\n",
    "from pandas import read_csv\n",
    "\n",
    "from tabulate import tabulate\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "9dm8JRHcZc9M",
    "outputId": "1080905d-f4fd-4e48-a9e0-444703c5a40b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n"
     ]
    }
   ],
   "source": [
    "fn = 'D:/berlin_sensordata.json.gz'\n",
    "print(os.path.isfile(fn))\n",
    "df = pd.read_json(fn, lines=True, compression='gzip')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 434
    },
    "colab_type": "code",
    "id": "31WRe9Y2Zgg_",
    "outputId": "e49358cc-2699-4961-baa3-194364cf7e1e"
   },
   "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>absoluteHumidity1</th>\n",
       "      <th>absoluteHumidity2</th>\n",
       "      <th>aggregated</th>\n",
       "      <th>air</th>\n",
       "      <th>airQualityIndexPollutants</th>\n",
       "      <th>currentActuatorConfig</th>\n",
       "      <th>differentialPressurePascal</th>\n",
       "      <th>fan1Rpm</th>\n",
       "      <th>fan2Rpm</th>\n",
       "      <th>humidity1</th>\n",
       "      <th>humidity2</th>\n",
       "      <th>id</th>\n",
       "      <th>moduleId</th>\n",
       "      <th>receivedTimestamp</th>\n",
       "      <th>state</th>\n",
       "      <th>temperature1</th>\n",
       "      <th>temperature2</th>\n",
       "      <th>treeId</th>\n",
       "      <th>weather</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.999457</td>\n",
       "      <td>5.963408</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>{'minimumHumidity': 80, 'maximumHumidity': 85,...</td>\n",
       "      <td>0.2</td>\n",
       "      <td>840.0</td>\n",
       "      <td>840.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>1570439258000-ct_berlin_1-d6</td>\n",
       "      <td>d6</td>\n",
       "      <td>1570439258000</td>\n",
       "      <td>{'areFansEnabled': True, 'isWateringEnabled': ...</td>\n",
       "      <td>16.4</td>\n",
       "      <td>16.3</td>\n",
       "      <td>ct_berlin_1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6.018530</td>\n",
       "      <td>6.018530</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>{'minimumHumidity': 80, 'maximumHumidity': 85,...</td>\n",
       "      <td>1.3</td>\n",
       "      <td>840.0</td>\n",
       "      <td>480.0</td>\n",
       "      <td>45.0</td>\n",
       "      <td>45.0</td>\n",
       "      <td>1570439258000-ct_berlin_1-d3</td>\n",
       "      <td>d3</td>\n",
       "      <td>1570439258000</td>\n",
       "      <td>{'areFansEnabled': False, 'isWateringEnabled':...</td>\n",
       "      <td>15.7</td>\n",
       "      <td>15.7</td>\n",
       "      <td>ct_berlin_1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5.895329</td>\n",
       "      <td>6.138979</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>{'minimumHumidity': 80, 'maximumHumidity': 85,...</td>\n",
       "      <td>0.9</td>\n",
       "      <td>840.0</td>\n",
       "      <td>840.0</td>\n",
       "      <td>42.0</td>\n",
       "      <td>44.0</td>\n",
       "      <td>1570439258000-ct_berlin_1-d5</td>\n",
       "      <td>d5</td>\n",
       "      <td>1570439258000</td>\n",
       "      <td>{'areFansEnabled': True, 'isWateringEnabled': ...</td>\n",
       "      <td>16.5</td>\n",
       "      <td>16.4</td>\n",
       "      <td>ct_berlin_1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5.849228</td>\n",
       "      <td>5.708862</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>{'minimumHumidity': 80, 'maximumHumidity': 85,...</td>\n",
       "      <td>1.2</td>\n",
       "      <td>840.0</td>\n",
       "      <td>840.0</td>\n",
       "      <td>44.0</td>\n",
       "      <td>44.0</td>\n",
       "      <td>1570439258000-ct_berlin_1-d4</td>\n",
       "      <td>d4</td>\n",
       "      <td>1570439258000</td>\n",
       "      <td>{'areFansEnabled': True, 'isWateringEnabled': ...</td>\n",
       "      <td>15.6</td>\n",
       "      <td>15.2</td>\n",
       "      <td>ct_berlin_1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10.757431</td>\n",
       "      <td>11.806965</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>{'minimumHumidity': 80, 'maximumHumidity': 85,...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>86.0</td>\n",
       "      <td>91.0</td>\n",
       "      <td>1570545523000-ct_berlin_1-d5</td>\n",
       "      <td>d5</td>\n",
       "      <td>1570545523000</td>\n",
       "      <td>{'areFansEnabled': False, 'isWateringEnabled':...</td>\n",
       "      <td>14.6</td>\n",
       "      <td>15.2</td>\n",
       "      <td>ct_berlin_1</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   absoluteHumidity1  absoluteHumidity2 aggregated  air  \\\n",
       "0           5.999457           5.963408        NaN  NaN   \n",
       "1           6.018530           6.018530        NaN  NaN   \n",
       "2           5.895329           6.138979        NaN  NaN   \n",
       "3           5.849228           5.708862        NaN  NaN   \n",
       "4          10.757431          11.806965        NaN  NaN   \n",
       "\n",
       "  airQualityIndexPollutants  \\\n",
       "0                       NaN   \n",
       "1                       NaN   \n",
       "2                       NaN   \n",
       "3                       NaN   \n",
       "4                       NaN   \n",
       "\n",
       "                               currentActuatorConfig  \\\n",
       "0  {'minimumHumidity': 80, 'maximumHumidity': 85,...   \n",
       "1  {'minimumHumidity': 80, 'maximumHumidity': 85,...   \n",
       "2  {'minimumHumidity': 80, 'maximumHumidity': 85,...   \n",
       "3  {'minimumHumidity': 80, 'maximumHumidity': 85,...   \n",
       "4  {'minimumHumidity': 80, 'maximumHumidity': 85,...   \n",
       "\n",
       "   differentialPressurePascal  fan1Rpm  fan2Rpm  humidity1  humidity2  \\\n",
       "0                         0.2    840.0    840.0       43.0       43.0   \n",
       "1                         1.3    840.0    480.0       45.0       45.0   \n",
       "2                         0.9    840.0    840.0       42.0       44.0   \n",
       "3                         1.2    840.0    840.0       44.0       44.0   \n",
       "4                         0.0      0.0      0.0       86.0       91.0   \n",
       "\n",
       "                             id moduleId  receivedTimestamp  \\\n",
       "0  1570439258000-ct_berlin_1-d6       d6      1570439258000   \n",
       "1  1570439258000-ct_berlin_1-d3       d3      1570439258000   \n",
       "2  1570439258000-ct_berlin_1-d5       d5      1570439258000   \n",
       "3  1570439258000-ct_berlin_1-d4       d4      1570439258000   \n",
       "4  1570545523000-ct_berlin_1-d5       d5      1570545523000   \n",
       "\n",
       "                                               state  temperature1  \\\n",
       "0  {'areFansEnabled': True, 'isWateringEnabled': ...          16.4   \n",
       "1  {'areFansEnabled': False, 'isWateringEnabled':...          15.7   \n",
       "2  {'areFansEnabled': True, 'isWateringEnabled': ...          16.5   \n",
       "3  {'areFansEnabled': True, 'isWateringEnabled': ...          15.6   \n",
       "4  {'areFansEnabled': False, 'isWateringEnabled':...          14.6   \n",
       "\n",
       "   temperature2       treeId weather  \n",
       "0          16.3  ct_berlin_1     NaN  \n",
       "1          15.7  ct_berlin_1     NaN  \n",
       "2          16.4  ct_berlin_1     NaN  \n",
       "3          15.2  ct_berlin_1     NaN  \n",
       "4          15.2  ct_berlin_1     NaN  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "3hjc0KZbfYNR"
   },
   "source": [
    "# Missing value ratio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 343
    },
    "colab_type": "code",
    "id": "Q0ektN_2ewM9",
    "outputId": "17f85979-3d9c-47d6-a3f8-6ab98213fa51"
   },
   "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>Missing Ratio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>airQualityIndexPollutants</th>\n",
       "      <td>86.942824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aggregated</th>\n",
       "      <td>76.857110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>weather</th>\n",
       "      <td>50.717054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>temperature2</th>\n",
       "      <td>27.230105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>temperature1</th>\n",
       "      <td>27.230105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>state</th>\n",
       "      <td>27.230105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>humidity2</th>\n",
       "      <td>27.230105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>humidity1</th>\n",
       "      <td>27.230105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fan2Rpm</th>\n",
       "      <td>27.230105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fan1Rpm</th>\n",
       "      <td>27.230105</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                           Missing Ratio\n",
       "airQualityIndexPollutants      86.942824\n",
       "aggregated                     76.857110\n",
       "weather                        50.717054\n",
       "temperature2                   27.230105\n",
       "temperature1                   27.230105\n",
       "state                          27.230105\n",
       "humidity2                      27.230105\n",
       "humidity1                      27.230105\n",
       "fan2Rpm                        27.230105\n",
       "fan1Rpm                        27.230105"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_data_na = (df.isnull().sum() / len(df)) * 100\n",
    "all_data_na = all_data_na.drop(all_data_na[all_data_na == 0].index).sort_values(ascending=False)[:30]\n",
    "missing_data = pd.DataFrame({'Missing Ratio' :all_data_na})\n",
    "missing_data.head(10) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Dgy1f9-UgNtR"
   },
   "source": [
    "# Delete some fileds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "aSyACOhhfdXY"
   },
   "outputs": [],
   "source": [
    "df=df.drop(\"airQualityIndexPollutants\",axis=1)\n",
    "df=df.drop(\"air\",axis=1)\n",
    "df=df.drop('aggregated'\t,axis=1)\n",
    "df=df.drop('weather'\t,axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "jF61VPVTgUUG"
   },
   "source": [
    "# Fill missing data by mean values in target features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XGhus3GzelbO"
   },
   "outputs": [],
   "source": [
    "df['receivedTimestamp'].fillna((df['receivedTimestamp'].mean()), inplace=True)\n",
    "df['fan2Rpm'].fillna((df['fan2Rpm'].mean()), inplace=True)\n",
    "df['fan1Rpm'].fillna((df['fan1Rpm'].mean()), inplace=True)\n",
    "df['differentialPressurePascal'].fillna((df['differentialPressurePascal'].mean()), inplace=True)\n",
    "df['absoluteHumidity2'].fillna((df['absoluteHumidity2'].mean()), inplace=True)\n",
    "df['absoluteHumidity1'].fillna((df['absoluteHumidity1'].mean()), inplace=True)\n",
    "\n",
    "df['humidity2'].fillna((df['humidity2'].mean()), inplace=True)\n",
    "df['humidity1'].fillna((df['humidity1'].mean()), inplace=True)\n",
    "df['temperature2'].fillna((df['temperature2'].mean()), inplace=True)\n",
    "df['temperature1'].fillna((df['temperature1'].mean()), inplace=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 136
    },
    "colab_type": "code",
    "id": "EtW0jPysg2z6",
    "outputId": "abb36acb-9196-495a-a03e-5fe404535345"
   },
   "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>Missing Ratio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>state</th>\n",
       "      <td>27.230105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>currentActuatorConfig</th>\n",
       "      <td>27.230105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>moduleId</th>\n",
       "      <td>24.574844</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Missing Ratio\n",
       "state                      27.230105\n",
       "currentActuatorConfig      27.230105\n",
       "moduleId                   24.574844"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_data_na = (df.isnull().sum() / len(df)) * 100\n",
    "all_data_na = all_data_na.drop(all_data_na[all_data_na == 0].index).sort_values(ascending=False)[:30]\n",
    "missing_data = pd.DataFrame({'Missing Ratio' :all_data_na})\n",
    "missing_data.head(11) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "xgP3xLbLeF6x"
   },
   "source": [
    "# Preparing the data & data split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 101
    },
    "colab_type": "code",
    "id": "JKNCn6Kskqoq",
    "outputId": "e9e8be0e-6a27-4de0-9c3c-759bd8338207"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['absoluteHumidity1', 'absoluteHumidity2', 'currentActuatorConfig',\n",
       "       'differentialPressurePascal', 'fan1Rpm', 'fan2Rpm', 'humidity1',\n",
       "       'humidity2', 'id', 'moduleId', 'receivedTimestamp', 'state',\n",
       "       'temperature1', 'temperature2', 'treeId'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "BcK2OPwGeIye"
   },
   "outputs": [],
   "source": [
    "col=['receivedTimestamp', 'temperature1',\n",
    "       'temperature2', 'humidity1', 'humidity2', 'absoluteHumidity1',\n",
    "       'absoluteHumidity2', 'differentialPressurePascal', 'fan1Rpm', 'fan2Rpm']\n",
    "\n",
    "col2=['temperature1','temperature2',  'absoluteHumidity1',\n",
    "       'absoluteHumidity2', 'differentialPressurePascal','fan1Rpm', 'fan2Rpm']\n",
    "\n",
    "probs = np.random.rand(len(df))\n",
    "training_mask = probs < 0.8\n",
    "test_mask = (probs >= 0.80)\n",
    "\n",
    "x_train_feat=df[col][training_mask]\n",
    "x_test_feat=df[col][test_mask]\n",
    "y_train_feat=df[col2][training_mask]\n",
    "y_test_feat =df[col2][test_mask]\n",
    "\n",
    "sc = StandardScaler()\n",
    "x_train_feat = sc.fit_transform(x_train_feat)\n",
    "x_test_feat = sc.transform(x_test_feat)\n",
    "y_train_feat = sc.fit_transform(y_train_feat)\n",
    "y_test_feat = sc.transform(y_test_feat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "uboMBxsEmO__"
   },
   "outputs": [],
   "source": [
    "x_train_feat = x_train_feat.reshape((x_train_feat.shape[0], 1, 10))\n",
    "x_test_feat = x_test_feat.reshape((x_test_feat.shape[0], 1, 10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 454
    },
    "colab_type": "code",
    "id": "dk3HQpVgoV2b",
    "outputId": "3ce71d03-a7a7-4378-bf99-af07ddb3c2fa"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_1 (LSTM)                (None, 1, 15)             1560      \n",
      "_________________________________________________________________\n",
      "lstm_2 (LSTM)                (None, 1, 15)             1860      \n",
      "_________________________________________________________________\n",
      "lstm_3 (LSTM)                (None, 1, 15)             1860      \n",
      "_________________________________________________________________\n",
      "lstm_4 (LSTM)                (None, 1, 15)             1860      \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 15)                0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 20)                320       \n",
      "_________________________________________________________________\n",
      "activation_1 (Activation)    (None, 20)                0         \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 20)                0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 7)                 147       \n",
      "=================================================================\n",
      "Total params: 7,607\n",
      "Trainable params: 7,607\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model = Sequential()\n",
    "model.add(LSTM(15,return_sequences=True ,input_shape=(1, 10)))\n",
    "model.add(LSTM(15,return_sequences=True,dropout=0.50 ))\n",
    "model.add(LSTM(15,return_sequences=True,dropout=0.50 ))\n",
    "model.add(LSTM(15,return_sequences=True,dropout=0.50 ))\n",
    "model.add(Flatten())\n",
    "\n",
    "model.add(Dense(units = 20))\n",
    "model.add(Activation('tanh'))\n",
    "model.add(Dropout(0.50))\n",
    "model.add(Dense(7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss=\"mse\", optimizer=\"adam\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "e8QcGX9-oiFj",
    "outputId": "04fd7d51-8ad5-4729-8e2c-cda3061fb61f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 556346 samples, validate on 140045 samples\n",
      "Epoch 1/100\n",
      "556346/556346 [==============================] - 17s 31us/step - loss: 0.5680 - val_loss: 0.1869\n",
      "Epoch 2/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2997 - val_loss: 0.1578\n",
      "Epoch 3/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2735 - val_loss: 0.1443\n",
      "Epoch 4/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2597 - val_loss: 0.1350\n",
      "Epoch 5/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2505 - val_loss: 0.1302\n",
      "Epoch 6/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2442 - val_loss: 0.1257\n",
      "Epoch 7/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2397 - val_loss: 0.1226\n",
      "Epoch 8/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2353 - val_loss: 0.1201\n",
      "Epoch 9/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2320 - val_loss: 0.1182\n",
      "Epoch 10/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2300 - val_loss: 0.1174\n",
      "Epoch 11/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2292 - val_loss: 0.1153\n",
      "Epoch 12/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2271 - val_loss: 0.1155\n",
      "Epoch 13/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2259 - val_loss: 0.1116\n",
      "Epoch 14/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2233 - val_loss: 0.1042\n",
      "Epoch 15/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2202 - val_loss: 0.1041\n",
      "Epoch 16/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2188 - val_loss: 0.1036\n",
      "Epoch 17/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2173 - val_loss: 0.1029\n",
      "Epoch 18/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2156 - val_loss: 0.0998\n",
      "Epoch 19/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2147 - val_loss: 0.0999\n",
      "Epoch 20/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2138 - val_loss: 0.0994\n",
      "Epoch 21/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2130 - val_loss: 0.0983\n",
      "Epoch 22/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2118 - val_loss: 0.1000\n",
      "Epoch 23/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2115 - val_loss: 0.0958\n",
      "Epoch 24/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2110 - val_loss: 0.0977\n",
      "Epoch 25/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2106 - val_loss: 0.0964\n",
      "Epoch 26/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2098 - val_loss: 0.0969\n",
      "Epoch 27/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2110 - val_loss: 0.0968\n",
      "Epoch 28/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2099 - val_loss: 0.0954\n",
      "Epoch 29/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2094 - val_loss: 0.0956\n",
      "Epoch 30/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2083 - val_loss: 0.0982\n",
      "Epoch 31/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2080 - val_loss: 0.0956\n",
      "Epoch 32/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2080 - val_loss: 0.0969\n",
      "Epoch 33/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2081 - val_loss: 0.1009\n",
      "Epoch 34/100\n",
      "556346/556346 [==============================] - 12s 21us/step - loss: 0.2090 - val_loss: 0.0986\n",
      "Epoch 35/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2081 - val_loss: 0.0963\n",
      "Epoch 36/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2069 - val_loss: 0.0949\n",
      "Epoch 37/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2064 - val_loss: 0.0955\n",
      "Epoch 38/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2060 - val_loss: 0.0968\n",
      "Epoch 39/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2067 - val_loss: 0.1008\n",
      "Epoch 40/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2068 - val_loss: 0.0952\n",
      "Epoch 41/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2056 - val_loss: 0.0957\n",
      "Epoch 42/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2054 - val_loss: 0.0938\n",
      "Epoch 43/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2053 - val_loss: 0.0940\n",
      "Epoch 44/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2053 - val_loss: 0.0962\n",
      "Epoch 45/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2047 - val_loss: 0.0934\n",
      "Epoch 46/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2053 - val_loss: 0.0956\n",
      "Epoch 47/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2050 - val_loss: 0.0926\n",
      "Epoch 48/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2042 - val_loss: 0.0968\n",
      "Epoch 49/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2055 - val_loss: 0.0945\n",
      "Epoch 50/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2047 - val_loss: 0.0941\n",
      "Epoch 51/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2043 - val_loss: 0.0949\n",
      "Epoch 52/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2042 - val_loss: 0.0943\n",
      "Epoch 53/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2037 - val_loss: 0.0940\n",
      "Epoch 54/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2036 - val_loss: 0.0934\n",
      "Epoch 55/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2036 - val_loss: 0.0929\n",
      "Epoch 56/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2030 - val_loss: 0.0927\n",
      "Epoch 57/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2030 - val_loss: 0.0937\n",
      "Epoch 58/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2031 - val_loss: 0.0940\n",
      "Epoch 59/100\n",
      "556346/556346 [==============================] - 11s 21us/step - loss: 0.2033 - val_loss: 0.0919\n",
      "Epoch 60/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2033 - val_loss: 0.0922 ETA: 0s - loss: \n",
      "Epoch 61/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2025 - val_loss: 0.0920\n",
      "Epoch 62/100\n",
      "556346/556346 [==============================] - 11s 21us/step - loss: 0.2032 - val_loss: 0.0924\n",
      "Epoch 63/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2028 - val_loss: 0.0916\n",
      "Epoch 64/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2025 - val_loss: 0.0916\n",
      "Epoch 65/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2024 - val_loss: 0.0952\n",
      "Epoch 66/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2028 - val_loss: 0.0919\n",
      "Epoch 67/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2028 - val_loss: 0.0928\n",
      "Epoch 68/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2017 - val_loss: 0.0922\n",
      "Epoch 69/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2018 - val_loss: 0.0908\n",
      "Epoch 70/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2012 - val_loss: 0.0948\n",
      "Epoch 71/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2014 - val_loss: 0.0942\n",
      "Epoch 72/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2015 - val_loss: 0.0912\n",
      "Epoch 73/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2014 - val_loss: 0.0905\n",
      "Epoch 74/100\n",
      "556346/556346 [==============================] - 12s 21us/step - loss: 0.2013 - val_loss: 0.0918\n",
      "Epoch 75/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2018 - val_loss: 0.0938\n",
      "Epoch 76/100\n",
      "556346/556346 [==============================] - 11s 19us/step - loss: 0.2016 - val_loss: 0.0914\n",
      "Epoch 77/100\n",
      "556346/556346 [==============================] - 11s 19us/step - loss: 0.2007 - val_loss: 0.0917\n",
      "Epoch 78/100\n",
      "556346/556346 [==============================] - 11s 19us/step - loss: 0.2007 - val_loss: 0.0898\n",
      "Epoch 79/100\n",
      "556346/556346 [==============================] - 11s 19us/step - loss: 0.2010 - val_loss: 0.0939\n",
      "Epoch 80/100\n",
      "556346/556346 [==============================] - 11s 19us/step - loss: 0.2009 - val_loss: 0.0911\n",
      "Epoch 81/100\n",
      "556346/556346 [==============================] - 11s 19us/step - loss: 0.2003 - val_loss: 0.0900\n",
      "Epoch 82/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2002 - val_loss: 0.0908\n",
      "Epoch 83/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2000 - val_loss: 0.0912\n",
      "Epoch 84/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2002 - val_loss: 0.0902\n",
      "Epoch 85/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2000 - val_loss: 0.0891\n",
      "Epoch 86/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2000 - val_loss: 0.0904\n",
      "Epoch 87/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.1999 - val_loss: 0.0888\n",
      "Epoch 88/100\n",
      "556346/556346 [==============================] - 12s 21us/step - loss: 0.2007 - val_loss: 0.0907\n",
      "Epoch 89/100\n",
      "556346/556346 [==============================] - 12s 21us/step - loss: 0.2000 - val_loss: 0.0901\n",
      "Epoch 90/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2000 - val_loss: 0.0898\n",
      "Epoch 91/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.1996 - val_loss: 0.0898\n",
      "Epoch 92/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.1997 - val_loss: 0.0902\n",
      "Epoch 93/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2004 - val_loss: 0.0885\n",
      "Epoch 94/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.2004 - val_loss: 0.0914\n",
      "Epoch 95/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.1991 - val_loss: 0.0920\n",
      "Epoch 96/100\n",
      "556346/556346 [==============================] - 12s 21us/step - loss: 0.1988 - val_loss: 0.0904\n",
      "Epoch 97/100\n",
      "556346/556346 [==============================] - 11s 19us/step - loss: 0.1991 - val_loss: 0.0879\n",
      "Epoch 98/100\n",
      "556346/556346 [==============================] - 11s 20us/step - loss: 0.1990 - val_loss: 0.0911\n",
      "Epoch 99/100\n",
      "556346/556346 [==============================] - 11s 19us/step - loss: 0.1991 - val_loss: 0.0909\n",
      "Epoch 100/100\n",
      "556346/556346 [==============================] - 11s 19us/step - loss: 0.1992 - val_loss: 0.0887\n"
     ]
    }
   ],
   "source": [
    "hist_text=model.fit([x_train_feat], \n",
    "                    y_train_feat,\n",
    "                    epochs=100,\n",
    "                    batch_size=16,\n",
    "                    validation_data=([x_test_feat],y_test_feat))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 134
    },
    "colab_type": "code",
    "id": "j9RpRAQuopOW",
    "outputId": "0b48e7ae-29dc-4428-de60-5b37fc70f1dd"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "y1 MSE:0.1225\n",
      "y2 MSE:0.1135\n",
      "y3 MSE:0.1235\n",
      "y4 MSE:0.1027\n",
      "y5 MSE:0.0776\n",
      "y6 MSE:0.0382\n",
      "y7 MSE:0.0428\n"
     ]
    }
   ],
   "source": [
    "#x_test_feat ],y_test_feat\n",
    "ypred = model.predict(x_test_feat)\n",
    "print(\"y1 MSE:%.4f\" % mean_squared_error(y_test_feat[:,0], ypred[:,0]))\n",
    "print(\"y2 MSE:%.4f\" % mean_squared_error(y_test_feat[:,1], ypred[:,1]))\n",
    "print(\"y3 MSE:%.4f\" % mean_squared_error(y_test_feat[:,2], ypred[:,2]))\n",
    "print(\"y4 MSE:%.4f\" % mean_squared_error(y_test_feat[:,3], ypred[:,3]))\n",
    "print(\"y5 MSE:%.4f\" % mean_squared_error(y_test_feat[:,4], ypred[:,4]))\n",
    "print(\"y6 MSE:%.4f\" % mean_squared_error(y_test_feat[:,5], ypred[:,5]))\n",
    "print(\"y7 MSE:%.4f\" % mean_squared_error(y_test_feat[:,6], ypred[:,6]))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "def mean_absolute_percentage_error(y_true, y_pred):     \n",
    "    return np.mean(np.abs(y_true - y_pred) / y_true)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Error Measurement        y0      y1      y2      y3      y4      y5      y6\n",
      "-------------------  ------  ------  ------  ------  ------  ------  ------\n",
      "MSE                  0.1225  0.1135  0.1235  0.1027  0.0776  0.0382  0.0428\n",
      "MAE                  0.2057  0.2003  0.2149  0.198   0.1578  0.1312  0.1388\n"
     ]
    }
   ],
   "source": [
    "#x_test_feat ],y_test_feat\n",
    "ypred = model.predict(x_test_feat)\n",
    "headers = []\n",
    "headers.append(\"Error Measurement\")\n",
    "for i in range(7):\n",
    "    headers.append(\"y\" + str(i))\n",
    "\n",
    "table = []\n",
    "\n",
    "table.append([])\n",
    "table[-1].append(\"MSE\")\n",
    "table[-1].append(\"%.4f\" % mean_squared_error(y_test_feat[:,0], ypred[:,0]))\n",
    "table[-1].append(\"%.4f\" % mean_squared_error(y_test_feat[:,1], ypred[:,1]))\n",
    "table[-1].append(\"%.4f\" % mean_squared_error(y_test_feat[:,2], ypred[:,2]))\n",
    "table[-1].append(\"%.4f\" % mean_squared_error(y_test_feat[:,3], ypred[:,3]))\n",
    "table[-1].append(\"%.4f\" % mean_squared_error(y_test_feat[:,4], ypred[:,4]))\n",
    "table[-1].append(\"%.4f\" % mean_squared_error(y_test_feat[:,5], ypred[:,5]))\n",
    "table[-1].append(\"%.4f\" % mean_squared_error(y_test_feat[:,6], ypred[:,6]))\n",
    "\n",
    "table.append([])\n",
    "table[-1].append(\"MAE\")\n",
    "table[-1].append(\"%.4f\" % mean_absolute_error(y_test_feat[:,0], ypred[:,0]))\n",
    "table[-1].append(\"%.4f\" % mean_absolute_error(y_test_feat[:,1], ypred[:,1]))\n",
    "table[-1].append(\"%.4f\" % mean_absolute_error(y_test_feat[:,2], ypred[:,2]))\n",
    "table[-1].append(\"%.4f\" % mean_absolute_error(y_test_feat[:,3], ypred[:,3]))\n",
    "table[-1].append(\"%.4f\" % mean_absolute_error(y_test_feat[:,4], ypred[:,4]))\n",
    "table[-1].append(\"%.4f\" % mean_absolute_error(y_test_feat[:,5], ypred[:,5]))\n",
    "table[-1].append(\"%.4f\" % mean_absolute_error(y_test_feat[:,6], ypred[:,6]))\n",
    "\n",
    "# table.append([])\n",
    "# table[-1].append(\"MAPE\")\n",
    "# table[-1].append(\"%.4f\" % mean_absolute_percentage_error(y_test_feat[:,0], ypred[:,0]))\n",
    "# table[-1].append(\"%.4f\" % mean_absolute_percentage_error(y_test_feat[:,1], ypred[:,1]))\n",
    "# table[-1].append(\"%.4f\" % mean_absolute_percentage_error(y_test_feat[:,2], ypred[:,2]))\n",
    "# table[-1].append(\"%.4f\" % mean_absolute_percentage_error(y_test_feat[:,3], ypred[:,3]))\n",
    "# table[-1].append(\"%.4f\" % mean_absolute_percentage_error(y_test_feat[:,4], ypred[:,4]))\n",
    "# table[-1].append(\"%.4f\" % mean_absolute_percentage_error(y_test_feat[:,5], ypred[:,5]))\n",
    "# table[-1].append(\"%.4f\" % mean_absolute_percentage_error(y_test_feat[:,6], ypred[:,6]))\n",
    "\n",
    "print(tabulate(table, headers=headers))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction of temperature1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0xecdefb31d0>]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAD8CAYAAABjAo9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzt3XeYE9X6B/Dvm2Q3oS0sTZa6IKKACMKqFBULKgi2q94LWLBwQcR6RQXbD+zKtSMKXmwIYhcFlSZcrtJclCosLLAICLL0suxmk5zfH5m0TZmZZJIp+36eh4dkcjLz7mzy7pkzp5AQAowxxqzDpncAjDHGtMWJnTHGLIYTO2OMWQwndsYYsxhO7IwxZjGc2BljzGI4sTPGmMVwYmeMMYvhxM4YYxbj0OOgDRs2FPn5+XocmjHGTGvlypX7hBCN5Mrpktjz8/NRWFiox6EZY8y0iGi7knLcFMMYYxbDiZ0xxiyGEztjjFkMJ3bGGLMYTuyMMWYxnNgZY8xiOLEzxpjFmCqxHy2vxMxVu/QOgzHGDE2XAUrJGv3FWsxeuxunNK6DDk1z9A6HMcYMyVQ19j8PnwAAnKj06hwJY4wZl6kSO2OMMXmc2BljzGJMldiF0DsCxhgzPlMldsYYY/JMldiJ9I6AMcaMT7PETkR2IvqNiGZptU/GGGPqaVljvxfABg33xxhjLAmaJHYiag6gP4D/aLE/xhhjydOqxv4qgIcA+DTaH2OMsSSlnNiJaACAvUKIlTLlhhFRIREVlpaWpnpYxhhjcWhRY+8F4EoiKgEwA8BFRPRR1UJCiMlCiAIhREGjRrKLbDPGGEtSyoldCDFGCNFcCJEPYCCAH4UQN6YcWcxjpWOvjDFmLabqx84YY0yeptP2CiEWAVik5T7D8QAlxhiTxzV2xhizGE7sjDFmMaZK7HzzlDHG5JkqsTPGGJNnqsTON08ZY0yeqRI7Y4wxeZzYGWPMYkyV2PnmKWOMyTNVYmeMMSbPVImdb54yxpg8UyV2xhhj8kyV2LmNnTHG5JkqsTPGGJNnqsTObeyMMSbPVImdMcaYPE7sjDFmMaZK7HzzlDHG5JkqsTPGGJNnqsTON08ZY0Y15ss16P/6//QOA4DGa54yxlh19fGKHXqHEGSqGju3sTPGmDxTJfYAbpJhRuDx+vDT5n16h8FYFFMmdq65MyMYP6cIN05Zjvm//6V3KIxFMFVi55o6M5JJi7cCAIZ+WKhzJIxFMlViZ4wxJs9UiZ2bYBhjRvfa/M14c2GxrjGYKrEHcJMM08qc9XvwczHfAGXaeWX+JoyfU6RrDKZM7FxzZ1oZPnUlbvjPcizeVKp3KIxpxlSJnWvqLF1ufneF3iEwphlTJXbGGGPyTJXYuQmGGYXb49M7BMbiMlVil3O8woP80bORP3o2BP8VYGl0wu3VOwTG4jJVYpdrYy/cfjD4uIJrVCyNjrk9eodgCUIIjJ+zEbsOndA7lJQYrSJpqsQu56tfdwYfV1RyYmfp89r8TXqHYAnr/zyCNxduwV3Tf9U7FEsxVWKX+6NYHpbMK7x8qczU+aXkgOKynxbulC/EFDNqRWzNzkP4bu1u2XI+Y1XYzZXYV+04BACYvyH2pEs/rN8TfDzDQHMjM3M4Vs7NK3oxWF4E4G9euXLCz7hzmvzVxPCpxpovyFSJPeCjpdtly0z4Ud8hvcx8Fm/mQUqZtvdoOQBgw+4jOkcS7eqJS4KPD5dVJiw7f8PeqG1Ltug3otmUif1ohXzNyu015qUdM67/FnFiz7SpCippelkttRAAoT9Aasxdr990zqZM7Iylw/7jbr1DqHYWWviPqZ49ZVJO7ETUgogWEtEGIlpPRPdqERhjmXb4ROLLbVZ9rNt1OOJ5pVd9ktbzvoEWNXYPgAeEEO0BdAcwkog6aLBfxjLus8LEN93LK71Yu/NwwjLM/Aa88VPE86nL1DcZ6dm1PeXELoTYLYT4VXp8FMAGAM1S3S9jenjw8zUJX3/o8zW4YsJPCcsw6zlSrv5qTuhYZ9e0jZ2I8gGcCWC5lvuNpXjv0XQfglVTngQ33n/bcTDua8y6FidxL+CjZX+kIRJlNEvsRFQbwBcA7hNCRPVdIqJhRFRIRIWlpanfMNn017GU98FYLIG1TGPZccDcQ99ZcpT0xDMSTRI7EWXBn9SnCSG+jFVGCDFZCFEghCho1KhRyse8c9qvyB89O+X9MFaV3qvfMGPbe6QcIz5aieMGTvZa9IohAFMAbBBCvJx6SOpsLT2GHQfK8PlKHuLNGEu/f88twvfr9mDWmj/1DiUuhwb76AXgJgBriWiVtO0RIcR3Guxb1kUv/TcTh2GMyfD5BJZs2Y9ebRuATLzcmVz/833H3FK5TESTHC16xfwkhCAhxBlCiC7Sv4wkdTnb9x/XZD9ujw//21yKT37R72YIY0b30fLtuHHKcsxWMGmWkR2TaWL5cWP09AFGo0WN3bDm/f4Xhp7XJuX9DJ9aGBwhl+PKQr9OeSnvkxlbpdeHLDsPzFZj+/4yAMCew+qH3xtJoEYux8AVdrNNKSDghPJh34Ul2nRNCx/2/KGB57Zg2jF7ctJDyT7/FfLTszfoHElqlm/dr3cIKTNVYh9h/xZFrltQ4hoMQnRf4znZD6HENTj4b+Me7WeMW6rglz5z1S6M+XKt5sfW2+ETlbj+7SXYcaBM71DSbv2fxptt0OgWmKCJQomx365XVM7SbeyZ9HDWjODjba4bo14/1RbZM2by0ZGaHPcSWyGmZI1XXP7eGavw8Qr17fF/n7QUncbOUf2+TJm15k/8UnIQz/+wUe9QmAX5DLJaRbnCRT/eXFhsuCXxAkyV2Ku6wLYq4evtbLs0Oc472S/jYvtvaIRDsmXn/Z78VJ0rth3AUQMv9vCONHBn9prdhv1AM/MyW41/16ETWLndmCORTZ3Y389+Meb2pd70zEH2i+tO2TL//DC0koqVmiwqPF6U7A/9PI98Zb2mpkj8h0upw2WVWP9n6hOjeX3mW0PBY5CrjKpMndjDhbe5P+MZrNl+yyuVr506e01kN6+hHyhfLutg2FzgxXuNN13CqY/9EPF81hpzd2mTs3yb8vVPq7vOT85F/9dTnxjtiIGvVs3GMom9IUI1hu2iSfCx25NaLWBNjCla401lsGCjvxnmZNqFZihF0V/KJyo786l5wcc/FytbUit/9Gzkj56N29//RfFxtGLkJiMtbOa5iDLOyEP0E9mk4nueKZZJ7PawGvvFXdoGH29OcRbIxZuUT1hWtMd/rAXOB/GzS/l6IwerrNzzf9/I35U/4Q5dSZitbdIMKjzKr9SYNszaxXSZAbtHWiaxX2gP3Ui9pVfr4OM5v/ye0n69cW4Szlm/J2rbobJK3G4PH3SrrP1tYZH6xLxqh/yN3HhemlsUrO2z2Lbts879EbMw63xPRpw8wfSJPdC2foltZXBbh7yc4OMau5ZEvUcNz+HYvVyGT10Ztc19aDcez/oo+HxC1uuKjiE3hDmWqctKIp6rWdbtjR+LVR8vFYUlB/DiDxtN1pPGTLEyFsn0iX2gfSEA4KKwGnu2I/RjrduZfM0WACrXfK647Pm2yNV3BtiXY2upfFvtEzOVDYgI993ayCuG25JsZ9+cgfbB695eiomLtsS8yjEqg3Z2sDQjnPLSoxXq32TACc9Mn9ify5qScJqBS+yRPVPU1hovrNJX/hyKP1y6pz06QWdq9slk+9MqaQZKZeBI+Pneuk+bSdkywVxXF9ZghHPuTrB6VjzGS+sWSOwAMO2aBsHHm32Ry61ebQ81xfxcvA/tx3yF1X8oT4K97ZG18E+cT8Utm4PY7bIHjiuf3yZg16H0rNRT9cvz7Hfyo0gr4/QvVtJ3ecpP24KPzdSThmvsmXewTP26olpT01kiIFGFPZlmVi1YIrE77aEz68qK/pECyWzxvK+w0XUrvDPvSul4V9n8fXar3ny8xB7d7g4kN1CpsCQ9/aiTqRQJAeTTbqxw3hkxXkBJt8zwCaHeWrQFK0zSP9wow9tZZk3Q+P7Tul2pD9xKhnkT+/2h3i42Cn0JfbasqKJen4AQAmP+GgUA6Lp/Fr76Lfk78K9lT4zalugy8oSKQU4lrsF4wvEh7p0Rf7qEyjiXi0ouZePVvhM5XuHBIucDaEyH8KhjWnD7Cbf6fb3x42bV79GDzwDNAizzkrlSpgSNMeFXrJlk3sSe0zT4sMG2b4OPN5w7AQAw3TUwuO23HYeiRnPe/8lq5I+erVkXq9JjsW+6XG5bhlfmbVK0j8ts/hugtzl+SFhuYVi/9fo4gsBtJyUfysWbQrXs2ijDa1kTgLLEtejwD2dbCi0H9sp8ZT9XODM1xzCmRKKmmFTmjkqFeRN72Nlssm5y8HHXM04HALjPvCW47a0fN+GSVxZH7aIJ9mPsZ0vjHiLRdALnST1gArXk4xWxy07Mfh0jdz4Ydz9+AhudQzAp+5Xglg5UErf0uz/7E22JazB+dd2BEtcNKHENRnmZ/M3J4xUetKbdqIejGOmYiavsS+CedGHC90xcFLo8bUyp9TJKpf99JnF9nSnFN09T9GzlIABA2bUfxS3TOLcuAKB7547Bbfs3L0dziuz9UeIajGWuu7HONTTuvpZuiT+ibGr28wCAiYu2AACGfRh/Xpjz7WshvPFvDL2d9SpcFPn6d85H4i7tt2zrgWDtPtz7E5+Je4yA577fgIXOB7DKNRwjHP4rnezDJQnfEz6q14HIGne67gXojZtiUpNUt0GTMmBvR3Ml9gBPq3Nly5zWJDRIaabzCYx1fBC3bLy26QN75Ztpxs8pAgBslpm4i55qGPe1vvbYfdATXcaF1+4Dns56L34AQgBC4KayD+OXiaNF2B/FqlMhX/d2/CuegH/YF6LENRjdqEj1sfWidE7u6u6DJSUxt6/cbs0/+GZhysQevPgZukDxO/rYf4v72t5lM2Ju77JyTMJ9NpAmHnt9wWbYYqzoFMUdXQNP1Nyj2RJjfywHxtUDxtXDXY6ZscuciN8FdJHzgYjngRWqAibEuSEa+NleyHoHAHCTwz/R2Zgv1+DvCv4gMON77+fYNwer0wpUiW6e6sVUiT1w+oIV7OYFmuy3cvmUmNtbHko8mnOlawQA4OV5m2QX/QAAPNs0atP2/drOSdJ69LfRG9+9VP6NL+Qjf/RsdBk9A6eMnon80bNx1Zs/Kzrmv+fGvom6pcqo26vtS3C9fRE+XrEDK0oO4Lq3lqRtUMriTaUYP4dXekq3eL+9vUeSa4oxwiAl1arkdSfcqtZmTgeTJXbpl65xo1bzOAk8i5R3U+xlUz8tAAB8szr+Kk/xrgJirfcaUA/HsGF3WG1p2duKYylxDcYq13Bsdt2MEtdgzCy9XPF7Y5mzbg/ezHo1YtsI+zfBx4XbD2LS4q04nIaBKTe/uwJvLtyi+X5ZJCGAdrQDg+2RV8+fFO5Ian9lbvPNqlk1GxW5bkGR6xY9QgkyVWIPCTuV10yKW2q577S0R7LU6R/sdLvj+6Tenyj5DLL/GHP732zxFzX4zXUHvp3wgL/r42e3Aj88nFRcAd6N8X+us8OmV4h1o5mI0N++ImJbG1vkfDHPf78RnZ+ci1GfreZBQSb0x4EyzHU+jGezYl/1qmXEKXDlUFhF8zRSv9ZxOpgqscesp3cO9VfHkFkRL831dlO+8yQvAfPoQER7s5zAnO1KPJP1LlbH6B44KuvThO97KOtTPPHieGD9l4qPFY99xsC4r93sCC0OMuidZVGvl8eZ07xqDyXAP2Vrn5f98+rsP1aR1DQM4U6mXchGJbb89l+UHUzPak8EH9Y7b8U4x3uAz3w1Ta3VxTGsdA7H9fZFKt4lUOIajGbwD+X/dvWfMuWNa43zdjzkCN2vm5z1EgB1q7BpxWSJPU5TzIglwPD/Aa3Pi9jcrO8oxfs+MuuxVMNT5MHXInuu5CBxb5pY7dx5JN/jYEr2S+oCS8IA+zKEt7JWHc076b9bY77vJ+d9qBNjXp2t+46jvNKLbk/PR9en5iW9+lXh/E+xwPkgNrmG4OSZV6Lma/Gv3JxwIxvJNQW9nPUWalEFhjjmAU/WB44rW/nKqla7hqEBHcX4rMlQOhLgvSz/usU/u+7FKbQTX68yTmIfap8d1VEgFoK/OSqHTkTMMnupfSVKXIPheqY+sGddmqONZKrEHvqwVEnsJ3UE8s6IKj3onFaK95yzckLE8z070tM++43z8Yjndzpi3OwMo6i3jY4edHwSfHz/J6vxbtgo1TMp/vQBa11DMdrxcdT20x4Pjbr1JLm48dZF8cc5VFXkugWbXEMw1vG+6uNcY6/yR/fP+D2vqpsS1w2KyvWwhaYGmed8KF3hqNbd9jseywpNn/GCY3LcskRArkwFDdP/rlVoipgqsT8UaIJQePO0RrZd3QHG1g3+azKlq8rolPOGtSXfIZPYt7puTFscWhjp+AZ1wz7UT84KfVEnZCdeaOQOx7cocQ3Gb85hiFXD6/DEHFz80iLVMf3doX6q5Fscc1W/J5rxur2lW48EnQb2HpFf6k4Y9JzNyH464nlv++q4ZYkAGyWuhFRm+P6RI6NH00iOK/mwL694Ft85H9EwGvWOLHwVuRffr7i8zydgsxnzCwD4L8Hzy6dHbW9Gym6E5dIxlLhuwP3uEfjKdx66UDG+dj4BAGhf+i4mLirGnRe0ldlL6gJd1CqQndwOjPsrSpv7HF/Efe3sZ+XHmax3Rp80Iy7ZaIcXY79Zj7FXdozYHlh7WO7K+lBZJRqlLbpopqqxBxApD7tX+WsRz38X+RpHo17u/8aqKj9n0cLgY6Ou5P5wjGYVtV7JfgslrsHBpA4AG1y34cUfijBzVfxuoVoxQjc1sznHltpYAa9JUlAWvHg/xijbVxdsBoEipt2IJdP9vcxxVqMorxrtCvs7ucTbAYPPaZmOgJJSonBFoVULQr1gZq9JTw+PVI1wfAsX/INSjlV4NB1oQvDh3hmrgrWjdGtLyc74WQ2r7CmLPGfJ3sjWVvRnNztsjqSOVIIi5xA0xkGccHv8TTGy98Iy+9kwaWJX7rQmdZBfPh355dMxuPIxPHtNJ71D8vNUYKbCHgBjskK14XcWGXe+lY2uW2GDDwVPz8NHy7Xrz9ua/H3f2z+ReDpjrcxXcBMv5hfZiLNB6ejqBOMt4vlQmlxPT7YYiT0rLLHfbJ8LJ1XiAvuq4Epbsd4TTqhoZdCCORO7ii/Q9/eeJ19ID083xoY/lM/VHOgL6zigfg70THrO8R+UV/rw9cz4ba9q/egMdVvNHz0bZz0zX7N9J6tpzPsH1Syxb56X8GUlg5aqpsPuNo3mR0pBU4ruthprFPqLWe/A5nODSElTDNfYFVB+kogIb9/o7+GyZqyCOVMyaN/m5YrLLlvnT+hvZL2RrnA08Q/HIpxMu/CFc5ym+73MFhrBWnq0Ale88VPMm2zbNFowO9Gi5UBkDS6outXYjyS+71GT5OeLMWKvmH85Pk/4+lVh3Vw7evz3GFrGGHQXIcOfDZMmdnX6np6Hkuf7I8cVvWyenj53Pqm47AUzu6O80ou2NuMM4IhngVNuYRH1JmW/6h/sIbXjr42zluSF/16kyfEutCee1K3AZtwmsUzxeFO/j2LExP43e+ImpMi1E3wgIfB4VuKxEyLDqdaU3R2rXc1IsvTlgUi81pH1PeyYgXGeIVHbKzxezFmv7TJk+45VoGFtZ8zX/KMrq6pen8vdv32PFjJl1Ey3kcp7MsEfV2SFoqV3J5r+OUf2vdwUo0Q1TewXntBiEI253eqI/BIV7/XPvfPy3E2452MtR34KrNwef456BrTYLZ/QrO5qz3fIdse+egwnuCmGMTmhJoA+Ly/G6h2HsDcNS7GNnParujd4qs9ycMzPDq+iC7VMN8VwYmemE5g4KmD7gTLYZGpEscYMJOprTxDwqB0GXk2vJKszu/ABGe7KqIQ529hZtXahfTXCx7EoaYK5QLqpWvxMPzjs/i+iEFq3inNir278i7vLJ3ZT1tiJqC8RFRFRMRGN1mKfjCWSh+QWZNgTNjEVL+vBUmWHF1Awj5Ppbp4SkR3AmwD6AegAYBARdUh1v4wlstR1d1LvO/eF0Lw7yUx7QPAFFzGPxn8qqhsbfBBQMItshi/mtGiKORtAsRBiKwAQ0QwAVwH4PeG7GEuRmm5xEWXH+v9L9OEf7piN4Y7ZwbIB21wJ3jTtOsXxMHP6btbnCF8JuD6OYFVxoez7Wrq3YMkn49HkQCFq9BuHvPz0LtupRVNMMwDhK9fulLZFIKJhRFRIRIWlpaUaHJYxxjLr8sLbo7ZddCjxSNWAnhueRp09S7F3e/qnTdCixh7rIiPqmlQIMRnAZAAoKCjQ95r1lu8gZv0LJLwQ7uNA5XFQXhf/CjgOF5BdE7BnA2X7AU8FhKsu6MguoFYjgOzAiQNA3eZAxTF/+cN/ALn5gM8HHP0TolYj0NHQLIzClgXUyAV8laDKcuz3urDDWx9dbFVWaXLUAE7qABzaARzfC9iz8Ye3ASp8NhyHCzb4cEzUQDfbZjipErA5IIQPRDaIeq2Awzv9P1OdPFBlGeD1ABWHgTp5EOVHQJXHw2JygJw5/p8loEFbYH8xRHYdkNcN4azt37fDCSo7CFQeh6jTFGRzALUbA7uiayrrfa3gghuVcOA02w7AVQ+i4ghI+CBqNgCV7YcgO0iEzb1RrxVwbC/gORG1Pw9sOCpqwgU3ikRzdLHFXm6P6W+7rzFsEGhhS63idljURF2KXjrRCI794wuM+WQ5Ong2gCBQ0bI3bj+vDRZNHy8tFRnfkX+uQFa9fLR3pn8EvBaJfScQMQCtOQBjj3vP7wW6yz9Pi5KmL7XNY1XLV32e4/Xh6ke/R2/banyQ/QJw8kXATV/F3FcLIdB6zHdR26cMKcDF7U8K7jv8GLHilYsp1mtK9x2waschXF1ljdaSsf0T7k/OkeNudH0qNNmUUUclVnexFlq53T4bj4ctL5fIPe6R+MbXK/i8h209Ps5+RrP4khH4mcI/c7Xb98G4B8/HX0fK8XPxPgw/tzWICG9k7cMA3zL8JerhpHHb/YXH1g2+b463AJc1OzVjsWuR2H8BcAoRtQawC8BAAPztSyDLrrwFjOL0jW53Uh2twtFM5+Z18dzfOqFvxyY486nEM/8pVb9WNgae1QL/3VSK3Yfll1pj+it5vr/0qD8wVlliv7lHK7x+Rf+wLf2BzV2BaddqHp9SwZ9jbOT2+rWyUb9WNtrn5QS3Xdu1OZCgqd1jtu6OQggPgLsAzAGwAcCnQoj4CyEyAMCS0ReFPVN/y7x5bg3tgtEIEWHQ2S2RWysbN5zTEm0a1dJkv89fewa+HtlLviAzrZgVmFP6AGMP+/8ZXKA7Y/hPcajfW8HHXiU9ZzSkyZ8RIcR3Qoh2QoiThRDpu34686a07TrTmtargQ9uPTup9w6VLv+M7JlrOuHHBy7QbH8NaiW5DikzBdkrUMMn9+jbhrWdoWROtsyOBTXeWNhEnDlAtvGaIDLtuoLmeoeQcQ67LewSn5lGVk1FxeoomVLb0Mk9uqLlqBFqY69TM1E/We2ZK7EzAEC7xtX3j9ufOZ31DoElcHWXppEbHtqm7QH+75C2+0unxqG+6j4yYVMMy5wfH+gNm4IhzFZ1PK+73iGwBF75R5fIDVka11SJMpPcVSbi2P23Q9/TGTnR/d/TiRO7CayVlvRb/OCFaNOots7R6KvtP/Rf7JjFl5F7P5k4RpveqooHZqegGCl+p2iIMkfdqO3pxIndBOq4slDyfH+0bKCsvdLKyMYfWdMZ9l+9I1DvqjdVFY9ZY3f4V9/aI+pnfEZn/pYw8xnyrd4RMDWadpEvozLzTTz1fewUDVF+95okg5KR01S+TJiT6vqbnCLWBajTBL/3fAXD3P/SMjJFTJbYefY8BqD1+XpHwGJI2Gvpsb2aHuvOQdeg+bgtcDVoBTxcoum+k5ElLbaRbY/8A/VXqwE4gJyMd082WWIHr1LD/PqM0zsCpobDCVzxWnr2XSM3PfsNaNxRtojX7m922VWjXeQLwbb3zDJfYmcMAHomNx+7rF73pWe/DOg8KH37Tmcf94sfly1Su14D/K1iLOZ1fCFie04Nf//8lvUze3+MEzszJ5sd6896Tvv99hmr/T4trqD8LflCQPBmYmyp12k9I5anvI+YWsv3kOnTvjH+OXgg7rg0cpxFt1a5mHRTNzzav316YouDE7uu+J5BKjp2v0z7nXJTn2r7kNmufPE4TkrT4hXZ8rVtIkK/TnkxJ/i7rGMTuLJ4gBJjyuTmA03PjNq8W9SPeB5rSllmMAqnHpD1SIIZw2+eCXQdktx+TTIZWQAndl1x7TAlNjswbBFwxeuhbY+VokfFhKjmgXKReC6SDb6WuF9kvluaVbx/61mp7UCrnk7ZtbBvVJweOG0uAK58PfZrFmOuxJ7E4sOsGug2BDj9WqBlD8DhnwWyApGJ/PSKKQl30c/9PLI6XZ22EK1O8TfziYOhx+c9AIwqBh7cqqi5Q6mGtZ3IL4+cB/5/3tM1278ZZHYuSU1YqZbLf6g0c927EU/d0kd7g68FVj1xCbo8OQ8dy6dgvSv2nB0rH+uDujXSv2SZVdWvqXBaZZstI00am5+5HPmPTkOJ6wYcETVx6LrPgq8tqD0AFx+blfYY9GTCxG5BfMNOcxXIxnkVr2CvyEWRlHSOowZ2i/rIowNR5RvUTtRjg8np3KKe3iFEyLLbkO2wB++vbO2UF3ytyeCJwGRrJ3ZzNcVYFTcxpcUOcRKuOetkAMCoS/0DR9b42ugZEsugDU/2DT4OnxG1Q14O8suno1P5f1Doaxf1vumeCzMSXzpxjV1XXFNPt39JCX1Iz3z8e+4mvOS5HpfZEyxOCfgXJCYTzfvNYrLbKOY0B4Hh/UdRE9e5x8Z8r9kXbeYaO7O0xnX8kzPVcWXhlMa1sUm0wGnl7+Ee98hgmfPtH0W8Z6LnqozGaHZnNDdGP3YWwoldV9wEk0lXSav7lMOJhb5Q//dvH7g0otwH3jQMfLKwb+46V+8QVBt3pfz8L2ZmssRu0UTIN08zolWDWsHHR1Ez2JZq4zneq50hPfP1DiGkR3rWAAAP+ElEQVStzPeJtmIO5JunGTHgjLyI5494/on88umRc2gDePCyUzMZlmFM9fRB2/IPg8+LfNVv0XSrMF9itxQr/pUyLiLC9d2ik1XVxO6opmvKPu65DR44kF8+Hfnl02Mu81bVR56LMxBZetzR+2S9Q0gbTuy64pp6po27yt+2+vXIXsFtVVvCbuzeCldVPJnJsHS3UzSM2qbkz9tjnswu0qylh/ta98qMuzsaAbexa6pNw1rYuu94zNdqZjuiusBlV5mRr5bTgWLRLG3xGcXn3vPxjbcH1vpaoxzRI0dt8OkQVeZkelWjTOIaO7Oci9s3VlRu4FktAEQOXgk4jhrYL+poGpeRlIocjKq8A4t9nXEQOTgBV9yyW3x5MbefVa5uwWcjSricn4lxYjcCvnmqqcHntFJU7tlrOqH4mX5xX+9W8bbiY15b8X+KyxrBWQp+tj3S9Mc3u0djq69JcPte4Z8+wLr1XfMzV2K3XALkr0Y6tG5YC2/d0BVrxl6asJzNRnDEWBgBACbe0BVqfj9GbrqZ4on/xyuRkZX34G73XdiFRhF9+33SeVFyc9UMip7uK1/IZMyV2AFYKxla44thRP065SHHlfxsjW0a1ZIvFOYw1JXPpKc8N+Fm98Oq33cIdfCtrycA4LjUVPOd92wI6Ttos8jn1+nI7OpGmWDCxG5BFr6JU108dXUnvUNIaLEvtBZnMs1GHuFPfodEbYuk80j1alprymZO7IzFkN9AXQ3cbqI/ziuF+m5+s3zd8ZbnCjzvGRTcZpWmGAD46s5e8oVMhBO7EVju3oH5BRYffqnyOkXljTqoaaLnyuDjke57MMvbXfF7n7kmtOqQBw684BmEI6iFBypH4BdfO/yFXE1j1VPrhrXQrF4NANHdX82I+7EzFkfJ8/1x3ZiNisrabIT1vlboaNue5qjUedEzMPh4tq87ZvuUJ/YbzmmFkxvVxsDJyyK2L/V1xPVu602i9fPoi/QOQTPm/9NkBSa6jK9u3FDW9ppbMwu7pe6BZjfq0nYofKwPAOCc1tb4maobkyV2brJgmbVGxF9x6YXKUG24IL8+Hqu8LRMhpdWL152Buy46BQ2lpQKJCPf1OUXnqJhaJkvs4Noty7D4n7fJ3tCoRSJgDxpkIqC0een6zvh7QYuo7ff1iV4+LiDbYb4UUh3wb8UI+OapoXUvfyPmdi9C/Z9rZ5v/dtW1MWa+lLPp6eQGP7H0Mv+nkbE0eurq0/H41+tky8Wab0Zvd7rvidr2WP/2eHr2huBztXOlWHVuFavhxG4E3LxkWKc0rq24bMnz/YGx6YtFrV98p0U8r1sjC0PPa4Oh58W/bxAPJ3RzSakphojGE9FGIlpDRF8RUT2tAmPMCAIjEntXvKyo/D3uu9IZjiql8H8de57cAIWP9cHSMdbpzscSS7WNfR6A04UQZwDYBGBM6iExZhynNckBAJQKZXWWWSr6iafTFRVPBx9PG3oOGtZ2oqYF7gMwZVJK7EKIuUIIj/R0GQBeJJFZzmP926MswXzl4XwG6Y9QJEK9W6y8oASLTctP4W0Avo/3IhENI6JCIiosLS1N7gjce4Tp4KYe/vndl/naKyp/fsUr6Qwnoa+9PdG2/MPgwKp2Jym/R8CsQzaxE9F8IloX499VYWUeBeABMC3efoQQk4UQBUKIgkaNGqUQMtc+WGYFpnUd6H4c0z0XJixrI+APcZLsPvPLp2sSW9V93ld5FzxhfSLm3t9b8+Mw45NtdBNC9En0OhENATAAwMVCcJWaWdsjnn/iEc8/476eWzMb+4+7kV8+HSWuwTHLnFb+HgDgG28PXGlfmnQsbmHHQPfjOIg62CZiL1/HqqdUe8X0BfAwgCuFEGXahJTAiQP+f4yZwA3u2H0JyuEfrn9P5d14vPKWpPa90dcC7Sqm4lfRjpM6i5JqG/sEAHUAzCOiVUSkfJHIZKz7Iq27z7hWPYHmZwOXPKl3JEzGikcuVv2en32dokatLq/St3yqN/HyfQHveUJL033l7YW+7hdUx8Oqj5T6Pwkh2moVSLWUXQsYOk/vKJgCjXOU9Yqp2hb5F3KxzpeP020lAIB73SOj3tOn4kXMdz6UcL/Pem7AblEf73r7RbShJ7L6/5T90WDWY4y+WYxZxB29I0d1CtgwwP0s+lc8i39XXh9zorBikbiX8ALvmaiEA5O9VyhO6oB/pCmrnsyV2Ad9AvQwzsg+Vr0MPbd1zO2HRc3g4ys6NwUAfDGiR0SZ9SIfE7zXxN33be5RcV97y3OFmjAZA+nRkaWgoEAUFhZm/LiMaW7DLKBJJyC3VdRL+aNnx33bpJu6YfjUlRHbulAxvnY+Eb2fJLtG8vwu1kNEK4UQBXLleIwxY6loPyCpt13WsUnUtlWiLcqEEzWpItWoMPaKDinvg5mXuZpiGLO4DhXv4ZTyD/Fk5U0AgKsrkusxZTfgNMIsc7jGzliGfTGiZ8LXK+HAu95+eNeb/CIWRpwfnmUO19gZy7D8BjXlC6Xo+m7RS9yx6oMTO2Npold3w23PXc5rkVZz/NtnLE2GnR97paJ0TqM7+JyWPE0v48TOWLrc0fvkmNtdWen72j17Tae07ZuZByd2xtIk3v1LXsmIpRt/whhLk0w2iQw6uyUu7xTdN55VT5zYGcugs/Pra77PyTd1w6UxBjyx6oubYhjLoH9d2k7T/Z3eLAc92zbUdJ/M/DixM5YBL/+9MwDg1JPqBLcN7x2714was+4+D7WdfOHNInFiZywD/ta1OUqe74/cWtnBbWP6teeJulhacGJnjDGL4Ws4xtJo1t3non5YLV1LG57sm5b9MvPjxM5YGp3erG7a9l0j2562fTNz46YYxhizGE7sjJlQ9zba94dn1sGJnTGTcTpsmDGsh3xBVm1xYmfMBFrWD83hnlszPTdjmXVwYmdMZ1l2+TllFo66IPi4cY4zjdEwK+DEzpjOHDb5r6HdRlg79lI0z62Bz+7gZhiWGCd2xnT246jeisrVcWXhp4cvgtPB3RxZYpzYGdNZXt0aOCs/V+8wmIVwYmfMALw+oXcIzEI4sTNmAB2bxh+huu25yzMYCbMCTuyMGcCgs1vGfY0Xp2ZqcWJnzAA4dzMtcWJnzADCF+BgLFWc2BkzAJuN8M7NBXqHwSyCEztjBnF6sxy9Q2AWwYmdMYPIq1tD7xCYRXBiZ4wxi+HEzpiBfHvXuXqHwCyAl8ZjzEA6Na+Lxwd0QK+2DfDJLzuw90iF3iExEyIhMj+UuaCgQBQWFmb8uIwxZmZEtFIIIdt9SpOmGCIaRUSCiBpqsT/GGGPJSzmxE1ELAJcA+CP1cBhjjKVKixr7KwAeAsDT0zHGmAGklNiJ6EoAu4QQqzWKhzHGWIpke8UQ0XwATWK89CiARwBcquRARDQMwDAAaNky/kx2jDHGUpN0rxgi6gRgAYAyaVNzAH8COFsIsSfRe7lXDGOMqae0V0zS/diFEGsBNA47YAmAAiHEvmT3yRhjLHU88pQxxixGlwFKRFQKYHuSb28IwExXBWaK10yxAuaK10yxAuaKtzrF2koI0UiukC6JPRVEVKikjckozBSvmWIFzBWvmWIFzBUvxxqNm2IYY8xiOLEzxpjFmDGxT9Y7AJXMFK+ZYgXMFa+ZYgXMFS/HWoXp2tgZY4wlZsYaO2OMsQRMldiJqC8RFRFRMRGNzuBxWxDRQiLaQETrieheaXt9IppHRJul/3Ol7UREr0txriGirmH7GiKV30xEQ8K2dyOitdJ7XiciSjFmOxH9RkSzpOetiWi5dNxPiChb2u6UnhdLr+eH7WOMtL2IiC4L267p74GI6hHR50S0UTrHPYx6bonofukzsI6IPiYil5HOLRG9S0R7iWhd2La0n8t4x0gi1vHS52ANEX1FRPWSPWfJ/F7Uxhv2WsTU5XqfWwghTPEPgB3AFgBtAGQDWA2gQ4aOnQegq/S4DoBNADoAeBHAaGn7aAAvSI8vB/A9AALQHcByaXt9AFul/3Olx7nSaysA9JDe8z2AfinG/C8A0wHMkp5/CmCg9PhtACOkx3cCeFt6PBDAJ9LjDtI5dgJoLZ17ezp+DwA+ADBUepwNoJ4Rzy2AZgC2AagRdk5vMdK5BXA+gK4A1oVtS/u5jHeMJGK9FIBDevxCWKyqz5na30sy8UrbWwCYA//YnIaGOLepfCEz+U/6geeEPR8DYIxOscyEfw76IgB50rY8AEXS40kABoWVL5JeHwRgUtj2SdK2PAAbw7ZHlEsivubwz+NzEYBZ0gdlX9gXJngupQ9kD+mxQypHVc9voJzWvwcAOfAnS6qy3XDnFv7EvkP6Ujqkc3uZ0c4tgHxEJsu0n8t4x1Aba5XXrgEwLda5kDtnyXzmk40XwOcAOgMoQSix63puzdQUE/hSBeyUtmWUdNl2JoDlAE4SQuwGAOn/wNw58WJNtH1njO3JehX+OfJ90vMGAA4JITwx9h+MSXr9sFRe7c+QrDYASgG8R/6mo/8QUS0Y8NwKIXYB+Df8i8rshv9crYRxz21AJs5lvGOk4jb4a67JxJrMZ141ij91ua7n1kyJPVa7aEa79BBRbQBfALhPCHEkUdEY20QS21UjogEA9gohViqIJ9FraY9V4oD/8vYtIcSZAI7Df7kZj57nNhfAVfA3BTQFUAtAvwT71/vcyjFsfET0KAAPgGmBTSpjSuYzrzbGmvBPX/5ErJdVxqXpuTVTYt8Jf1tWQGCa4Iwgoiz4k/o0IcSX0ua/iChPej0PwF6ZWBNtbx5jezJ6AbiS/LNtzoC/OeZVAPWIKDCbZ/j+gzFJr9cFcCCJnyFZOwHsFEIsl55/Dn+iN+K57QNgmxCiVAhRCeBLAD1h3HMbkIlzGe8Yqkk3FAcAuEFI7Q9JxLoP6n8vap0M/x/51dL3rTmAX4moSRLxantuk2lr1OMf/DW7rdKJDNwk6ZihYxOADwG8WmX7eETe1HhRetwfkTdOVkjb68Pfnpwr/dsGoL702i9S2cCNk8s1iPsChG6efobIG0l3So9HIvJG0qfS446IvFm1Ff4bVZr/HgD8D8Cp0uOx0nk13LkFcA6A9QBqSvv6AMDdRju3iG5jT/u5jHeMJGLtC+B3AI2qlFN9ztT+XpKJt8prJQi1set6btOeFLX8B/+d5k3w3wV/NIPHPRf+y6I1AFZJ/y6Hv11uAYDN0v+BXxABeFOKcy3889QH9nUbgGLp361h2wsArJPeMwEKb+bIxH0BQom9Dfx33YulD7xT2u6SnhdLr7cJe/+jUjxFCOtJovXvAUAXAIXS+f1a+sAb8twCGAdgo7S/qfAnGsOcWwAfw9/+Xwl/LfD2TJzLeMdIItZi+NugA9+zt5M9Z8n8XtTGW+X1EoQSu67nlkeeMsaYxZipjZ0xxpgCnNgZY8xiOLEzxpjFcGJnjDGL4cTOGGMWw4mdMcYshhM7Y4xZDCd2xhizmP8H9Epwok/ib4oAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(y_test_feat[:,0])\n",
    "plt.plot(ypred[:,0])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction of temperature2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 282
    },
    "colab_type": "code",
    "id": "vA5pcMdE6YIK",
    "outputId": "18da2ec7-6d0c-415d-fa71-46bb910cdb91"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0xecdf032a58>]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(y_test_feat[:,1])\n",
    "plt.plot(ypred[:,1])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction of absoluteHumidity1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "-MlxlN2QoTQD"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0xecdf07d438>]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(y_test_feat[:,2])\n",
    "plt.plot(ypred[:,2])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction of absoluteHumidity2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "m5pZ4IBKmLuq"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0xecdf13eb00>]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(y_test_feat[:,3])\n",
    "plt.plot(ypred[:,3])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction of differentialPressurePascal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "3d23vXo0l8Nb"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0xecdf1c8e48>]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(y_test_feat[:,4])\n",
    "plt.plot(ypred[:,4])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction of fan1Rpm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0xecdf253518>]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(y_test_feat[:,5])\n",
    "plt.plot(ypred[:,5])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction of fan2Rpm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0xecdf2e92e8>]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(y_test_feat[:,6])\n",
    "plt.plot(ypred[:,6])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Performance of loss plot on train & test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,8))\n",
    "plt.plot(model.history.history['loss'])\n",
    "plt.plot(model.history.history['val_loss'])\n",
    "plt.title('Loss Performance of Predictive LSTM Model', fontsize=24 ,fontweight='bold')\n",
    "plt.ylabel('Loss [MSE]', fontsize=20)\n",
    "plt.xlabel('Epoch', fontsize=20)\n",
    "plt.rcParams['xtick.labelsize']=15\n",
    "plt.rcParams['ytick.labelsize']=15\n",
    "plt.legend(['Train data', 'Test data'], fontsize=20)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.savetxt(\"D:\\\\berlin_sensordata.csv\", ypred, delimiter=\",\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dff=pd.(ypred())\n",
    "dff.to_csv('D:\\\\berlin_sensordata_.csv', sep=',', header=None, index=None)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "name": "Untitled8.ipynb",
   "provenance": []
  },
  "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": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}