Random Forest Model to predict Volve production rate -checkpoint.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import packages\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "plt.style.use('seaborn-whitegrid')\n",
    "from IPython.display import display, HTML, display_html\n",
    "import seaborn as sns\n",
    "import datetime\n",
    "\n",
    "# Machine Learning Models\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.neural_network import MLPRegressor\n",
    "from sklearn import preprocessing\n",
    "from sklearn.preprocessing import StandardScaler, MinMaxScaler\n",
    "from sklearn.model_selection import TimeSeriesSplit\n",
    "\n",
    "# Train-test split\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "\n",
    "# Hyperparameter tuning\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import RandomizedSearchCV, GridSearchCV\n",
    "\n",
    "\n",
    "# Metrics\n",
    "\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.metrics import explained_variance_score\n",
    "\n",
    "# Additionals\n",
    "\n",
    "from pprint import pprint\n",
    "from math import sqrt\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DataSet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>Unnamed: 0</th>\n",
       "      <th>DATEPRD</th>\n",
       "      <th>NPD_WELL_BORE_CODE</th>\n",
       "      <th>ON_STREAM_HRS</th>\n",
       "      <th>AVG_DOWNHOLE_PRESSURE</th>\n",
       "      <th>AVG_DP_TUBING</th>\n",
       "      <th>AVG_WHP_P</th>\n",
       "      <th>AVG_WHT_P</th>\n",
       "      <th>DP_CHOKE_SIZE</th>\n",
       "      <th>BORE_OIL_VOL</th>\n",
       "      <th>BORE_GAS_VOL</th>\n",
       "      <th>BORE_WAT_VOL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2014-04-22</td>\n",
       "      <td>7405</td>\n",
       "      <td>24.0</td>\n",
       "      <td>289.421362</td>\n",
       "      <td>182.059312</td>\n",
       "      <td>107.362050</td>\n",
       "      <td>37.939251</td>\n",
       "      <td>78.935409</td>\n",
       "      <td>631.47</td>\n",
       "      <td>90439.09</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2014-04-23</td>\n",
       "      <td>7405</td>\n",
       "      <td>24.0</td>\n",
       "      <td>270.239793</td>\n",
       "      <td>171.052782</td>\n",
       "      <td>99.187011</td>\n",
       "      <td>60.756579</td>\n",
       "      <td>70.627109</td>\n",
       "      <td>1166.46</td>\n",
       "      <td>165720.39</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2014-04-24</td>\n",
       "      <td>7405</td>\n",
       "      <td>24.0</td>\n",
       "      <td>262.842748</td>\n",
       "      <td>168.241977</td>\n",
       "      <td>94.600770</td>\n",
       "      <td>63.046800</td>\n",
       "      <td>66.049151</td>\n",
       "      <td>1549.81</td>\n",
       "      <td>221707.31</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2014-04-25</td>\n",
       "      <td>7405</td>\n",
       "      <td>24.0</td>\n",
       "      <td>255.526995</td>\n",
       "      <td>165.538903</td>\n",
       "      <td>89.988092</td>\n",
       "      <td>64.547229</td>\n",
       "      <td>61.405386</td>\n",
       "      <td>1248.70</td>\n",
       "      <td>178063.52</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>2014-04-26</td>\n",
       "      <td>7405</td>\n",
       "      <td>24.0</td>\n",
       "      <td>247.199052</td>\n",
       "      <td>162.422242</td>\n",
       "      <td>84.776810</td>\n",
       "      <td>65.723694</td>\n",
       "      <td>56.147906</td>\n",
       "      <td>1345.78</td>\n",
       "      <td>192602.19</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0     DATEPRD  NPD_WELL_BORE_CODE  ON_STREAM_HRS  \\\n",
       "0           0  2014-04-22                7405           24.0   \n",
       "1           1  2014-04-23                7405           24.0   \n",
       "2           2  2014-04-24                7405           24.0   \n",
       "3           3  2014-04-25                7405           24.0   \n",
       "4           4  2014-04-26                7405           24.0   \n",
       "\n",
       "   AVG_DOWNHOLE_PRESSURE  AVG_DP_TUBING   AVG_WHP_P  AVG_WHT_P  DP_CHOKE_SIZE  \\\n",
       "0             289.421362     182.059312  107.362050  37.939251      78.935409   \n",
       "1             270.239793     171.052782   99.187011  60.756579      70.627109   \n",
       "2             262.842748     168.241977   94.600770  63.046800      66.049151   \n",
       "3             255.526995     165.538903   89.988092  64.547229      61.405386   \n",
       "4             247.199052     162.422242   84.776810  65.723694      56.147906   \n",
       "\n",
       "   BORE_OIL_VOL  BORE_GAS_VOL  BORE_WAT_VOL  \n",
       "0        631.47      90439.09           0.0  \n",
       "1       1166.46     165720.39           0.0  \n",
       "2       1549.81     221707.31           0.0  \n",
       "3       1248.70     178063.52           0.0  \n",
       "4       1345.78     192602.19           0.0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('data/final_data.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(7879, 12)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.96</td>\n",
       "      <td>0.939509</td>\n",
       "      <td>0.702695</td>\n",
       "      <td>0.888107</td>\n",
       "      <td>0.405726</td>\n",
       "      <td>0.744427</td>\n",
       "      <td>0.106970</td>\n",
       "      <td>0.106227</td>\n",
       "      <td>0.054006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.000127</td>\n",
       "      <td>0.96</td>\n",
       "      <td>0.877243</td>\n",
       "      <td>0.660213</td>\n",
       "      <td>0.820483</td>\n",
       "      <td>0.649736</td>\n",
       "      <td>0.666073</td>\n",
       "      <td>0.197620</td>\n",
       "      <td>0.194679</td>\n",
       "      <td>0.054006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.000254</td>\n",
       "      <td>0.96</td>\n",
       "      <td>0.853231</td>\n",
       "      <td>0.649364</td>\n",
       "      <td>0.782545</td>\n",
       "      <td>0.674228</td>\n",
       "      <td>0.622899</td>\n",
       "      <td>0.262577</td>\n",
       "      <td>0.260460</td>\n",
       "      <td>0.054006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.000381</td>\n",
       "      <td>0.96</td>\n",
       "      <td>0.829482</td>\n",
       "      <td>0.638931</td>\n",
       "      <td>0.744389</td>\n",
       "      <td>0.690274</td>\n",
       "      <td>0.579104</td>\n",
       "      <td>0.211555</td>\n",
       "      <td>0.209181</td>\n",
       "      <td>0.054006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.000508</td>\n",
       "      <td>0.96</td>\n",
       "      <td>0.802449</td>\n",
       "      <td>0.626901</td>\n",
       "      <td>0.701281</td>\n",
       "      <td>0.702855</td>\n",
       "      <td>0.529522</td>\n",
       "      <td>0.228005</td>\n",
       "      <td>0.226263</td>\n",
       "      <td>0.054006</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0     1         2         3         4         5         6         7  \\\n",
       "0  0.000000  0.96  0.939509  0.702695  0.888107  0.405726  0.744427  0.106970   \n",
       "1  0.000127  0.96  0.877243  0.660213  0.820483  0.649736  0.666073  0.197620   \n",
       "2  0.000254  0.96  0.853231  0.649364  0.782545  0.674228  0.622899  0.262577   \n",
       "3  0.000381  0.96  0.829482  0.638931  0.744389  0.690274  0.579104  0.211555   \n",
       "4  0.000508  0.96  0.802449  0.626901  0.701281  0.702855  0.529522  0.228005   \n",
       "\n",
       "          8         9  \n",
       "0  0.106227  0.054006  \n",
       "1  0.194679  0.054006  \n",
       "2  0.260460  0.054006  \n",
       "3  0.209181  0.054006  \n",
       "4  0.226263  0.054006  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfnew = df.drop(['DATEPRD' , 'NPD_WELL_BORE_CODE'], axis=1)\n",
    "\n",
    "scaler_all_data = MinMaxScaler()\n",
    "data_scaled = scaler_all_data.fit_transform(dfnew)\n",
    "df_data_scaled=pd.DataFrame(data_scaled)\n",
    "df_data_scaled.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "color = {'boxes': 'DarkGreen', 'whiskers': 'DarkOrange', 'medians': 'DarkBlue', 'caps': 'Gray'}\n",
    "bp_scaled_all_data = df_data_scaled.plot.box(color=color, sym='r+' ,figsize=(30,10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Oil & Gas Production Rate Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>ON_STREAM_HRS</th>\n",
       "      <th>AVG_DOWNHOLE_PRESSURE</th>\n",
       "      <th>AVG_DP_TUBING</th>\n",
       "      <th>AVG_WHP_P</th>\n",
       "      <th>AVG_WHT_P</th>\n",
       "      <th>DP_CHOKE_SIZE</th>\n",
       "      <th>BORE_OIL_VOL</th>\n",
       "      <th>BORE_GAS_VOL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>289.421362</td>\n",
       "      <td>182.059312</td>\n",
       "      <td>107.362050</td>\n",
       "      <td>37.939251</td>\n",
       "      <td>78.935409</td>\n",
       "      <td>631.47</td>\n",
       "      <td>90439.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>24.0</td>\n",
       "      <td>270.239793</td>\n",
       "      <td>171.052782</td>\n",
       "      <td>99.187011</td>\n",
       "      <td>60.756579</td>\n",
       "      <td>70.627109</td>\n",
       "      <td>1166.46</td>\n",
       "      <td>165720.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>24.0</td>\n",
       "      <td>262.842748</td>\n",
       "      <td>168.241977</td>\n",
       "      <td>94.600770</td>\n",
       "      <td>63.046800</td>\n",
       "      <td>66.049151</td>\n",
       "      <td>1549.81</td>\n",
       "      <td>221707.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>24.0</td>\n",
       "      <td>255.526995</td>\n",
       "      <td>165.538903</td>\n",
       "      <td>89.988092</td>\n",
       "      <td>64.547229</td>\n",
       "      <td>61.405386</td>\n",
       "      <td>1248.70</td>\n",
       "      <td>178063.52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>24.0</td>\n",
       "      <td>247.199052</td>\n",
       "      <td>162.422242</td>\n",
       "      <td>84.776810</td>\n",
       "      <td>65.723694</td>\n",
       "      <td>56.147906</td>\n",
       "      <td>1345.78</td>\n",
       "      <td>192602.19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   ON_STREAM_HRS  AVG_DOWNHOLE_PRESSURE  AVG_DP_TUBING   AVG_WHP_P  AVG_WHT_P  \\\n",
       "0           24.0             289.421362     182.059312  107.362050  37.939251   \n",
       "1           24.0             270.239793     171.052782   99.187011  60.756579   \n",
       "2           24.0             262.842748     168.241977   94.600770  63.046800   \n",
       "3           24.0             255.526995     165.538903   89.988092  64.547229   \n",
       "4           24.0             247.199052     162.422242   84.776810  65.723694   \n",
       "\n",
       "   DP_CHOKE_SIZE  BORE_OIL_VOL  BORE_GAS_VOL  \n",
       "0      78.935409        631.47      90439.09  \n",
       "1      70.627109       1166.46     165720.39  \n",
       "2      66.049151       1549.81     221707.31  \n",
       "3      61.405386       1248.70     178063.52  \n",
       "4      56.147906       1345.78     192602.19  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfnew= df.filter(['ON_STREAM_HRS','AVG_DOWNHOLE_PRESSURE','AVG_DP_TUBING','AVG_WHP_P','AVG_WHT_P','DP_CHOKE_SIZE','BORE_OIL_VOL', 'BORE_GAS_VOL'], axis=1)\n",
    "dfnew.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 51
    },
    "colab_type": "code",
    "id": "grvtB0MMeTjK",
    "outputId": "5fafa491-3737-4c90-80b2-575c8946fcdc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Features shape: (7879, 6)\n",
      "Target shape (7879, 2)\n"
     ]
    }
   ],
   "source": [
    "X = dfnew.filter(['DATEPRD','NPD_WELL_BORE_CODE','ON_STREAM_HRS', 'AVG_DOWNHOLE_PRESSURE', 'AVG_DP_TUBING', 'AVG_WHP_P', 'AVG_WHT_P', 'DP_CHOKE_SIZE'])\n",
    "Y = dfnew[['BORE_OIL_VOL', 'BORE_GAS_VOL']]\n",
    "print('Features shape:', X.shape)\n",
    "print('Target shape', Y.shape)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "PJYOx6aoxGc0"
   },
   "source": [
    "### Selection of data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.3, random_state=42, shuffle = True)\n",
    "\n",
    "\n",
    "Time=df.iloc[:,0]\n",
    "Time_train, Time_test = train_test_split(Time, test_size=0.3, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\DELL\\Anaconda3\\lib\\site-packages\\sklearn\\ensemble\\forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n",
      "  \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model score: 0.97\n",
      "Mean absolute error: 7779.38\n",
      "Root mean squared error: 24968.18\n",
      "R2: 0.97\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\DELL\\Anaconda3\\lib\\site-packages\\sklearn\\base.py:420: FutureWarning: The default value of multioutput (not exposed in score method) will change from 'variance_weighted' to 'uniform_average' in 0.23 to keep consistent with 'metrics.r2_score'. To specify the default value manually and avoid the warning, please either call 'metrics.r2_score' directly or make a custom scorer with 'metrics.make_scorer' (the built-in scorer 'r2' uses multioutput='uniform_average').\n",
      "  \"multioutput='uniform_average').\", FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "# Random analysis to fit Random forest \n",
    "rf = RandomForestRegressor(random_state=0).fit(X_train, y_train) #Random_state is the seed used by the random number generator\n",
    "predictions_rf = rf.predict(X_test)\n",
    "\n",
    "# Metrics\n",
    "print('Model score:',              round(rf.score(X_test, y_test),2))\n",
    "print('Mean absolute error:',      round(mean_absolute_error(y_test, predictions_rf),2))\n",
    "print('Root mean squared error:',  round(sqrt(mean_squared_error(y_test, predictions_rf)),2))\n",
    "print('R2:',                       round(r2_score(y_test, predictions_rf),2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Features importances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Variable: AVG_WHP_P                      Importance: 0.51\n",
      "Variable: AVG_WHT_P                      Importance: 0.32\n",
      "Variable: AVG_DP_TUBING                  Importance: 0.06\n",
      "Variable: ON_STREAM_HRS                  Importance: 0.04\n",
      "Variable: AVG_DOWNHOLE_PRESSURE          Importance: 0.04\n",
      "Variable: DP_CHOKE_SIZE                  Importance: 0.02\n"
     ]
    }
   ],
   "source": [
    "# List of features\n",
    "feature_list = list(X_train.columns)\n",
    "\n",
    "# Get numerical feature importances (Gini importance)\n",
    "importances = list(rf.feature_importances_)\n",
    "\n",
    "# List of tuples with variable and importance\n",
    "feature_importances = [(feature, round(importance, 2)) for feature, importance in zip(feature_list, importances)]\n",
    "\n",
    "# Sort the feature importances by most important first\n",
    "feature_importances = sorted(feature_importances, key = lambda x: x[1], reverse = True)\n",
    "\n",
    "# Print out the feature and importances \n",
    "[print('Variable: {:30} Importance: {}'.format(*pair)) for pair in feature_importances];"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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": [
    "# list of x locations for plotting\n",
    "X_testues = list(range(len(importances)))\n",
    "\n",
    "# Make a bar chart\n",
    "plt.bar(X_testues, importances, orientation = 'vertical', color = 'b', edgecolor = 'k', linewidth = 1.2)\n",
    "\n",
    "# Tick labels for x axis\n",
    "plt.xticks(X_testues, feature_list, rotation='vertical')\n",
    "\n",
    "# Axis labels and title\n",
    "plt.ylabel('Importance'); plt.xlabel('Variable'); plt.title('Variable Importances');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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": [
    "# List of features sorted from most to least important\n",
    "sorted_importances = [importance[1] for importance in feature_importances]\n",
    "sorted_features = [importance[0] for importance in feature_importances]\n",
    "\n",
    "# Cumulative importances\n",
    "cumulative_importances = np.cumsum(sorted_importances)\n",
    "\n",
    "# Make a line graph\n",
    "plt.plot(X_testues, cumulative_importances, 'b-')\n",
    "\n",
    "# Draw line at 95% of importance retained\n",
    "plt.hlines(y = 0.95, xmin=0, xmax=len(sorted_importances), color = 'r', linestyles = 'dashed')\n",
    "\n",
    "# Format x ticks and labels\n",
    "plt.xticks(X_testues, sorted_features, rotation = 'vertical')\n",
    "\n",
    "# Axis labels and title\n",
    "plt.xlabel('Variable'); plt.ylabel('Cumulative Importance'); plt.title('Cumulative Importances');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of features for 95% importance: 5\n"
     ]
    }
   ],
   "source": [
    "# Find number of features for cumulative importance of 95%\n",
    "# Add 1 because Python is zero-indexed\n",
    "print('Number of features for 95% importance:', np.where(cumulative_importances > 0.95)[0][0] + 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Important train features shape: (5515, 5)\n",
      "Important val features shape: (2364, 5)\n"
     ]
    }
   ],
   "source": [
    "# Extract the names of the most important features\n",
    "important_feature_names = [feature[0] for feature in feature_importances[0:5]]\n",
    "\n",
    "# Find the columns of the most important features\n",
    "important_indices = [feature_list.index(feature) for feature in important_feature_names]\n",
    "\n",
    "# # Create training and testing sets with only the important features\n",
    "important_train_features = X_train.iloc[:, important_indices]\n",
    "important_val_features = X_test.iloc[:, important_indices]\n",
    "\n",
    "# # Sanity check on operations\n",
    "print('Important train features shape:', important_train_features.shape)\n",
    "print('Important val features shape:', important_val_features.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training and testing in features selected \n",
    "Then, we evaluate on the **test data set** the performance using the important features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 71
    },
    "colab_type": "code",
    "id": "tSzQ1y1o1Wtd",
    "outputId": "af1fbe02-98b1-4552-ddf7-ba506b052d2d"
   },
   "outputs": [],
   "source": [
    "# Train the expanded model on only the important features\n",
    "rf.fit(important_train_features, y_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 68
    },
    "colab_type": "code",
    "id": "eJyGsucT1hI2",
    "outputId": "b87ffb17-7030-4a23-91f8-5359c98176fb"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean absolute error: 8338.13\n",
      "Root mean squared error: 25507.03\n",
      "R2: 0.97\n",
      "Explained variance score: 0.9659093146903484\n"
     ]
    }
   ],
   "source": [
    "# Make predictions on validation data\n",
    "predictions_import = rf.predict(important_val_features)\n",
    "\n",
    "# Metrics\n",
    "print('Mean absolute error:',  round(mean_absolute_error(y_test, predictions_import),2))\n",
    "print('Root mean squared error:',   round(sqrt(mean_squared_error(y_test, predictions_import)),2))\n",
    "print('R2:',                   round(r2_score(y_test, predictions_import),2))\n",
    "print('Explained variance score:', explained_variance_score(y_test, predictions_import))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 394
    },
    "colab_type": "code",
    "id": "lgBmXhq84MpI",
    "outputId": "c8eabeea-fa6f-45cd-837b-1f8481095a09"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "All features total training and testing time: 0.26 seconds.\n"
     ]
    }
   ],
   "source": [
    "# We require to import the time library for the run time evaluation\n",
    "import time\n",
    "\n",
    "# All features training and testing time\n",
    "all_features_time = []\n",
    "\n",
    "# We decide to do 10 iterations and take average for all features\n",
    "for _ in range(10):\n",
    "    start_time = time.time()\n",
    "    rf.fit(X_train, y_train)\n",
    "    all_features_predictions = rf.predict(X_test)\n",
    "    end_time = time.time()\n",
    "    all_features_time.append(end_time - start_time)\n",
    "\n",
    "all_features_time = np.mean(all_features_time)\n",
    "print('All features total training and testing time:', round(all_features_time, 2), 'seconds.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 394
    },
    "colab_type": "code",
    "id": "xmLqjwfQ4etv",
    "outputId": "9f7199a8-2186-4515-bf81-e71f020ee4a4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reduced features total training and testing time: 0.21 seconds.\n"
     ]
    }
   ],
   "source": [
    "# Time training and testing for reduced feature set\n",
    "reduced_features_time = []\n",
    "\n",
    "# We decide to do 10 iterations and take average for reduced features\n",
    "for _ in range(10):\n",
    "    start_time = time.time()\n",
    "    rf.fit(important_train_features, y_train)\n",
    "    reduced_features_predictions = rf.predict(important_val_features)\n",
    "    end_time = time.time()\n",
    "    reduced_features_time.append(end_time - start_time)\n",
    "\n",
    "reduced_features_time = np.mean(reduced_features_time)\n",
    "print('Reduced features total training and testing time:', round(reduced_features_time, 2), 'seconds.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 111
    },
    "colab_type": "code",
    "id": "zWSuqk4p42c4",
    "outputId": "68197c9a-44aa-4d31-9e1b-3f99f8e41f88"
   },
   "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>Features</th>\n",
       "      <th>R2</th>\n",
       "      <th>Mean absolute error</th>\n",
       "      <th>Root mean squared error</th>\n",
       "      <th>Run time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>all (8)</td>\n",
       "      <td>0.97</td>\n",
       "      <td>7779.38</td>\n",
       "      <td>24968.18</td>\n",
       "      <td>0.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>reduced (5)</td>\n",
       "      <td>0.97</td>\n",
       "      <td>8338.13</td>\n",
       "      <td>25507.03</td>\n",
       "      <td>0.21</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Features    R2  Mean absolute error  Root mean squared error  Run time\n",
       "0      all (8)  0.97              7779.38                 24968.18      0.26\n",
       "1  reduced (5)  0.97              8338.13                 25507.03      0.21"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Mean absolute error\n",
    "all_mean_absolute_error=mean_absolute_error(y_test, predictions_rf)\n",
    "reduced_mean_absolute_error=mean_absolute_error(y_test, predictions_import)\n",
    "\n",
    "# Root squared mean error\n",
    "all_root_mean_squared_error=sqrt(mean_squared_error(y_test, predictions_rf))\n",
    "reduced_root_mean_squared_error=sqrt(mean_squared_error(y_test, predictions_import))\n",
    "\n",
    "# R2 score\n",
    "all_r2_score=r2_score(y_test, predictions_rf)\n",
    "reduced_r2_score=r2_score(y_test, predictions_import)\n",
    "\n",
    "# Summarize the data in a Dataframe\n",
    "comparison = pd.DataFrame({'Features': ['all (8)', 'reduced (5)'], \n",
    "                           'R2': [round(all_r2_score, 2), round(reduced_r2_score, 2)],\n",
    "                           'Mean absolute error': [round(all_mean_absolute_error, 2), round(reduced_mean_absolute_error, 2)],\n",
    "                           'Root mean squared error': [round(all_root_mean_squared_error, 2), round(reduced_root_mean_squared_error, 2)],\n",
    "                           'Run time': [round(all_features_time, 2), round(reduced_features_time, 2)]})\n",
    "\n",
    "comparison[['Features','R2', 'Mean absolute error', 'Root mean squared error', 'Run time']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Examine the default Random Forest to determine parameters\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 323
    },
    "colab_type": "code",
    "id": "lGRD5bAp6azA",
    "outputId": "695c89ba-a16b-48b7-fed5-f8199f0ca3af"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parameters currently in use:\n",
      "\n",
      "{'bootstrap': True,\n",
      " 'criterion': 'mse',\n",
      " 'max_depth': None,\n",
      " 'max_features': 'auto',\n",
      " 'max_leaf_nodes': None,\n",
      " 'min_impurity_decrease': 0.0,\n",
      " 'min_impurity_split': None,\n",
      " 'min_samples_leaf': 1,\n",
      " 'min_samples_split': 2,\n",
      " 'min_weight_fraction_leaf': 0.0,\n",
      " 'n_estimators': 10,\n",
      " 'n_jobs': None,\n",
      " 'oob_score': False,\n",
      " 'random_state': 0,\n",
      " 'verbose': 0,\n",
      " 'warm_start': False}\n"
     ]
    }
   ],
   "source": [
    "# Look at parameters used by our current forest\n",
    "print('Parameters currently in use:\\n')\n",
    "pprint(rf.get_params())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "trC_FYHq-vx0"
   },
   "source": [
    "### Defining hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 442
    },
    "colab_type": "code",
    "id": "CC-8lgZj7vh3",
    "outputId": "19d7d9d5-bf95-45c4-e36c-9452c2fff3b7"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'bootstrap': [True, False],\n",
      " 'max_depth': [4, 30, None],\n",
      " 'max_features': ['auto', 'sqrt'],\n",
      " 'min_samples_leaf': [1, 2, 4],\n",
      " 'min_samples_split': [2, 3, 4, 5, 10],\n",
      " 'n_estimators': [2,\n",
      "                  107,\n",
      "                  212,\n",
      "                  317,\n",
      "                  422,\n",
      "                  527,\n",
      "                  632,\n",
      "                  738,\n",
      "                  843,\n",
      "                  948,\n",
      "                  1053,\n",
      "                  1158,\n",
      "                  1263,\n",
      "                  1369,\n",
      "                  1474,\n",
      "                  1579,\n",
      "                  1684,\n",
      "                  1789,\n",
      "                  1894,\n",
      "                  2000]}\n"
     ]
    }
   ],
   "source": [
    "# Definition of specific parameters for Random forest \n",
    "# Number of trees in random forest\n",
    "n_estimators = [int(x) for x in np.linspace(start = 2, stop = 2000, num = 20)]\n",
    "# Number of features to consider at every split\n",
    "max_features = ['auto', 'sqrt']\n",
    "# Maximum number of levels in tree\n",
    "max_depth = [int(x) for x in np.linspace(4, 30, num = 2)]\n",
    "max_depth.append(None)\n",
    "# Minimum number of samples required to split a node\n",
    "min_samples_split = [2, 3, 4, 5, 10]\n",
    "# Minimum number of samples required at each leaf node\n",
    "min_samples_leaf = [1, 2, 4]\n",
    "# Method of selecting samples for training each tree\n",
    "bootstrap = [True, False]\n",
    "\n",
    "# Create the random grid\n",
    "random_grid = {'n_estimators': n_estimators,\n",
    "               'max_features': max_features,\n",
    "               'max_depth': max_depth,\n",
    "               'min_samples_split': min_samples_split,\n",
    "               'min_samples_leaf': min_samples_leaf,\n",
    "               'bootstrap': bootstrap}\n",
    "\n",
    "pprint(random_grid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random search with cross validation\n",
    "In cross validation, the training set is split into k smaller sets. We require to follow the following steps: \n",
    "* A model is trained using $k-1$of the folds as training data (Block of the data to train).\n",
    "* The resulting model is validated on the remaining part of the data (Test data). Therefore, we can test the best parameters in this part od the data.\n",
    "* The performance measure reported by k-fold cross-validation is then the average of the values computed in the loop. \n",
    "* The method is computationally expensive. \n",
    "![alt text](https://scikit-learn.org/stable/_images/grid_search_cross_validation.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 173
    },
    "colab_type": "code",
    "id": "e6YccpHS8Z3D",
    "outputId": "aee84df2-1f86-4c27-82c4-399a2cdc149e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 15 candidates, totalling 45 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  25 tasks      | elapsed:  1.6min\n",
      "[Parallel(n_jobs=-1)]: Done  45 out of  45 | elapsed:  2.4min finished\n"
     ]
    }
   ],
   "source": [
    "rf = RandomForestRegressor(random_state = 42)\n",
    "\n",
    "# Use the random grid to search for best hyperparameters\n",
    "# Random search of parameters, using 3 fold cross validation, \n",
    "# search across different combinations.\n",
    "rf_random = RandomizedSearchCV(estimator=rf, param_distributions=random_grid, n_iter = 15, scoring='neg_mean_absolute_error',  cv = 3, verbose=2, random_state=42, n_jobs=-1, return_train_score=True)\n",
    "\n",
    "# Fit the random search model\n",
    "rf_random.fit(X_train, y_train);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "JPr64VFx_eTs"
   },
   "source": [
    "### Definition of best hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 119
    },
    "colab_type": "code",
    "id": "Mv_lx5cp8wxt",
    "outputId": "6f273b9d-d748-4aa0-c99f-6af06a478640"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n_estimators': 422,\n",
       " 'min_samples_split': 4,\n",
       " 'min_samples_leaf': 2,\n",
       " 'max_features': 'sqrt',\n",
       " 'max_depth': None,\n",
       " 'bootstrap': False}"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Obtaining the best parameters\n",
    "rf_random.best_params_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "0FmxMfGnc_Uz"
   },
   "source": [
    "### Evaluation of parameters in test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 139
    },
    "colab_type": "code",
    "id": "7OgbdW-DYzEW",
    "outputId": "b885767e-e5fb-40eb-a768-a6d8d4a7d386"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model score: 0.97\n",
      "Mean absolute error: 7275.38\n",
      "Root mean squared error: 23746.05\n",
      "R2: 0.97\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\DELL\\Anaconda3\\lib\\site-packages\\sklearn\\base.py:420: FutureWarning: The default value of multioutput (not exposed in score method) will change from 'variance_weighted' to 'uniform_average' in 0.23 to keep consistent with 'metrics.r2_score'. To specify the default value manually and avoid the warning, please either call 'metrics.r2_score' directly or make a custom scorer with 'metrics.make_scorer' (the built-in scorer 'r2' uses multioutput='uniform_average').\n",
      "  \"multioutput='uniform_average').\", FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "best_random = rf_random.best_estimator_.fit(X_train, y_train)\n",
    "predictions_best_random_test = best_random.predict(X_test)\n",
    "predictions_best_random_train = best_random.predict(X_train)\n",
    "\n",
    "print('Model score:', round(best_random.score(X_test, y_test),2))\n",
    "print('Mean absolute error:', round(mean_absolute_error(y_test, predictions_best_random_test),2))\n",
    "print('Root mean squared error:', round(sqrt(mean_squared_error(y_test, predictions_best_random_test)),2))\n",
    "print('R2:', round(r2_score(y_test, predictions_best_random_test),2))\n",
    "\n",
    "r2_rf=r2_score(y_test, predictions_best_random_test)\n",
    "Mean_absolute_error_rf=mean_absolute_error(y_test, predictions_best_random_test)\n",
    "Root_mean_squared_error_rf=sqrt(mean_squared_error(y_test, predictions_best_random_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "HMxFV2XF4P8O"
   },
   "source": [
    "### Visualization of  prediction with best hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "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>BORE_OIL_VOL</th>\n",
       "      <th>Predicted</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>162.02</td>\n",
       "      <td>159.288661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>3186.96</td>\n",
       "      <td>3231.966848</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>569.14</td>\n",
       "      <td>601.815872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>2905.92</td>\n",
       "      <td>3117.794198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>2530.04</td>\n",
       "      <td>2631.232472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2359</td>\n",
       "      <td>1310.29</td>\n",
       "      <td>1189.380575</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2360</td>\n",
       "      <td>175.96</td>\n",
       "      <td>185.637449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2361</td>\n",
       "      <td>3015.97</td>\n",
       "      <td>2980.405707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2362</td>\n",
       "      <td>303.82</td>\n",
       "      <td>367.867128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2363</td>\n",
       "      <td>1846.83</td>\n",
       "      <td>1937.247852</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2364 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      BORE_OIL_VOL    Predicted\n",
       "0           162.02   159.288661\n",
       "1          3186.96  3231.966848\n",
       "2           569.14   601.815872\n",
       "3          2905.92  3117.794198\n",
       "4          2530.04  2631.232472\n",
       "...            ...          ...\n",
       "2359       1310.29  1189.380575\n",
       "2360        175.96   185.637449\n",
       "2361       3015.97  2980.405707\n",
       "2362        303.82   367.867128\n",
       "2363       1846.83  1937.247852\n",
       "\n",
       "[2364 rows x 2 columns]"
      ]
     },
     "execution_count": 285,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_df = pd.DataFrame({'BORE_OIL_VOL': y_test, 'Predicted': predictions_best_random_test})\n",
    "result_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "dark"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from scipy import stats\n",
    "import statsmodels as sm\n",
    "\n",
    "y = y_train\n",
    "X_train_np = np.array(X_train)\n",
    "y_np = np.array(y)\n",
    "\n",
    "def abline(slope, intercept):\n",
    "    axes = plt.gca()\n",
    "    x_vals = np.array(axes.get_xlim())\n",
    "    y_vals = intercept + slope * x_vals\n",
    "    plt.plot(x_vals, y_vals, '-')\n",
    "\n",
    "#predict y values for training data\n",
    "y_hat = predictions_best_random_train \n",
    "\n",
    "#plot predicted vs actual\n",
    "plt.plot(y_hat,y_np,'o')\n",
    "plt.xlabel('Predicted')#,color='white')\n",
    "plt.ylabel('Actual')#,color='white')\n",
    "plt.title('Predicted vs. Actual: Visual Linearity Test')#,color='white')\n",
    "plt.tick_params(axis='x', colors='white')\n",
    "plt.tick_params(axis='y', colors='white')\n",
    "abline(1,0)\n",
    "plt.show()"
   ]
  }
 ],
 "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
}