notebooks/3 - modeling/c - Multiple Linear Regression.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "7b685402-f9be-4f62-bb9f-091a8d71d4be",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing dependencies\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from scipy import stats\n",
    "\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import preprocessing\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "import statsmodels.api as sm\n",
    "from statsmodels.formula.api import ols\n",
    "from statsmodels.stats.outliers_influence import variance_inflation_factor as vif\n",
    "from statsmodels.stats.diagnostic import het_white\n",
    "\n",
    "from math import sqrt\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "9e7149d5-d37b-4846-8335-256005a3c939",
   "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>Solar Energy (kWh)</th>\n",
       "      <th>Inverter 1 Energy (kWh)</th>\n",
       "      <th>Inverter 2 Energy (kWh)</th>\n",
       "      <th>Daylength in decimal hours</th>\n",
       "      <th>Billing Period Duration (Days)</th>\n",
       "      <th>Billable kWh</th>\n",
       "      <th>Power Bill Total</th>\n",
       "      <th>Tesla Bill Total</th>\n",
       "      <th>Total Cost</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Billing Period End</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12/17/2021</th>\n",
       "      <td>590.2</td>\n",
       "      <td>282.2</td>\n",
       "      <td>307.8</td>\n",
       "      <td>251.94</td>\n",
       "      <td>30</td>\n",
       "      <td>1123</td>\n",
       "      <td>195.13</td>\n",
       "      <td>185</td>\n",
       "      <td>380.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1/19/2022</th>\n",
       "      <td>366.8</td>\n",
       "      <td>77.1</td>\n",
       "      <td>289.7</td>\n",
       "      <td>266.70</td>\n",
       "      <td>32</td>\n",
       "      <td>2046</td>\n",
       "      <td>347.39</td>\n",
       "      <td>185</td>\n",
       "      <td>532.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2/16/2022</th>\n",
       "      <td>738.0</td>\n",
       "      <td>355.5</td>\n",
       "      <td>382.6</td>\n",
       "      <td>244.05</td>\n",
       "      <td>27</td>\n",
       "      <td>829</td>\n",
       "      <td>146.63</td>\n",
       "      <td>185</td>\n",
       "      <td>331.63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3/21/2022</th>\n",
       "      <td>1280.5</td>\n",
       "      <td>608.4</td>\n",
       "      <td>672.1</td>\n",
       "      <td>320.33</td>\n",
       "      <td>32</td>\n",
       "      <td>160</td>\n",
       "      <td>36.27</td>\n",
       "      <td>185</td>\n",
       "      <td>221.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4/19/2022</th>\n",
       "      <td>1323.8</td>\n",
       "      <td>641.0</td>\n",
       "      <td>682.7</td>\n",
       "      <td>312.83</td>\n",
       "      <td>28</td>\n",
       "      <td>473</td>\n",
       "      <td>87.90</td>\n",
       "      <td>185</td>\n",
       "      <td>272.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5/18/2022</th>\n",
       "      <td>1511.0</td>\n",
       "      <td>736.2</td>\n",
       "      <td>775.2</td>\n",
       "      <td>339.47</td>\n",
       "      <td>28</td>\n",
       "      <td>352</td>\n",
       "      <td>61.97</td>\n",
       "      <td>185</td>\n",
       "      <td>246.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6/17/2022</th>\n",
       "      <td>1546.6</td>\n",
       "      <td>752.0</td>\n",
       "      <td>794.8</td>\n",
       "      <td>369.80</td>\n",
       "      <td>29</td>\n",
       "      <td>882</td>\n",
       "      <td>133.15</td>\n",
       "      <td>185</td>\n",
       "      <td>318.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7/19/2022</th>\n",
       "      <td>1594.7</td>\n",
       "      <td>775.0</td>\n",
       "      <td>819.6</td>\n",
       "      <td>395.68</td>\n",
       "      <td>31</td>\n",
       "      <td>1410</td>\n",
       "      <td>206.94</td>\n",
       "      <td>185</td>\n",
       "      <td>391.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8/16/2022</th>\n",
       "      <td>1123.7</td>\n",
       "      <td>547.8</td>\n",
       "      <td>576.0</td>\n",
       "      <td>330.32</td>\n",
       "      <td>27</td>\n",
       "      <td>1448</td>\n",
       "      <td>212.25</td>\n",
       "      <td>185</td>\n",
       "      <td>397.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9/19/2022</th>\n",
       "      <td>1108.3</td>\n",
       "      <td>622.9</td>\n",
       "      <td>485.2</td>\n",
       "      <td>368.61</td>\n",
       "      <td>33</td>\n",
       "      <td>1655</td>\n",
       "      <td>260.42</td>\n",
       "      <td>185</td>\n",
       "      <td>445.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10/17/2022</th>\n",
       "      <td>482.5</td>\n",
       "      <td>482.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>273.81</td>\n",
       "      <td>27</td>\n",
       "      <td>657</td>\n",
       "      <td>120.82</td>\n",
       "      <td>185</td>\n",
       "      <td>305.82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11/17/2022</th>\n",
       "      <td>285.4</td>\n",
       "      <td>285.4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>273.39</td>\n",
       "      <td>30</td>\n",
       "      <td>1115</td>\n",
       "      <td>198.85</td>\n",
       "      <td>185</td>\n",
       "      <td>383.85</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Solar Energy (kWh)  Inverter 1 Energy (kWh)  \\\n",
       "Billing Period End                                                \n",
       "12/17/2021                       590.2                    282.2   \n",
       "1/19/2022                        366.8                     77.1   \n",
       "2/16/2022                        738.0                    355.5   \n",
       "3/21/2022                       1280.5                    608.4   \n",
       "4/19/2022                       1323.8                    641.0   \n",
       "5/18/2022                       1511.0                    736.2   \n",
       "6/17/2022                       1546.6                    752.0   \n",
       "7/19/2022                       1594.7                    775.0   \n",
       "8/16/2022                       1123.7                    547.8   \n",
       "9/19/2022                       1108.3                    622.9   \n",
       "10/17/2022                       482.5                    482.5   \n",
       "11/17/2022                       285.4                    285.4   \n",
       "\n",
       "                    Inverter 2 Energy (kWh)  Daylength in decimal hours  \\\n",
       "Billing Period End                                                        \n",
       "12/17/2021                            307.8                      251.94   \n",
       "1/19/2022                             289.7                      266.70   \n",
       "2/16/2022                             382.6                      244.05   \n",
       "3/21/2022                             672.1                      320.33   \n",
       "4/19/2022                             682.7                      312.83   \n",
       "5/18/2022                             775.2                      339.47   \n",
       "6/17/2022                             794.8                      369.80   \n",
       "7/19/2022                             819.6                      395.68   \n",
       "8/16/2022                             576.0                      330.32   \n",
       "9/19/2022                             485.2                      368.61   \n",
       "10/17/2022                              0.0                      273.81   \n",
       "11/17/2022                              0.0                      273.39   \n",
       "\n",
       "                    Billing Period Duration (Days)  Billable kWh  \\\n",
       "Billing Period End                                                 \n",
       "12/17/2021                                      30          1123   \n",
       "1/19/2022                                       32          2046   \n",
       "2/16/2022                                       27           829   \n",
       "3/21/2022                                       32           160   \n",
       "4/19/2022                                       28           473   \n",
       "5/18/2022                                       28           352   \n",
       "6/17/2022                                       29           882   \n",
       "7/19/2022                                       31          1410   \n",
       "8/16/2022                                       27          1448   \n",
       "9/19/2022                                       33          1655   \n",
       "10/17/2022                                      27           657   \n",
       "11/17/2022                                      30          1115   \n",
       "\n",
       "                    Power Bill Total  Tesla Bill Total  Total Cost  \n",
       "Billing Period End                                                  \n",
       "12/17/2021                    195.13               185      380.13  \n",
       "1/19/2022                     347.39               185      532.39  \n",
       "2/16/2022                     146.63               185      331.63  \n",
       "3/21/2022                      36.27               185      221.27  \n",
       "4/19/2022                      87.90               185      272.90  \n",
       "5/18/2022                      61.97               185      246.97  \n",
       "6/17/2022                     133.15               185      318.15  \n",
       "7/19/2022                     206.94               185      391.94  \n",
       "8/16/2022                     212.25               185      397.25  \n",
       "9/19/2022                     260.42               185      445.42  \n",
       "10/17/2022                    120.82               185      305.82  \n",
       "11/17/2022                    198.85               185      383.85  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Reading feature set into memory and instantiating dataframe\n",
    "\n",
    "df = pd.read_csv('../../data/processed/cleaned_feature_set.csv', index_col = 0)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "17b7463a-0f78-4689-9ff6-07b2ad6979b2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creating NaNs from 0s in `Inverter 2 Energy (kWh)` column\n",
    "# and forward-fill them with the mean for measuring error later\n",
    "\n",
    "df.replace(0, np.nan, inplace=True)\n",
    "df = df.fillna(df['Inverter 2 Energy (kWh)'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "b5d712c3-e356-4884-89d2-c6b4cf49e87d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Correlation matrix\n",
    "\n",
    "sns.set(rc = {'figure.figsize':(15,8)})\n",
    "corr = df.corr()\n",
    "mask = np.triu(corr)\n",
    "sns.heatmap(corr, cmap=\"Blues\", mask=mask)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e4b986b8-bf7d-4bfe-9fe4-fd9580018c46",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Test set is two rows, Oct. and Nov. of 2022\n",
    "\n",
    "X_test = df[['Daylength in decimal hours', 'Billing Period Duration (Days)']][-2:]\n",
    "y_test = df['Inverter 2 Energy (kWh)'][-2:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "1a0d9aeb-185c-4ada-be38-e7fdd9afb9f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Training set is first ten rows (12-2021 thru 9-2022)\n",
    "\n",
    "X_train = df[['Daylength in decimal hours', 'Billing Period Duration (Days)']][:-2]\n",
    "y_train = df['Inverter 2 Energy (kWh)'][:-2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7f4a7729-60ef-4c4d-bde5-f916168bf412",
   "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>feature</th>\n",
       "      <th>weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Daylength in decimal hours</td>\n",
       "      <td>3.4591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Billing Period Duration (Days)</td>\n",
       "      <td>-33.4487</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          feature   weight\n",
       "0      Daylength in decimal hours   3.4591\n",
       "1  Billing Period Duration (Days) -33.4487"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fitting a regressor to the training data\n",
    "\n",
    "model = LinearRegression(fit_intercept=True).fit(X_train, y_train)\n",
    "\n",
    "# Using feature weights as a proxy for feature importance\n",
    "\n",
    "weights = model.coef_\n",
    "\n",
    "feature = []\n",
    "weight = []\n",
    "\n",
    "for i, v in enumerate(weights):\n",
    "    feature.append(i)\n",
    "    weight.append(f'{v.round(4)}')\n",
    "    \n",
    "# Packaging it all up as a dataframe\n",
    "    \n",
    "df = pd.DataFrame({'feature': feature,\n",
    "                   'weight': weight}).sort_values(by='weight', ascending=False)\n",
    "\n",
    "df.feature = X_train.columns\n",
    "\n",
    "df.weight = pd.to_numeric(df.weight)\n",
    "\n",
    "df.sort_values(by='weight', ascending=False).reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "54a5737a-1614-425d-a7a6-d1ee98a648c6",
   "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": [
    "# Residuals seem reasonably close to the line\n",
    "\n",
    "y_pred = model.predict(X_test)\n",
    "\n",
    "residuals = y_test - y_pred\n",
    "\n",
    "fig = sm.graphics.qqplot(residuals, dist=stats.norm, line='45', fit=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "c82843c0-981e-4505-8651-413e1033d061",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([509.19796447, 407.39892034])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Predicted energy generated by inverter 2 for the months\n",
    "# it was malfunctioning passes the sniff test\n",
    "\n",
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "720f17ca-df5a-4c42-afd1-7df6fa1a6abc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculating RMSE\n",
    "\n",
    "rmse = sqrt(mean_squared_error(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "7b544d53-8f54-4b81-bb8c-5b9134af6ca7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "130.59865586404544"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# This suggests that the model is predicting within\n",
    "# 130 kWh for the month based on the mean that was calculated earlier\n",
    "\n",
    "rmse"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}