Agritech_AB5Consulting.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 275,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from sklearn.preprocessing import LabelEncoder,OneHotEncoder\n",
    "from sklearn.model_selection import train_test_split\n",
    "import pandas as pd\n",
    "from sklearn import model_selection\n",
    "import sklearn as read_csv\n",
    "import numpy as np\n",
    "from sklearn.metrics import classification_report,accuracy_score,confusion_matrix\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn import svm\n",
    "from sklearn import metrics\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import label_binarize\n",
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "from scipy import interp\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn import preprocessing\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "import numpy as np\n",
    "from random import shuffle\n",
    "from operator import itemgetter\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "import pandas as pd\n",
    "from sklearn import model_selection\n",
    "import sklearn as read_csv\n",
    "import numpy as np\n",
    "from sklearn.metrics import classification_report,accuracy_score,confusion_matrix\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn import svm\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import label_binarize\n",
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "from scipy import interp\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn import preprocessing\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "import numpy as np\n",
    "from random import shuffle\n",
    "from operator import itemgetter\n",
    "from sklearn import preprocessing\n",
    "from gplearn.genetic import SymbolicRegressor\n",
    "from sklearn.utils.random import check_random_state\n",
    "import numpy as np\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble import BaggingClassifier\n",
    "from sklearn.model_selection import cross_val_score, train_test_split\n",
    "from mlxtend.plotting import plot_learning_curves\n",
    "from mlxtend.plotting import plot_decision_regions\n",
    "import itertools\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "import numpy as np\n",
    "import xgboost as xgb\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.gridspec as gridspec\n",
    "from mlxtend.classifier import StackingClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import GridSearchCV \n",
    "from matplotlib import pyplot\n",
    "#We choose 5 cross validation for our machine learning model\n",
    "n_folds = KFold(n_splits =5,shuffle= False )\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.metrics import jaccard_score\n",
    "from keras.models import Sequential\n",
    "from keras.layers.core import Dense,Activation\n",
    "from keras.utils import np_utils\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "import xgboost as xgb\n",
    "from sklearn.metrics import accuracy_score\n",
    "from numpy import mean\n",
    "from numpy import std\n",
    "from sklearn.model_selection import RepeatedStratifiedKFold # evaluate a given model using cross-validation\n",
    "\n",
    "# import packages for hyperparameters tuning\n",
    "from hyperopt import STATUS_OK, Trials, fmin, hp, tpe\n",
    "\n",
    "import keras\n",
    "import keras.utils\n",
    "from keras import utils as np_utils\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "from keras.layers import Dropout\n",
    "# Example of Dropout on the Sonar Dataset: Hidden Layer\n",
    "from pandas import read_csv\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import Dropout\n",
    "from keras.wrappers.scikit_learn import KerasClassifier\n",
    "from keras.constraints import maxnorm\n",
    "from keras.optimizers import SGD\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.pipeline import Pipeline\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "from torch import optim\n",
    "from torch.optim import lr_scheduler\n",
    "from torch.utils.data import DataLoader, RandomSampler, SequentialSampler\n",
    "\n",
    "from src.config import config\n",
    "from src.dataset import CustomDataset, DataTransformer\n",
    "from src.engine import engine\n",
    "from src.metrics import mbe\n",
    "from src.model import NeuralNet\n",
    "from src.utils import get_cat_emb_dims, init_random_seed\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "import os\n",
    "import pickle\n",
    "import random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "colnames=['date','time','voltage','humidity','target']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "df=pd.read_csv(\"sensor_readings.csv\",names=colnames, header=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 675,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>time</th>\n",
       "      <th>voltage</th>\n",
       "      <th>humidity</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2021-11-25</td>\n",
       "      <td>18:40:44</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2021-11-25</td>\n",
       "      <td>18:40:45</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2021-11-25</td>\n",
       "      <td>18:40:46</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2021-11-25</td>\n",
       "      <td>18:40:47</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2021-11-25</td>\n",
       "      <td>18:40:48</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334056</th>\n",
       "      <td>2021-11-29</td>\n",
       "      <td>18:51:17</td>\n",
       "      <td>1.43</td>\n",
       "      <td>-0.52</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334057</th>\n",
       "      <td>2021-11-29</td>\n",
       "      <td>18:51:18</td>\n",
       "      <td>1.43</td>\n",
       "      <td>-0.52</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334058</th>\n",
       "      <td>2021-11-29</td>\n",
       "      <td>18:51:19</td>\n",
       "      <td>1.43</td>\n",
       "      <td>-0.52</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334059</th>\n",
       "      <td>2021-11-29</td>\n",
       "      <td>18:51:21</td>\n",
       "      <td>1.43</td>\n",
       "      <td>-0.52</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334060</th>\n",
       "      <td>2021-11-29</td>\n",
       "      <td>18:51:22</td>\n",
       "      <td>1.43</td>\n",
       "      <td>-0.52</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>334061 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              date      time  voltage  humidity target\n",
       "0       2021-11-25  18:40:44     1.32     -0.44    dry\n",
       "1       2021-11-25  18:40:45     1.32     -0.44    dry\n",
       "2       2021-11-25  18:40:46     1.32     -0.44    dry\n",
       "3       2021-11-25  18:40:47     1.32     -0.44    dry\n",
       "4       2021-11-25  18:40:48     1.32     -0.44    dry\n",
       "...            ...       ...      ...       ...    ...\n",
       "334056  2021-11-29  18:51:17     1.43     -0.52    dry\n",
       "334057  2021-11-29  18:51:18     1.43     -0.52    dry\n",
       "334058  2021-11-29  18:51:19     1.43     -0.52    dry\n",
       "334059  2021-11-29  18:51:21     1.43     -0.52    dry\n",
       "334060  2021-11-29  18:51:22     1.43     -0.52    dry\n",
       "\n",
       "[334061 rows x 5 columns]"
      ]
     },
     "execution_count": 675,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>time</th>\n",
       "      <th>voltage</th>\n",
       "      <th>humidity</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2021-11-25</td>\n",
       "      <td>18:40:44</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2021-11-25</td>\n",
       "      <td>18:40:45</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2021-11-25</td>\n",
       "      <td>18:40:46</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2021-11-25</td>\n",
       "      <td>18:40:47</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2021-11-25</td>\n",
       "      <td>18:40:48</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         date      time  voltage  humidity target\n",
       "0  2021-11-25  18:40:44     1.32     -0.44    dry\n",
       "1  2021-11-25  18:40:45     1.32     -0.44    dry\n",
       "2  2021-11-25  18:40:46     1.32     -0.44    dry\n",
       "3  2021-11-25  18:40:47     1.32     -0.44    dry\n",
       "4  2021-11-25  18:40:48     1.32     -0.44    dry"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>time</th>\n",
       "      <th>voltage</th>\n",
       "      <th>humidity</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>334056</th>\n",
       "      <td>2021-11-29</td>\n",
       "      <td>18:51:17</td>\n",
       "      <td>1.43</td>\n",
       "      <td>-0.52</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334057</th>\n",
       "      <td>2021-11-29</td>\n",
       "      <td>18:51:18</td>\n",
       "      <td>1.43</td>\n",
       "      <td>-0.52</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334058</th>\n",
       "      <td>2021-11-29</td>\n",
       "      <td>18:51:19</td>\n",
       "      <td>1.43</td>\n",
       "      <td>-0.52</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334059</th>\n",
       "      <td>2021-11-29</td>\n",
       "      <td>18:51:21</td>\n",
       "      <td>1.43</td>\n",
       "      <td>-0.52</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>334060</th>\n",
       "      <td>2021-11-29</td>\n",
       "      <td>18:51:22</td>\n",
       "      <td>1.43</td>\n",
       "      <td>-0.52</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              date      time  voltage  humidity target\n",
       "334056  2021-11-29  18:51:17     1.43     -0.52    dry\n",
       "334057  2021-11-29  18:51:18     1.43     -0.52    dry\n",
       "334058  2021-11-29  18:51:19     1.43     -0.52    dry\n",
       "334059  2021-11-29  18:51:21     1.43     -0.52    dry\n",
       "334060  2021-11-29  18:51:22     1.43     -0.52    dry"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date        0\n",
       "time        0\n",
       "voltage     0\n",
       "humidity    0\n",
       "target      0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.set_index('date')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time</th>\n",
       "      <th>voltage</th>\n",
       "      <th>humidity</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2021-11-25</th>\n",
       "      <td>18:40:44</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-11-25</th>\n",
       "      <td>18:40:45</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-11-25</th>\n",
       "      <td>18:40:46</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-11-25</th>\n",
       "      <td>18:40:47</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-11-25</th>\n",
       "      <td>18:40:48</td>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>dry</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                time  voltage  humidity target\n",
       "date                                          \n",
       "2021-11-25  18:40:44     1.32     -0.44    dry\n",
       "2021-11-25  18:40:45     1.32     -0.44    dry\n",
       "2021-11-25  18:40:46     1.32     -0.44    dry\n",
       "2021-11-25  18:40:47     1.32     -0.44    dry\n",
       "2021-11-25  18:40:48     1.32     -0.44    dry"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "wet    188669\n",
       "dry    145392\n",
       "Name: target, dtype: int64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['target'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x2bcf79ac130>"
      ]
     },
     "execution_count": 26,
     "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": [
    "df.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "x=df['voltage']\n",
    "\n",
    "y=df['humidity']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:               humidity   R-squared (uncentered):                   0.992\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.992\n",
      "Method:                 Least Squares   F-statistic:                          4.074e+07\n",
      "Date:                Tue, 30 Nov 2021   Prob (F-statistic):                        0.00\n",
      "Time:                        15:37:46   Log-Likelihood:                      6.3486e+05\n",
      "No. Observations:              334061   AIC:                                 -1.270e+06\n",
      "Df Residuals:                  334060   BIC:                                 -1.270e+06\n",
      "Df Model:                           1                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "voltage       -0.3129    4.9e-05  -6382.557      0.000      -0.313      -0.313\n",
      "==============================================================================\n",
      "Omnibus:                    35213.544   Durbin-Watson:                   0.002\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            18643.263\n",
      "Skew:                          -0.427   Prob(JB):                         0.00\n",
      "Kurtosis:                       2.220   Cond. No.                         1.00\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "import statsmodels.api as sm\n",
    "model= sm.OLS(y, x)\n",
    "\n",
    "results=model.fit()\n",
    "\n",
    "print(results.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DEEP LEARNING FOR SOIL MOISTURE SENSOR "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 617,
   "metadata": {},
   "outputs": [],
   "source": [
    "colnames=['date','time','voltage','humidity','target']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 618,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1=pd.read_csv(\"sensor_readings.csv\",names=colnames, header=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 619,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['date', 'time', 'voltage', 'humidity', 'target'], dtype='object')"
      ]
     },
     "execution_count": 619,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 620,
   "metadata": {},
   "outputs": [],
   "source": [
    "encoded=df1[['date','time','target']].apply(LabelEncoder().fit_transform)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 621,
   "metadata": {},
   "outputs": [],
   "source": [
    "remain=df1[['voltage','humidity']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 622,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Adding both the dataframes encoded and remaining (without encoding)\n",
    "data =pd.concat([remain,encoded], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 623,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>voltage</th>\n",
       "      <th>humidity</th>\n",
       "      <th>date</th>\n",
       "      <th>time</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>0</td>\n",
       "      <td>67244</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>0</td>\n",
       "      <td>67245</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>0</td>\n",
       "      <td>67246</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>0</td>\n",
       "      <td>67247</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.32</td>\n",
       "      <td>-0.44</td>\n",
       "      <td>0</td>\n",
       "      <td>67248</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   voltage  humidity  date   time  target\n",
       "0     1.32     -0.44     0  67244       0\n",
       "1     1.32     -0.44     0  67245       0\n",
       "2     1.32     -0.44     0  67246       0\n",
       "3     1.32     -0.44     0  67247       0\n",
       "4     1.32     -0.44     0  67248       0"
      ]
     },
     "execution_count": 623,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 624,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "target\n",
       "1         188669\n",
       "0         145392\n",
       "dtype: int64"
      ]
     },
     "execution_count": 624,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[['target']].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 625,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['voltage', 'humidity', 'date', 'time', 'target'], dtype='object')"
      ]
     },
     "execution_count": 625,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 626,
   "metadata": {},
   "outputs": [],
   "source": [
    "X=data[['voltage', 'humidity', 'date', 'time']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 627,
   "metadata": {},
   "outputs": [],
   "source": [
    "y=data['target']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 628,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Split data in training and testing sample with 70% for training and 30% for testing\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 629,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(267248, 4) (66813, 4) (267248,) (66813,)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape,X_test.shape,y_train.shape,y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 630,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Nous pouvons maintenant standardiser les variables, \n",
    "#c'est-à-dire les centrer (ramener leur moyenne à 0) \n",
    "#et les réduire (ramener leur écart-type à 1), \n",
    "#afin qu'elles se placent toutes à peu près sur la même échelle\n",
    "# standardiser les données\n",
    "std_scale = preprocessing.StandardScaler().fit(X_train)\n",
    "x_train_std = std_scale.transform(X_train)\n",
    "x_test_std = std_scale.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 631,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(66813, 4)"
      ]
     },
     "execution_count": 631,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_test_std.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 635,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(267248, 4)"
      ]
     },
     "execution_count": 635,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train_std.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 632,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# Create model\n",
    "# Using a sequential that uses three layers. Here are there two hinden layers with 16 nodes each and one output layer with one node\n",
    "# the keras.sequential creates the single stack of layers that are sequentially connected\n",
    "# model.add(Dense(16, input_dim=4, activation='relu', kernel_constraint=maxnorm(3))) is the first layer added to the model. it's function is to take the input with 4 paramters\n",
    "# model.add(Dense(16, activation='relu', kernel_constraint=maxnorm(3)) adds a dense hiden layers  \n",
    "##### They use ReLU (Rectified Linear Units) as the activation function. \n",
    "##### ReLU is a half rectified function; that is, for all the inputs less than 0 (e.g. -120,-6.7, -0.0344, 0)\n",
    "# model.add(Dense(1, activation='sigmoid')) is the output layers which uses the Sigmoid activation function \n",
    "\n",
    "\n",
    "model = Sequential()\n",
    "model.add(Dense(16, input_dim=4, activation='relu', kernel_constraint=maxnorm(3)))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(16, activation='relu', kernel_constraint=maxnorm(3)))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "    \n",
    "    # Compile model\n",
    "    \n",
    "#sgd = SGD(lr=0.1, momentum=0.9)\n",
    "model.compile(loss='binary_crossentropy', optimizer='Adagrad', metrics=['accuracy'])\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 669,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_65\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_195 (Dense)            (None, 16)                80        \n",
      "_________________________________________________________________\n",
      "dropout_72 (Dropout)         (None, 16)                0         \n",
      "_________________________________________________________________\n",
      "dense_196 (Dense)            (None, 16)                272       \n",
      "_________________________________________________________________\n",
      "dropout_73 (Dropout)         (None, 16)                0         \n",
      "_________________________________________________________________\n",
      "dense_197 (Dense)            (None, 1)                 17        \n",
      "=================================================================\n",
      "Total params: 369\n",
      "Trainable params: 369\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 634,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "24053/24053 [==============================] - 22s 911us/step - loss: 0.5313 - accuracy: 0.8433 - val_loss: 0.4074 - val_accuracy: 0.9862\n",
      "Epoch 2/10\n",
      "24053/24053 [==============================] - 22s 920us/step - loss: 0.3646 - accuracy: 0.9801 - val_loss: 0.2879 - val_accuracy: 0.9862\n",
      "Epoch 3/10\n",
      "24053/24053 [==============================] - 22s 930us/step - loss: 0.2634 - accuracy: 0.9831 - val_loss: 0.1937 - val_accuracy: 0.9862\n",
      "Epoch 4/10\n",
      "24053/24053 [==============================] - 23s 953us/step - loss: 0.1891 - accuracy: 0.9831 - val_loss: 0.1344 - val_accuracy: 0.9862\n",
      "Epoch 5/10\n",
      "24053/24053 [==============================] - 21s 873us/step - loss: 0.1457 - accuracy: 0.9832 - val_loss: 0.1009 - val_accuracy: 0.9862\n",
      "Epoch 6/10\n",
      "24053/24053 [==============================] - 22s 905us/step - loss: 0.1190 - accuracy: 0.9831 - val_loss: 0.0785 - val_accuracy: 0.9862\n",
      "Epoch 7/10\n",
      "24053/24053 [==============================] - 24s 985us/step - loss: 0.1003 - accuracy: 0.9835 - val_loss: 0.0632 - val_accuracy: 0.9862\n",
      "Epoch 8/10\n",
      "24053/24053 [==============================] - 27s 1ms/step - loss: 0.0882 - accuracy: 0.9836 - val_loss: 0.0540 - val_accuracy: 0.9862\n",
      "Epoch 9/10\n",
      "24053/24053 [==============================] - 25s 1ms/step - loss: 0.0795 - accuracy: 0.9839 - val_loss: 0.0482 - val_accuracy: 0.9862\n",
      "Epoch 10/10\n",
      "24053/24053 [==============================] - 25s 1ms/step - loss: 0.0739 - accuracy: 0.9843 - val_loss: 0.0441 - val_accuracy: 0.9862\n"
     ]
    }
   ],
   "source": [
    "# history = model1.fit(train_x, train_y,validation_split = 0.1, epochs=50, batch_size=4)\n",
    "history=model.fit(x_train_std, y_train,validation_split = 0.1,epochs=10, batch_size=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 636,
   "metadata": {},
   "outputs": [
    {
     "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": [
    "from matplotlib import pyplot as plt\n",
    "# history = model1.fit(train_x, train_y,validation_split = 0.1, epochs=50, batch_size=4)\n",
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 676,
   "metadata": {},
   "outputs": [
    {
     "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(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('loss graph by epoch')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 651,
   "metadata": {},
   "outputs": [],
   "source": [
    "result = model.predict(x_test_std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 652,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.00648215],\n",
       "       [0.9131422 ],\n",
       "       [0.93969595],\n",
       "       ...,\n",
       "       [0.8772094 ],\n",
       "       [0.99552685],\n",
       "       [0.997649  ]], dtype=float32)"
      ]
     },
     "execution_count": 652,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 672,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The soil is dry\n"
     ]
    }
   ],
   "source": [
    "if result[0]<=0.5:\n",
    "    print(\"The soil is dry\")\n",
    "else:\n",
    "    print(\"The soil is wet\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 673,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The soil is wet\n"
     ]
    }
   ],
   "source": [
    "if result[1]<=0.5:\n",
    "    print(\"The soil is dry\")\n",
    "else:\n",
    "    print(\"The soil is wet\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 653,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\couli\\anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:70: FutureWarning: Pass threshold=0.5 as keyword args. From version 1.0 (renaming of 0.25) passing these as positional arguments will result in an error\n",
      "  warnings.warn(f\"Pass {args_msg} as keyword args. From version \"\n"
     ]
    }
   ],
   "source": [
    "from sklearn.preprocessing import binarize\n",
    "\n",
    "#everything together \n",
    "predict = np.ravel(binarize(result.reshape(-1,1), 0.5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 654,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0., 1., 1., ..., 1., 1., 1.], dtype=float32)"
      ]
     },
     "execution_count": 654,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 655,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save it as a h5 file\n",
    "from tensorflow.keras.models import load_model\n",
    "model.save('model_rwet_dry.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 663,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[29064    78]\n",
      " [  806 36865]]\n"
     ]
    }
   ],
   "source": [
    "mythreshold=0.8\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "y_pred = (model.predict(x_test_std)>= mythreshold).astype(int)\n",
    "cm=confusion_matrix(y_test, predict)  \n",
    "print(cm)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 662,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot the confusion matrix\n",
    "f,ax = plt.subplots(figsize=(6, 5))\n",
    "sns.heatmap(cm, annot=True, linewidths=0.01,cmap=\"Greens\",linecolor=\"gray\", fmt= '.1f',ax=ax)\n",
    "plt.xlabel(\"Predicted Label\")\n",
    "plt.ylabel(\"True Label\")\n",
    "plt.title(\"Confusion Matrix \")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 670,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.97      1.00      0.99     29142\n",
      "           1       1.00      0.98      0.99     37671\n",
      "\n",
      "    accuracy                           0.99     66813\n",
      "   macro avg       0.99      0.99      0.99     66813\n",
      "weighted avg       0.99      0.99      0.99     66813\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#VISUALISATION \n",
    "print(classification_report(y_test, predict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}