predictive-maintenance-group-project-experiment.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "0b424103",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.linear_model import LinearRegression, Lasso, Ridge, BayesianRidge\n",
    "from sklearn.metrics import mean_squared_error, r2_score, mean_absolute_error, make_scorer\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.svm import SVR\n",
    "from sklearn.neighbors import KNeighborsRegressor\n",
    "from sklearn.ensemble import BaggingRegressor, AdaBoostRegressor, VotingRegressor, StackingRegressor, RandomForestRegressor\n",
    "from sklearn.model_selection import GridSearchCV, cross_val_score, train_test_split, cross_validate\n",
    "from sklearn.preprocessing import MinMaxScaler"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "355401c0",
   "metadata": {},
   "source": [
    "### Reading Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "fa698c96",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_fd001 = pd.read_csv('data/train_fd001_with_rul.csv')\n",
    "df_test_fd001 = pd.read_csv('data/test_fd001_with_rul.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2863e667",
   "metadata": {},
   "outputs": [],
   "source": [
    "index_names =  ['unit_number', 'time_cycles']\n",
    "# setting_names = ['setting_1', 'setting_2', 'setting_3']\n",
    "# drop_sensors = ['sensor_1','sensor_5','sensor_6','sensor_10','sensor_16','sensor_18','sensor_19']\n",
    "target = [\"rul\"]\n",
    "drop_labels = index_names + target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e1110286",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train = df_train_fd001.drop(drop_labels, axis = 1)\n",
    "y_train = df_train_fd001[\"rul\"]\n",
    "\n",
    "y_train = y_train.clip(upper = 125)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "4e66a8f6",
   "metadata": {},
   "outputs": [],
   "source": [
    "test = df_test_fd001.groupby('unit_number').last().reset_index()\n",
    "x_test = test.drop(drop_labels, axis = 1)\n",
    "y_test = test[\"rul\"]\n",
    "\n",
    "y_test = y_test.clip(upper = 125)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "e61a3bca",
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler()\n",
    "scaler.fit(x_train)\n",
    "\n",
    "x_train_scaled = scaler.transform(x_train)\n",
    "x_test_scaled = scaler.transform(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "082534ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_x_train_scaled = pd.DataFrame(x_train_scaled, columns = x_train.columns)\n",
    "df_x_test_scaled = pd.DataFrame(x_test_scaled, columns = x_test.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "378fd731",
   "metadata": {},
   "source": [
    "### Experiment Set 1: Comparing machine learning algorithms"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0ddb877c",
   "metadata": {},
   "source": [
    "• Goal: Identify the most suitable machine learning algorithm. <br> \n",
    "• Explore at least 3 supervised machine learning algorithms with parameter tuning to obtain the best performing model for the data you have selected."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c3986498",
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate(y_true, y_pred, label = ''):\n",
    "    # MAE\n",
    "    mae = mean_absolute_error(y_true, y_pred)\n",
    "    # RMSE\n",
    "    mse = mean_squared_error(y_true, y_pred)\n",
    "    rmse = np.sqrt(mse)\n",
    "    # R Squared\n",
    "    variance = r2_score(y_true, y_pred)\n",
    "    print('{} set - MAE: {}, RMSE: {}, R2: {}'.format(label, mae, rmse, variance))\n",
    "\n",
    "    return { 'label': label, 'mae': mae, 'rmse': rmse, 'r2': variance}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "4799b80a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 0.0, RMSE: 0.0, R2: 1.0\n",
      "test set - MAE: 19.81, RMSE: 25.99826917315843, R2: 0.5790997700619758\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 19.81,\n",
       " 'rmse': 25.99826917315843,\n",
       " 'r2': 0.5790997700619758}"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Decision tree regression\n",
    "\n",
    "dt_regressor = DecisionTreeRegressor()\n",
    "dt_regressor.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = dt_regressor.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = dt_regressor.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "14f59d4b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 17.55620783069864, RMSE: 21.468325520200416, R2: 0.7346046578123817\n",
      "test set - MAE: 16.56873644302445, RMSE: 20.830452418998043, R2: 0.7297985369402261\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 16.56873644302445,\n",
       " 'rmse': 20.830452418998043,\n",
       " 'r2': 0.7297985369402261}"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Linear regression\n",
    "\n",
    "lm = LinearRegression()\n",
    "lm.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = lm.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = lm.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "fe78b0eb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 20.683910107419383, RMSE: 24.467312702040964, R2: 0.6552776054180337\n",
      "test set - MAE: 20.550098420754153, RMSE: 24.071326877399574, R2: 0.6391802078075437\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 20.550098420754153,\n",
       " 'rmse': 24.071326877399574,\n",
       " 'r2': 0.6391802078075437}"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Lasso regression\n",
    "\n",
    "lasso = Lasso()\n",
    "lasso.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = lasso.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = lasso.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "b0932a4c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 17.556410899296328, RMSE: 21.4683732129695, R2: 0.7346034786377118\n",
      "test set - MAE: 16.56478227313438, RMSE: 20.82460206085889, R2: 0.7299502910463482\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 16.56478227313438,\n",
       " 'rmse': 20.82460206085889,\n",
       " 'r2': 0.7299502910463482}"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Ridge regression\n",
    "\n",
    "ridge = Ridge()\n",
    "ridge.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = ridge.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = ridge.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f0c31b25",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 17.52291774871658, RMSE: 21.55202918033055, R2: 0.7325311036271246\n",
      "test set - MAE: 16.148181298376162, RMSE: 20.389718132213673, R2: 0.7411115141746607\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 16.148181298376162,\n",
       " 'rmse': 20.389718132213673,\n",
       " 'r2': 0.7411115141746607}"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### SVM regression\n",
    "\n",
    "regressor = SVR(kernel = 'linear')\n",
    "regressor.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = regressor.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = regressor.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "33ff54e8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 11.71491444912995, RMSE: 16.684766401932727, R2: 0.839698637327521\n",
      "test set - MAE: 14.468000000000002, RMSE: 20.170235496889966, R2: 0.7466550633847437\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 14.468000000000002,\n",
       " 'rmse': 20.170235496889966,\n",
       " 'r2': 0.7466550633847437}"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### KNN regression\n",
    "\n",
    "neigh = KNeighborsRegressor()\n",
    "neigh.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = neigh.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = neigh.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d3d09a20",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 4.940583103097281, RMSE: 6.841757007957831, R2: 0.9730454401578147\n",
      "test set - MAE: 11.873199999999999, RMSE: 16.93273982555688, R2: 0.8214562048238725\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 11.873199999999999,\n",
       " 'rmse': 16.93273982555688,\n",
       " 'r2': 0.8214562048238725}"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Random forest regression\n",
    "\n",
    "regr = RandomForestRegressor()\n",
    "regr.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = regr.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = regr.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "312fd0ed",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 17.556325024506652, RMSE: 21.468346882919988, R2: 0.7346041296324743\n",
      "test set - MAE: 16.56599158648626, RMSE: 20.82644714688442, R2: 0.7299024354363169\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 16.56599158648626,\n",
       " 'rmse': 20.82644714688442,\n",
       " 'r2': 0.7299024354363169}"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Bayesian ridge regression\n",
    "\n",
    "clf = BayesianRidge()\n",
    "clf.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = clf.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = clf.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2e906ab9",
   "metadata": {},
   "source": [
    "### Find Best Parameter For Selected Algorithms\n",
    "Linear regression will be skipped for finding best parameters as there ar not parameters we can fine tuning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "69deabca",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=5, estimator=KNeighborsRegressor(),\n",
       "             param_grid={&#x27;n_neighbors&#x27;: [3, 5, 7], &#x27;p&#x27;: [1, 3],\n",
       "                         &#x27;weights&#x27;: [&#x27;uniform&#x27;, &#x27;distance&#x27;]})</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=5, estimator=KNeighborsRegressor(),\n",
       "             param_grid={&#x27;n_neighbors&#x27;: [3, 5, 7], &#x27;p&#x27;: [1, 3],\n",
       "                         &#x27;weights&#x27;: [&#x27;uniform&#x27;, &#x27;distance&#x27;]})</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: KNeighborsRegressor</label><div class=\"sk-toggleable__content\"><pre>KNeighborsRegressor()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">KNeighborsRegressor</label><div class=\"sk-toggleable__content\"><pre>KNeighborsRegressor()</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(cv=5, estimator=KNeighborsRegressor(),\n",
       "             param_grid={'n_neighbors': [3, 5, 7], 'p': [1, 3],\n",
       "                         'weights': ['uniform', 'distance']})"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### KNN Regressor\n",
    "\n",
    "param_grid = {\n",
    "    'n_neighbors': [3, 5, 7],\n",
    "    'weights': ['uniform', 'distance'],\n",
    "    'p': [1, 3]\n",
    "}\n",
    "\n",
    "knn = KNeighborsRegressor()\n",
    "\n",
    "grid_search = GridSearchCV(estimator = knn, param_grid = param_grid, cv = 5)\n",
    "grid_search.fit(x_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "c1ed913e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters: {'n_neighbors': 7, 'p': 1, 'weights': 'distance'}\n"
     ]
    }
   ],
   "source": [
    "best_params = grid_search.best_params_\n",
    "\n",
    "print(\"Best parameters:\", best_params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "77825dc4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-2 {color: black;background-color: white;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=5, estimator=DecisionTreeRegressor(),\n",
       "             param_grid={&#x27;max_depth&#x27;: [None, 5, 10],\n",
       "                         &#x27;max_features&#x27;: [1.0, None],\n",
       "                         &#x27;min_samples_leaf&#x27;: [1, 2, 4],\n",
       "                         &#x27;min_samples_split&#x27;: [2, 5, 10]})</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" ><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=5, estimator=DecisionTreeRegressor(),\n",
       "             param_grid={&#x27;max_depth&#x27;: [None, 5, 10],\n",
       "                         &#x27;max_features&#x27;: [1.0, None],\n",
       "                         &#x27;min_samples_leaf&#x27;: [1, 2, 4],\n",
       "                         &#x27;min_samples_split&#x27;: [2, 5, 10]})</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" ><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: DecisionTreeRegressor</label><div class=\"sk-toggleable__content\"><pre>DecisionTreeRegressor()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-6\" type=\"checkbox\" ><label for=\"sk-estimator-id-6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">DecisionTreeRegressor</label><div class=\"sk-toggleable__content\"><pre>DecisionTreeRegressor()</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(cv=5, estimator=DecisionTreeRegressor(),\n",
       "             param_grid={'max_depth': [None, 5, 10],\n",
       "                         'max_features': [1.0, None],\n",
       "                         'min_samples_leaf': [1, 2, 4],\n",
       "                         'min_samples_split': [2, 5, 10]})"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Decision Tree Regressor\n",
    "\n",
    "param_grid = {\n",
    "    'max_depth': [None, 5, 10],\n",
    "    'min_samples_split': [2, 5, 10],\n",
    "    'min_samples_leaf': [1, 2, 4],\n",
    "    'max_features': [1.0, None]\n",
    "}\n",
    "\n",
    "tree = DecisionTreeRegressor()\n",
    "\n",
    "grid_search = GridSearchCV(estimator = tree, param_grid = param_grid, cv = 5)\n",
    "grid_search.fit(x_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "95ba8d55",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters: {'max_depth': 10, 'max_features': None, 'min_samples_leaf': 4, 'min_samples_split': 10}\n"
     ]
    }
   ],
   "source": [
    "best_params = grid_search.best_params_\n",
    "\n",
    "print(\"Best parameters:\", best_params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "65350523",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-3 {color: black;background-color: white;}#sk-container-id-3 pre{padding: 0;}#sk-container-id-3 div.sk-toggleable {background-color: white;}#sk-container-id-3 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-3 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-3 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-3 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-3 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-3 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-3 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-3 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-3 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-3 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-3 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-3 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-3 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-3 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-3 div.sk-item {position: relative;z-index: 1;}#sk-container-id-3 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-3 div.sk-item::before, #sk-container-id-3 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-3 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-3 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-3 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-3 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-3 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-3 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-3 div.sk-label-container {text-align: center;}#sk-container-id-3 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-3 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-3\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=5, estimator=BayesianRidge(),\n",
       "             param_grid={&#x27;alpha_1&#x27;: [1e-06, 1e-05, 0.0001],\n",
       "                         &#x27;alpha_2&#x27;: [1e-06, 1e-05, 0.0001],\n",
       "                         &#x27;lambda_1&#x27;: [1e-06, 1e-05, 0.0001],\n",
       "                         &#x27;lambda_2&#x27;: [1e-06, 1e-05, 0.0001]})</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-7\" type=\"checkbox\" ><label for=\"sk-estimator-id-7\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=5, estimator=BayesianRidge(),\n",
       "             param_grid={&#x27;alpha_1&#x27;: [1e-06, 1e-05, 0.0001],\n",
       "                         &#x27;alpha_2&#x27;: [1e-06, 1e-05, 0.0001],\n",
       "                         &#x27;lambda_1&#x27;: [1e-06, 1e-05, 0.0001],\n",
       "                         &#x27;lambda_2&#x27;: [1e-06, 1e-05, 0.0001]})</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-8\" type=\"checkbox\" ><label for=\"sk-estimator-id-8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: BayesianRidge</label><div class=\"sk-toggleable__content\"><pre>BayesianRidge()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-9\" type=\"checkbox\" ><label for=\"sk-estimator-id-9\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">BayesianRidge</label><div class=\"sk-toggleable__content\"><pre>BayesianRidge()</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(cv=5, estimator=BayesianRidge(),\n",
       "             param_grid={'alpha_1': [1e-06, 1e-05, 0.0001],\n",
       "                         'alpha_2': [1e-06, 1e-05, 0.0001],\n",
       "                         'lambda_1': [1e-06, 1e-05, 0.0001],\n",
       "                         'lambda_2': [1e-06, 1e-05, 0.0001]})"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Bayesian ridge regression\n",
    "\n",
    "param_grid = {\n",
    "    'alpha_1': [1e-6, 1e-5, 1e-4],\n",
    "    'alpha_2': [1e-6, 1e-5, 1e-4],\n",
    "    'lambda_1': [1e-6, 1e-5, 1e-4],\n",
    "    'lambda_2': [1e-6, 1e-5, 1e-4],\n",
    "}\n",
    "\n",
    "bayesian_ridge = BayesianRidge()\n",
    "\n",
    "grid_search = GridSearchCV(estimator = bayesian_ridge, param_grid = param_grid, cv = 5)\n",
    "grid_search.fit(x_train_scaled, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "35480b88",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best parameters: {'alpha_1': 1e-06, 'alpha_2': 0.0001, 'lambda_1': 0.0001, 'lambda_2': 1e-06}\n"
     ]
    }
   ],
   "source": [
    "best_params = grid_search.best_params_\n",
    "\n",
    "print(\"Best parameters:\", best_params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "e654c4b3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 12.136687509088265, RMSE: 16.97431679289148, R2: 0.8340865632994863\n",
      "test set - MAE: 13.595714285714287, RMSE: 18.369823302450822, R2: 0.7898641026340808\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 13.595714285714287,\n",
       " 'rmse': 18.369823302450822,\n",
       " 'r2': 0.7898641026340808}"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "neigh = KNeighborsRegressor(n_neighbors = 7, p = 1, weights = 'uniform')\n",
    "# scores = cross_validate(neigh, x_train_scaled, y_train,  scoring = ('r2', 'neg_mean_absolute_error', 'neg_root_mean_squared_error'), cv = 5)\n",
    "# mae = np.mean(-scores['test_neg_mean_absolute_error'])\n",
    "# rmse = np.mean(-scores['test_neg_root_mean_squared_error'])\n",
    "# r2 = np.mean(scores['test_r2'])\n",
    "\n",
    "# print('train set - MAE: {}, RMSE: {}, R2: {}'.format(mae, rmse, r2))\n",
    "\n",
    "neigh.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = neigh.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = neigh.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "7510d4a1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 11.79795306648188, RMSE: 16.779325930854867, R2: 0.8378764991701514\n",
      "test set - MAE: 14.472401108423746, RMSE: 19.39517427688668, R2: 0.7657510440738063\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 14.472401108423746,\n",
       " 'rmse': 19.39517427688668,\n",
       " 'r2': 0.7657510440738063}"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt_regressor = DecisionTreeRegressor(max_depth = 10, max_features = None, min_samples_leaf = 4, min_samples_split = 10)\n",
    "# scores = cross_validate(dt_regressor, x_train_scaled, y_train,  scoring = ('r2', 'neg_mean_absolute_error', 'neg_root_mean_squared_error'), cv = 5)\n",
    "# mae = np.mean(-scores['test_neg_mean_absolute_error'])\n",
    "# rmse = np.mean(-scores['test_neg_root_mean_squared_error'])\n",
    "# r2 = np.mean(scores['test_r2'])\n",
    "\n",
    "# print('train set - MAE: {}, RMSE: {}, R2: {}'.format(mae, rmse, r2))\n",
    "\n",
    "dt_regressor.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = dt_regressor.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = dt_regressor.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "1226a55b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 17.556325026265185, RMSE: 21.468346883397693, R2: 0.7346041296206633\n",
      "test set - MAE: 16.56599155802789, RMSE: 20.826447104130434, R2: 0.7299024365452673\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'label': 'test',\n",
       " 'mae': 16.56599155802789,\n",
       " 'rmse': 20.826447104130434,\n",
       " 'r2': 0.7299024365452673}"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = BayesianRidge(alpha_1 = 1e-06, alpha_2 = 0.0001, lambda_1 = 0.0001, lambda_2 = 1e-06)\n",
    "clf.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = clf.predict(x_train_scaled)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = clf.predict(x_test_scaled)\n",
    "evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4dd5418",
   "metadata": {},
   "source": [
    "### Experiment Set 2: Selecting features"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e1ef6dd6",
   "metadata": {},
   "source": [
    "• Goal: Find the best set of features. <br>\n",
    "• Explore at least 3 feature selection methods to identify the most relevant features for the problem."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab638644",
   "metadata": {},
   "source": [
    "#### Correlation Coefficient : Winner"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "aee666cf",
   "metadata": {},
   "outputs": [],
   "source": [
    "names = x_train.columns \n",
    "x_train_scaled_pd = pd.DataFrame(x_train_scaled, columns = names)\n",
    "\n",
    "# Place the x_train and y_train data frames side by side\n",
    "x_train_scaled_pd_con = pd.concat([x_train_scaled_pd, y_train], axis = 1)\n",
    "\n",
    "# Generate correlation matrix\n",
    "x_train_scaled_cor = x_train_scaled_pd_con.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "beae2781",
   "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>setting_1</th>\n",
       "      <th>setting_2</th>\n",
       "      <th>setting_3</th>\n",
       "      <th>sensor_1</th>\n",
       "      <th>sensor_2</th>\n",
       "      <th>sensor_3</th>\n",
       "      <th>sensor_4</th>\n",
       "      <th>sensor_5</th>\n",
       "      <th>sensor_6</th>\n",
       "      <th>sensor_7</th>\n",
       "      <th>...</th>\n",
       "      <th>sensor_13</th>\n",
       "      <th>sensor_14</th>\n",
       "      <th>sensor_15</th>\n",
       "      <th>sensor_16</th>\n",
       "      <th>sensor_17</th>\n",
       "      <th>sensor_18</th>\n",
       "      <th>sensor_19</th>\n",
       "      <th>sensor_20</th>\n",
       "      <th>sensor_21</th>\n",
       "      <th>rul</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>setting_1</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.011660</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.009030</td>\n",
       "      <td>-0.005651</td>\n",
       "      <td>0.009544</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.001340</td>\n",
       "      <td>-0.009437</td>\n",
       "      <td>...</td>\n",
       "      <td>0.002318</td>\n",
       "      <td>-0.004469</td>\n",
       "      <td>0.007652</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.002599</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.005713</td>\n",
       "      <td>-0.014559</td>\n",
       "      <td>-0.005556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>setting_2</th>\n",
       "      <td>0.011660</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.007266</td>\n",
       "      <td>0.009068</td>\n",
       "      <td>0.014673</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.014355</td>\n",
       "      <td>-0.016678</td>\n",
       "      <td>...</td>\n",
       "      <td>0.018156</td>\n",
       "      <td>-0.006310</td>\n",
       "      <td>0.014156</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.012280</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.010554</td>\n",
       "      <td>-0.007846</td>\n",
       "      <td>-0.007091</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>setting_3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_2</th>\n",
       "      <td>0.009030</td>\n",
       "      <td>0.007266</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.602610</td>\n",
       "      <td>0.714949</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.132242</td>\n",
       "      <td>-0.702136</td>\n",
       "      <td>...</td>\n",
       "      <td>0.661792</td>\n",
       "      <td>0.179752</td>\n",
       "      <td>0.675975</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.629886</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.661841</td>\n",
       "      <td>-0.668050</td>\n",
       "      <td>-0.678458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_3</th>\n",
       "      <td>-0.005651</td>\n",
       "      <td>0.009068</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.602610</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.678413</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.116039</td>\n",
       "      <td>-0.664595</td>\n",
       "      <td>...</td>\n",
       "      <td>0.600963</td>\n",
       "      <td>0.237137</td>\n",
       "      <td>0.639921</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.600017</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.625941</td>\n",
       "      <td>-0.633901</td>\n",
       "      <td>-0.655030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_4</th>\n",
       "      <td>0.009544</td>\n",
       "      <td>0.014673</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.714949</td>\n",
       "      <td>0.678413</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.150480</td>\n",
       "      <td>-0.793130</td>\n",
       "      <td>...</td>\n",
       "      <td>0.745158</td>\n",
       "      <td>0.190748</td>\n",
       "      <td>0.758459</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.703499</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.748067</td>\n",
       "      <td>-0.745193</td>\n",
       "      <td>-0.757157</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_5</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_6</th>\n",
       "      <td>-0.001340</td>\n",
       "      <td>0.014355</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.132242</td>\n",
       "      <td>0.116039</td>\n",
       "      <td>0.150480</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.155720</td>\n",
       "      <td>...</td>\n",
       "      <td>0.158276</td>\n",
       "      <td>-0.002112</td>\n",
       "      <td>0.149042</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.130810</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.141419</td>\n",
       "      <td>-0.137419</td>\n",
       "      <td>-0.108289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_7</th>\n",
       "      <td>-0.009437</td>\n",
       "      <td>-0.016678</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.702136</td>\n",
       "      <td>-0.664595</td>\n",
       "      <td>-0.793130</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.155720</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.764611</td>\n",
       "      <td>-0.110053</td>\n",
       "      <td>-0.747051</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.692893</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.736163</td>\n",
       "      <td>0.737447</td>\n",
       "      <td>0.733021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_8</th>\n",
       "      <td>-0.000427</td>\n",
       "      <td>0.013285</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.662325</td>\n",
       "      <td>0.602481</td>\n",
       "      <td>0.746852</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.152161</td>\n",
       "      <td>-0.767132</td>\n",
       "      <td>...</td>\n",
       "      <td>0.826084</td>\n",
       "      <td>-0.144787</td>\n",
       "      <td>0.700949</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.627785</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.687030</td>\n",
       "      <td>-0.688840</td>\n",
       "      <td>-0.624568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_9</th>\n",
       "      <td>-0.004279</td>\n",
       "      <td>-0.005425</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.273764</td>\n",
       "      <td>0.322964</td>\n",
       "      <td>0.297429</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.019347</td>\n",
       "      <td>-0.217835</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.034763</td>\n",
       "      <td>0.963157</td>\n",
       "      <td>0.293753</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.337110</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.285280</td>\n",
       "      <td>-0.292795</td>\n",
       "      <td>-0.462151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_10</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_11</th>\n",
       "      <td>0.011658</td>\n",
       "      <td>0.011690</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.740020</td>\n",
       "      <td>0.695900</td>\n",
       "      <td>0.830136</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.160014</td>\n",
       "      <td>-0.822805</td>\n",
       "      <td>...</td>\n",
       "      <td>0.780761</td>\n",
       "      <td>0.163408</td>\n",
       "      <td>0.780913</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.722296</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.771510</td>\n",
       "      <td>-0.772554</td>\n",
       "      <td>-0.775230</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_12</th>\n",
       "      <td>-0.001491</td>\n",
       "      <td>-0.010923</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.724578</td>\n",
       "      <td>-0.680307</td>\n",
       "      <td>-0.815591</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.155884</td>\n",
       "      <td>0.812713</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.788441</td>\n",
       "      <td>-0.098141</td>\n",
       "      <td>-0.766052</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.703485</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.751943</td>\n",
       "      <td>0.756263</td>\n",
       "      <td>0.748870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_13</th>\n",
       "      <td>0.002318</td>\n",
       "      <td>0.018156</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.661792</td>\n",
       "      <td>0.600963</td>\n",
       "      <td>0.745158</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.158276</td>\n",
       "      <td>-0.764611</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.147036</td>\n",
       "      <td>0.697662</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.627410</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.686172</td>\n",
       "      <td>-0.688948</td>\n",
       "      <td>-0.624034</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_14</th>\n",
       "      <td>-0.004469</td>\n",
       "      <td>-0.006310</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.179752</td>\n",
       "      <td>0.237137</td>\n",
       "      <td>0.190748</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.002112</td>\n",
       "      <td>-0.110053</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.147036</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.193761</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.246621</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.187364</td>\n",
       "      <td>-0.192415</td>\n",
       "      <td>-0.369753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_15</th>\n",
       "      <td>0.007652</td>\n",
       "      <td>0.014156</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.675975</td>\n",
       "      <td>0.639921</td>\n",
       "      <td>0.758459</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.149042</td>\n",
       "      <td>-0.747051</td>\n",
       "      <td>...</td>\n",
       "      <td>0.697662</td>\n",
       "      <td>0.193761</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.669205</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.707671</td>\n",
       "      <td>-0.704043</td>\n",
       "      <td>-0.720858</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_16</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_17</th>\n",
       "      <td>0.002599</td>\n",
       "      <td>0.012280</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.629886</td>\n",
       "      <td>0.600017</td>\n",
       "      <td>0.703499</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.130810</td>\n",
       "      <td>-0.692893</td>\n",
       "      <td>...</td>\n",
       "      <td>0.627410</td>\n",
       "      <td>0.246621</td>\n",
       "      <td>0.669205</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.651021</td>\n",
       "      <td>-0.658710</td>\n",
       "      <td>-0.680829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_18</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_19</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_20</th>\n",
       "      <td>-0.005713</td>\n",
       "      <td>-0.010554</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.661841</td>\n",
       "      <td>-0.625941</td>\n",
       "      <td>-0.748067</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.141419</td>\n",
       "      <td>0.736163</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.686172</td>\n",
       "      <td>-0.187364</td>\n",
       "      <td>-0.707671</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.651021</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.693480</td>\n",
       "      <td>0.704626</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensor_21</th>\n",
       "      <td>-0.014559</td>\n",
       "      <td>-0.007846</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.668050</td>\n",
       "      <td>-0.633901</td>\n",
       "      <td>-0.745193</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.137419</td>\n",
       "      <td>0.737447</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.688948</td>\n",
       "      <td>-0.192415</td>\n",
       "      <td>-0.704043</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.658710</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.693480</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.707334</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rul</th>\n",
       "      <td>-0.005556</td>\n",
       "      <td>-0.007091</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.678458</td>\n",
       "      <td>-0.655030</td>\n",
       "      <td>-0.757157</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.108289</td>\n",
       "      <td>0.733021</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.624034</td>\n",
       "      <td>-0.369753</td>\n",
       "      <td>-0.720858</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.680829</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.704626</td>\n",
       "      <td>0.707334</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>25 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           setting_1  setting_2  setting_3  sensor_1  sensor_2  sensor_3  \\\n",
       "setting_1   1.000000   0.011660        NaN       NaN  0.009030 -0.005651   \n",
       "setting_2   0.011660   1.000000        NaN       NaN  0.007266  0.009068   \n",
       "setting_3        NaN        NaN        NaN       NaN       NaN       NaN   \n",
       "sensor_1         NaN        NaN        NaN       NaN       NaN       NaN   \n",
       "sensor_2    0.009030   0.007266        NaN       NaN  1.000000  0.602610   \n",
       "sensor_3   -0.005651   0.009068        NaN       NaN  0.602610  1.000000   \n",
       "sensor_4    0.009544   0.014673        NaN       NaN  0.714949  0.678413   \n",
       "sensor_5         NaN        NaN        NaN       NaN       NaN       NaN   \n",
       "sensor_6   -0.001340   0.014355        NaN       NaN  0.132242  0.116039   \n",
       "sensor_7   -0.009437  -0.016678        NaN       NaN -0.702136 -0.664595   \n",
       "sensor_8   -0.000427   0.013285        NaN       NaN  0.662325  0.602481   \n",
       "sensor_9   -0.004279  -0.005425        NaN       NaN  0.273764  0.322964   \n",
       "sensor_10        NaN        NaN        NaN       NaN       NaN       NaN   \n",
       "sensor_11   0.011658   0.011690        NaN       NaN  0.740020  0.695900   \n",
       "sensor_12  -0.001491  -0.010923        NaN       NaN -0.724578 -0.680307   \n",
       "sensor_13   0.002318   0.018156        NaN       NaN  0.661792  0.600963   \n",
       "sensor_14  -0.004469  -0.006310        NaN       NaN  0.179752  0.237137   \n",
       "sensor_15   0.007652   0.014156        NaN       NaN  0.675975  0.639921   \n",
       "sensor_16        NaN        NaN        NaN       NaN       NaN       NaN   \n",
       "sensor_17   0.002599   0.012280        NaN       NaN  0.629886  0.600017   \n",
       "sensor_18        NaN        NaN        NaN       NaN       NaN       NaN   \n",
       "sensor_19        NaN        NaN        NaN       NaN       NaN       NaN   \n",
       "sensor_20  -0.005713  -0.010554        NaN       NaN -0.661841 -0.625941   \n",
       "sensor_21  -0.014559  -0.007846        NaN       NaN -0.668050 -0.633901   \n",
       "rul        -0.005556  -0.007091        NaN       NaN -0.678458 -0.655030   \n",
       "\n",
       "           sensor_4  sensor_5  sensor_6  sensor_7  ...  sensor_13  sensor_14  \\\n",
       "setting_1  0.009544       NaN -0.001340 -0.009437  ...   0.002318  -0.004469   \n",
       "setting_2  0.014673       NaN  0.014355 -0.016678  ...   0.018156  -0.006310   \n",
       "setting_3       NaN       NaN       NaN       NaN  ...        NaN        NaN   \n",
       "sensor_1        NaN       NaN       NaN       NaN  ...        NaN        NaN   \n",
       "sensor_2   0.714949       NaN  0.132242 -0.702136  ...   0.661792   0.179752   \n",
       "sensor_3   0.678413       NaN  0.116039 -0.664595  ...   0.600963   0.237137   \n",
       "sensor_4   1.000000       NaN  0.150480 -0.793130  ...   0.745158   0.190748   \n",
       "sensor_5        NaN       NaN       NaN       NaN  ...        NaN        NaN   \n",
       "sensor_6   0.150480       NaN  1.000000 -0.155720  ...   0.158276  -0.002112   \n",
       "sensor_7  -0.793130       NaN -0.155720  1.000000  ...  -0.764611  -0.110053   \n",
       "sensor_8   0.746852       NaN  0.152161 -0.767132  ...   0.826084  -0.144787   \n",
       "sensor_9   0.297429       NaN  0.019347 -0.217835  ...  -0.034763   0.963157   \n",
       "sensor_10       NaN       NaN       NaN       NaN  ...        NaN        NaN   \n",
       "sensor_11  0.830136       NaN  0.160014 -0.822805  ...   0.780761   0.163408   \n",
       "sensor_12 -0.815591       NaN -0.155884  0.812713  ...  -0.788441  -0.098141   \n",
       "sensor_13  0.745158       NaN  0.158276 -0.764611  ...   1.000000  -0.147036   \n",
       "sensor_14  0.190748       NaN -0.002112 -0.110053  ...  -0.147036   1.000000   \n",
       "sensor_15  0.758459       NaN  0.149042 -0.747051  ...   0.697662   0.193761   \n",
       "sensor_16       NaN       NaN       NaN       NaN  ...        NaN        NaN   \n",
       "sensor_17  0.703499       NaN  0.130810 -0.692893  ...   0.627410   0.246621   \n",
       "sensor_18       NaN       NaN       NaN       NaN  ...        NaN        NaN   \n",
       "sensor_19       NaN       NaN       NaN       NaN  ...        NaN        NaN   \n",
       "sensor_20 -0.748067       NaN -0.141419  0.736163  ...  -0.686172  -0.187364   \n",
       "sensor_21 -0.745193       NaN -0.137419  0.737447  ...  -0.688948  -0.192415   \n",
       "rul       -0.757157       NaN -0.108289  0.733021  ...  -0.624034  -0.369753   \n",
       "\n",
       "           sensor_15  sensor_16  sensor_17  sensor_18  sensor_19  sensor_20  \\\n",
       "setting_1   0.007652        NaN   0.002599        NaN        NaN  -0.005713   \n",
       "setting_2   0.014156        NaN   0.012280        NaN        NaN  -0.010554   \n",
       "setting_3        NaN        NaN        NaN        NaN        NaN        NaN   \n",
       "sensor_1         NaN        NaN        NaN        NaN        NaN        NaN   \n",
       "sensor_2    0.675975        NaN   0.629886        NaN        NaN  -0.661841   \n",
       "sensor_3    0.639921        NaN   0.600017        NaN        NaN  -0.625941   \n",
       "sensor_4    0.758459        NaN   0.703499        NaN        NaN  -0.748067   \n",
       "sensor_5         NaN        NaN        NaN        NaN        NaN        NaN   \n",
       "sensor_6    0.149042        NaN   0.130810        NaN        NaN  -0.141419   \n",
       "sensor_7   -0.747051        NaN  -0.692893        NaN        NaN   0.736163   \n",
       "sensor_8    0.700949        NaN   0.627785        NaN        NaN  -0.687030   \n",
       "sensor_9    0.293753        NaN   0.337110        NaN        NaN  -0.285280   \n",
       "sensor_10        NaN        NaN        NaN        NaN        NaN        NaN   \n",
       "sensor_11   0.780913        NaN   0.722296        NaN        NaN  -0.771510   \n",
       "sensor_12  -0.766052        NaN  -0.703485        NaN        NaN   0.751943   \n",
       "sensor_13   0.697662        NaN   0.627410        NaN        NaN  -0.686172   \n",
       "sensor_14   0.193761        NaN   0.246621        NaN        NaN  -0.187364   \n",
       "sensor_15   1.000000        NaN   0.669205        NaN        NaN  -0.707671   \n",
       "sensor_16        NaN        NaN        NaN        NaN        NaN        NaN   \n",
       "sensor_17   0.669205        NaN   1.000000        NaN        NaN  -0.651021   \n",
       "sensor_18        NaN        NaN        NaN        NaN        NaN        NaN   \n",
       "sensor_19        NaN        NaN        NaN        NaN        NaN        NaN   \n",
       "sensor_20  -0.707671        NaN  -0.651021        NaN        NaN   1.000000   \n",
       "sensor_21  -0.704043        NaN  -0.658710        NaN        NaN   0.693480   \n",
       "rul        -0.720858        NaN  -0.680829        NaN        NaN   0.704626   \n",
       "\n",
       "           sensor_21       rul  \n",
       "setting_1  -0.014559 -0.005556  \n",
       "setting_2  -0.007846 -0.007091  \n",
       "setting_3        NaN       NaN  \n",
       "sensor_1         NaN       NaN  \n",
       "sensor_2   -0.668050 -0.678458  \n",
       "sensor_3   -0.633901 -0.655030  \n",
       "sensor_4   -0.745193 -0.757157  \n",
       "sensor_5         NaN       NaN  \n",
       "sensor_6   -0.137419 -0.108289  \n",
       "sensor_7    0.737447  0.733021  \n",
       "sensor_8   -0.688840 -0.624568  \n",
       "sensor_9   -0.292795 -0.462151  \n",
       "sensor_10        NaN       NaN  \n",
       "sensor_11  -0.772554 -0.775230  \n",
       "sensor_12   0.756263  0.748870  \n",
       "sensor_13  -0.688948 -0.624034  \n",
       "sensor_14  -0.192415 -0.369753  \n",
       "sensor_15  -0.704043 -0.720858  \n",
       "sensor_16        NaN       NaN  \n",
       "sensor_17  -0.658710 -0.680829  \n",
       "sensor_18        NaN       NaN  \n",
       "sensor_19        NaN       NaN  \n",
       "sensor_20   0.693480  0.704626  \n",
       "sensor_21   1.000000  0.707334  \n",
       "rul         0.707334  1.000000  \n",
       "\n",
       "[25 rows x 25 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Print correlation matrix\n",
    "x_train_scaled_cor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "d9f3f1a7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2500x2500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Set plot size\n",
    "plt.figure(figsize=(25,25))\n",
    "\n",
    "# Generate the heatmap\n",
    "sns.heatmap(\n",
    "    x_train_scaled_cor,\n",
    "    vmin = -1, vmax = 1, center = 0,\n",
    "    cmap = sns.diverging_palette(20, 220, n = 200),\n",
    "    square = True,\n",
    "    annot = True\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "842862db",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Select features above a correlation threshold to target\n",
    "\n",
    "# Correlation with target\n",
    "# Apply abs() to get the absolute value so no need to deal with negative correlations\n",
    "x_train_scaled_cor_target = abs(x_train_scaled_cor['rul'])\n",
    "\n",
    "# Selecting highly correlated features\n",
    "# Say we set the correlation threshold to 0.2\n",
    "#relevant_features = cor_target[cor_target > 0.7]\n",
    "#relevant_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "faacd1e1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "setting_1    0.005556\n",
       "setting_2    0.007091\n",
       "sensor_2     0.678458\n",
       "sensor_3     0.655030\n",
       "sensor_4     0.757157\n",
       "sensor_6     0.108289\n",
       "sensor_7     0.733021\n",
       "sensor_8     0.624568\n",
       "sensor_9     0.462151\n",
       "sensor_11    0.775230\n",
       "sensor_12    0.748870\n",
       "sensor_13    0.624034\n",
       "sensor_14    0.369753\n",
       "sensor_15    0.720858\n",
       "sensor_17    0.680829\n",
       "sensor_20    0.704626\n",
       "sensor_21    0.707334\n",
       "rul          1.000000\n",
       "Name: rul, dtype: float64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train_scaled_correlated_features = x_train_scaled_cor_target[x_train_scaled_cor_target > 0] # Change to 0.1, 0.2, 0.5, 0.7\n",
    "x_train_scaled_correlated_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "9c2717c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "x_train_correlated_features_1 = [column for column in x_train_scaled_pd.columns if abs(x_train_scaled_cor_target[column]> 0)] \n",
    "x_train_correlated_features_2 = [column for column in x_train_scaled_pd.columns if abs(x_train_scaled_cor_target[column]> 0.1)] \n",
    "x_train_correlated_features_3 = [column for column in x_train_scaled_pd.columns if abs(x_train_scaled_cor_target[column]> 0.2)] \n",
    "x_train_correlated_features_4 = [column for column in x_train_scaled_pd.columns if abs(x_train_scaled_cor_target[column]> 0.5)] \n",
    "x_train_correlated_features_5 = [column for column in x_train_scaled_pd.columns if abs(x_train_scaled_cor_target[column]> 0.7)] \n",
    "\n",
    "# Change to 0.1, 0.2, 0.5, 0.7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "61457ee9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['setting_1',\n",
       " 'setting_2',\n",
       " 'sensor_2',\n",
       " 'sensor_3',\n",
       " 'sensor_4',\n",
       " 'sensor_6',\n",
       " 'sensor_7',\n",
       " 'sensor_8',\n",
       " 'sensor_9',\n",
       " 'sensor_11',\n",
       " 'sensor_12',\n",
       " 'sensor_13',\n",
       " 'sensor_14',\n",
       " 'sensor_15',\n",
       " 'sensor_17',\n",
       " 'sensor_20',\n",
       " 'sensor_21']"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train_correlated_features_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "11f00f82",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Features:  ['setting_1', 'setting_2', 'sensor_2', 'sensor_3', 'sensor_4', 'sensor_6', 'sensor_7', 'sensor_8', 'sensor_9', 'sensor_11', 'sensor_12', 'sensor_13', 'sensor_14', 'sensor_15', 'sensor_17', 'sensor_20', 'sensor_21']\n",
      "train set - MAE: 12.136687509088265, RMSE: 16.97431679289148, R2: 0.8340865632994863\n",
      "test set - MAE: 13.595714285714287, RMSE: 18.369823302450822, R2: 0.7898641026340808\n",
      "\n",
      "Features:  ['sensor_2', 'sensor_3', 'sensor_4', 'sensor_6', 'sensor_7', 'sensor_8', 'sensor_9', 'sensor_11', 'sensor_12', 'sensor_13', 'sensor_14', 'sensor_15', 'sensor_17', 'sensor_20', 'sensor_21']\n",
      "train set - MAE: 11.913943649293365, RMSE: 16.86228982911834, R2: 0.8362693249800367\n",
      "test set - MAE: 13.227142857142859, RMSE: 18.531831412739653, R2: 0.7861412753479324\n",
      "\n",
      "Features:  ['sensor_2', 'sensor_3', 'sensor_4', 'sensor_7', 'sensor_8', 'sensor_9', 'sensor_11', 'sensor_12', 'sensor_13', 'sensor_14', 'sensor_15', 'sensor_17', 'sensor_20', 'sensor_21']\n",
      "train set - MAE: 11.894728459945851, RMSE: 16.846392742778516, R2: 0.8365778968023765\n",
      "test set - MAE: 13.115714285714285, RMSE: 18.3985636085305, R2: 0.7892060566284933\n",
      "\n",
      "Features:  ['sensor_2', 'sensor_3', 'sensor_4', 'sensor_7', 'sensor_8', 'sensor_11', 'sensor_12', 'sensor_13', 'sensor_15', 'sensor_17', 'sensor_20', 'sensor_21']\n",
      "train set - MAE: 14.131092599901674, RMSE: 19.23844669922587, R2: 0.7868737410153217\n",
      "test set - MAE: 15.362857142857145, RMSE: 20.383656918908226, R2: 0.7412654098873318\n",
      "\n",
      "Features:  ['sensor_4', 'sensor_7', 'sensor_11', 'sensor_12', 'sensor_15', 'sensor_20', 'sensor_21']\n",
      "train set - MAE: 14.956487117167647, RMSE: 20.217857241460322, R2: 0.7646212759587662\n",
      "test set - MAE: 16.89, RMSE: 22.70297562084795, R2: 0.6790363451275923\n"
     ]
    }
   ],
   "source": [
    "### x_train_correlated_features_2 yields the highest performance.\n",
    "### train set - MAE: 11.39000533178227, RMSE: 16.333464836910608, R2: 0.8463779356669445\n",
    "### test set - MAE: 12.606000000000002, RMSE: 17.615754312546485, R2: 0.8067618903801216\n",
    "x_train_correlated_features_list = [x_train_correlated_features_1, x_train_correlated_features_2, \n",
    "    x_train_correlated_features_3, x_train_correlated_features_4, x_train_correlated_features_5]\n",
    "neigh = KNeighborsRegressor(n_neighbors = 7, p = 1, weights = 'uniform')\n",
    "\n",
    "for selected_features in x_train_correlated_features_list:\n",
    "    print(\"\\nFeatures: \", selected_features)\n",
    "    x_train_filtered = df_x_train_scaled.loc[:, selected_features]\n",
    "    df_x_test_scaled_filtered = df_x_test_scaled.loc[:, selected_features]\n",
    "    \n",
    "    ### KNN regression\n",
    "    neigh.fit(x_train_filtered, y_train)\n",
    "\n",
    "    y_pred_train = neigh.predict(x_train_filtered)\n",
    "    evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "    y_pred_test = neigh.predict(df_x_test_scaled_filtered)#using scaled_x_test\n",
    "    #y_pred_test = neigh.predict(x_test)#using x_test\n",
    "    evaluateValue = evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68aef03b",
   "metadata": {},
   "source": [
    "#### Variance Threshold "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "07396579",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "setting_1    0.015802\n",
       "setting_2    0.059643\n",
       "setting_3    0.000000\n",
       "sensor_1     0.000000\n",
       "sensor_2     0.022686\n",
       "sensor_3     0.017866\n",
       "sensor_4     0.023084\n",
       "sensor_5     0.000000\n",
       "sensor_6     0.019293\n",
       "sensor_7     0.020314\n",
       "sensor_8     0.011568\n",
       "sensor_9     0.009819\n",
       "sensor_10    0.000000\n",
       "sensor_11    0.025275\n",
       "sensor_12    0.024731\n",
       "sensor_13    0.011186\n",
       "sensor_14    0.009691\n",
       "sensor_15    0.020824\n",
       "sensor_16    0.000000\n",
       "sensor_17    0.016657\n",
       "sensor_18    0.000000\n",
       "sensor_19    0.000000\n",
       "sensor_20    0.019632\n",
       "sensor_21    0.022343\n",
       "dtype: float64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Compute the variance of each column\n",
    "x_train_scaled_pd.var()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "cf12e3f3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Features:  Index(['setting_1', 'setting_2', 'sensor_2', 'sensor_3', 'sensor_4',\n",
      "       'sensor_6', 'sensor_7', 'sensor_8', 'sensor_11', 'sensor_12',\n",
      "       'sensor_13', 'sensor_15', 'sensor_17', 'sensor_20', 'sensor_21'],\n",
      "      dtype='object')\n",
      "train set - MAE: 14.277017248661863, RMSE: 19.353532746507547, R2: 0.7843162348937347\n",
      "test set - MAE: 15.235714285714284, RMSE: 20.610939239310046, R2: 0.735463345309292\n",
      "\n",
      "Features:  Index(['setting_2', 'sensor_2', 'sensor_4', 'sensor_7', 'sensor_11',\n",
      "       'sensor_12', 'sensor_15', 'sensor_21'],\n",
      "      dtype='object')\n",
      "train set - MAE: 14.901839811102572, RMSE: 20.141543920921514, R2: 0.7663948201638436\n",
      "test set - MAE: 15.65714285714286, RMSE: 21.213184194586848, R2: 0.7197781363197964\n",
      "\n",
      "Features:  Index(['setting_2'], dtype='object')\n",
      "train set - MAE: 37.45892104115166, RMSE: 53.70584238376998, R2: -0.6608867437873349\n",
      "test set - MAE: 47.472857142857144, RMSE: 61.28323504808786, R2: -1.338695376772482\n"
     ]
    }
   ],
   "source": [
    "### 0.01 variance threshold yield highest model performance.\n",
    "### train set - MAE: 13.612456982211235, RMSE: 18.78759384068961, R2: 0.7967459169715116\n",
    "### test set - MAE: 15.689999999999998, RMSE: 21.055422104531647, R2: 0.723930648076507\n",
    "\n",
    "# Import VarianceThreshold module\n",
    "from sklearn.feature_selection import VarianceThreshold\n",
    "neigh = KNeighborsRegressor(n_neighbors = 7, p = 1, weights = 'uniform')\n",
    "\n",
    "for x in [0.01, 0.02, 0.05]:\n",
    "    # Create VarianceThreshold object with different variance thresholds \n",
    "    thresholder = VarianceThreshold(threshold = x) # Change to 0.01, 0.02 and 0.05\n",
    "    # Conduct variance thresholding - fit_transform() takes in an array\n",
    "    features_high_variance = thresholder.fit_transform(x_train_scaled_pd)\n",
    "    # Use the get_support() function to identify the feature(s) above the variance threshold\n",
    "    selected_indices = thresholder.get_support(indices = True)\n",
    "    selected_features = x_train_scaled_pd.columns[selected_indices]\n",
    "    \n",
    "    print(\"\\nFeatures: \", selected_features)\n",
    "    x_train_filtered = df_x_train_scaled.loc[:, selected_features]\n",
    "    df_x_test_scaled_filtered = df_x_test_scaled.loc[:, selected_features]\n",
    "    \n",
    "    ### KNN regression\n",
    "    neigh.fit(x_train_filtered, y_train)\n",
    "\n",
    "    y_pred_train = neigh.predict(x_train_filtered)\n",
    "    evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "    y_pred_test = neigh.predict(df_x_test_scaled_filtered)#using scaled_x_test\n",
    "    #y_pred_test = neigh.predict(x_test)#using x_test\n",
    "    evaluateValue = evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "af20c079",
   "metadata": {},
   "source": [
    "#### Mutual Information"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "1785ae1c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_selection import SelectKBest\n",
    "from sklearn.feature_selection import f_regression\n",
    "\n",
    "# Assuming 'X' is your feature matrix and 'y' is your target variable\n",
    "# Select top 5 features based on f_regression score\n",
    "k = 23 # Change to different values like 1, 5, 10 and 15 \n",
    "selector = SelectKBest(score_func=f_regression, k=k)\n",
    "# Learn the relationship from training data\n",
    "selector.fit_transform(x_train_scaled, y_train) \n",
    "\n",
    "# Get the indices of the selected features\n",
    "selected_indices = selector.get_support(indices=True)\n",
    "selected_features = x_train_scaled_pd.columns[selected_indices]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "e91a7b4e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['setting_1', 'setting_2', 'sensor_1', 'sensor_2', 'sensor_3',\n",
       "       'sensor_4', 'sensor_5', 'sensor_6', 'sensor_7', 'sensor_8', 'sensor_9',\n",
       "       'sensor_10', 'sensor_11', 'sensor_12', 'sensor_13', 'sensor_14',\n",
       "       'sensor_15', 'sensor_16', 'sensor_17', 'sensor_18', 'sensor_19',\n",
       "       'sensor_20', 'sensor_21'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "ebb2d204",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scores of Selected Features (Descending Order):\n",
      "Feature: sensor_11, Score: 31070.328368\n",
      "Feature: sensor_4, Score: 27714.893659\n",
      "Feature: sensor_12, Score: 26341.140302\n",
      "Feature: sensor_7, Score: 23956.874707\n",
      "Feature: sensor_15, Score: 22315.602210\n",
      "Feature: sensor_21, Score: 20655.553949\n",
      "Feature: sensor_20, Score: 20341.971636\n",
      "Feature: sensor_17, Score: 17824.136428\n",
      "Feature: sensor_2, Score: 17594.486606\n",
      "Feature: sensor_3, Score: 15502.946945\n",
      "Feature: sensor_8, Score: 13193.758040\n",
      "Feature: sensor_13, Score: 13156.844657\n",
      "Feature: sensor_9, Score: 5602.645998\n",
      "Feature: sensor_14, Score: 3266.986488\n",
      "Feature: sensor_6, Score: 244.775779\n",
      "Feature: setting_2, Score: 1.037227\n",
      "Feature: setting_1, Score: 0.636893\n",
      "Feature: sensor_16, Score: 0.000000\n",
      "Feature: sensor_1, Score: 0.000000\n",
      "Feature: sensor_18, Score: 0.000000\n",
      "Feature: sensor_19, Score: 0.000000\n",
      "Feature: sensor_5, Score: 0.000000\n",
      "Feature: sensor_10, Score: 0.000000\n",
      "Feature with highscores:  ['sensor_11', 'sensor_4', 'sensor_12', 'sensor_7', 'sensor_15', 'sensor_21', 'sensor_20', 'sensor_17', 'sensor_2', 'sensor_3', 'sensor_8', 'sensor_13']\n"
     ]
    }
   ],
   "source": [
    "# Show the scores of the selected features\n",
    "selected_scores = selector.scores_[selected_indices]\n",
    "feature_name_highscore=[]\n",
    "# Sort the selected features and scores in descending order\n",
    "sorted_indices = np.argsort(-selected_scores)\n",
    "sorted_features = selected_features[sorted_indices]\n",
    "sorted_scores = selected_scores[sorted_indices]\n",
    "\n",
    "# Print the scores of the selected features\n",
    "print(\"Scores of Selected Features (Descending Order):\")\n",
    "# for feature_name, score in zip(selected_features, selected_scores):\n",
    "#     print(\"Feature: %s, Score: %f\" % (feature_name, score))\n",
    "    \n",
    "for i in range(len(sorted_features)):\n",
    "    feature_name = sorted_features[i]\n",
    "    score = sorted_scores[i]\n",
    "    \n",
    "    if(score > 10000):\n",
    "        feature_name_highscore.append(feature_name)\n",
    "    print(\"Feature: %s, Score: %f\" % (feature_name, score))\n",
    "\n",
    "print(\"Feature with highscores: \", feature_name_highscore)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "56291c56",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Features:  ['sensor_11', 'sensor_4', 'sensor_12', 'sensor_7', 'sensor_15', 'sensor_21', 'sensor_20', 'sensor_17', 'sensor_2', 'sensor_3', 'sensor_8', 'sensor_13']\n",
      "train set - MAE: 14.131092599901674, RMSE: 19.23844669922587, R2: 0.7868737410153217\n",
      "test set - MAE: 15.362857142857145, RMSE: 20.383656918908226, R2: 0.7412654098873318\n"
     ]
    }
   ],
   "source": [
    "### Selecting Features with score > 10000\n",
    "### train set - MAE: 13.367059279724685, RMSE: 18.528960913017126, R2: 0.8023034532919944\n",
    "### test set - MAE: 14.806000000000001, RMSE: 19.918624450498584, R2: 0.7529362789893936\n",
    "neigh = KNeighborsRegressor(n_neighbors = 7, p = 1, weights = 'uniform')\n",
    "\n",
    "x_train_filtered = df_x_train_scaled.loc[:, feature_name_highscore]\n",
    "df_x_test_scaled_filtered = df_x_test_scaled.loc[:, feature_name_highscore]\n",
    "\n",
    "print(\"\\nFeatures: \", x_train_filtered.columns.tolist())\n",
    "### KNN regression\n",
    "neigh.fit(x_train_filtered, y_train)\n",
    "\n",
    "y_pred_train = neigh.predict(x_train_filtered)\n",
    "evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "y_pred_test = neigh.predict(df_x_test_scaled_filtered)#using scaled_x_test\n",
    "#y_pred_test = neigh.predict(x_test)#using x_test\n",
    "evaluateValue = evaluate(y_test, y_pred_test, 'test')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "825f2faf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the feature scores as a bar chart\n",
    "plt.figure(figsize=(10, 6))\n",
    "plt.bar(range(len(sorted_scores)), sorted_scores)\n",
    "plt.xticks(range(len(sorted_scores)), sorted_features, rotation='vertical')\n",
    "plt.xlabel(\"Selected Features\")\n",
    "plt.ylabel(\"Feature Score\")\n",
    "plt.title(\"Feature Scores of Selected Features\")\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a4270e3d",
   "metadata": {},
   "source": [
    "### Experiment Set 3: Ensemble learning"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c7a74238",
   "metadata": {},
   "source": [
    "• Goal: Examine if ensemble learning can help improve model performance. <br>\n",
    "• Explore at least 3 ensemble learning methods to examine if model performance can be improved."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "80c08edf",
   "metadata": {},
   "outputs": [],
   "source": [
    "metric_df = pd.DataFrame(columns = ['ensemble type','model', 'train/ test', 'MAE', 'RMSE', 'R2'])\n",
    "\n",
    "linear_reg = LinearRegression()\n",
    "tree_reg = DecisionTreeRegressor(max_depth = 10, max_features = None, min_samples_leaf = 4, min_samples_split = 10)\n",
    "knn_reg = KNeighborsRegressor(n_neighbors = 7, p = 1, weights = 'distance')\n",
    "br_reg = BayesianRidge()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "247fe498",
   "metadata": {},
   "source": [
    "### Bagging"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "840ca0ff",
   "metadata": {},
   "source": [
    "Linear Regressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "7e015428",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 17.562470534187053, RMSE: 21.469586457353124, R2: 0.7345734810282857\n",
      "test set - MAE: 16.571424022552705, RMSE: 20.832435972921644, R2: 0.7297470752936468\n"
     ]
    }
   ],
   "source": [
    "bagging_reg = BaggingRegressor(estimator = linear_reg, n_estimators = 10, random_state = 42)\n",
    "bagging_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = bagging_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['bagging', 'linear regressor', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = bagging_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['bagging', 'linear regressor', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b070786",
   "metadata": {},
   "source": [
    "Decision Tree Regressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "2fd55b54",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 11.435702268253877, RMSE: 15.861465457681469, R2: 0.85512829463455\n",
      "test set - MAE: 12.541042247829475, RMSE: 17.61517650147877, R2: 0.8067745669065163\n"
     ]
    }
   ],
   "source": [
    "bagging_reg = BaggingRegressor(estimator = tree_reg, n_estimators = 10, random_state = 42)\n",
    "bagging_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = bagging_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['bagging', 'decision tree regressor', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = bagging_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['bagging', 'decision tree regressor', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "7fd8d75f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 5.170746478500697, RMSE: 7.764651412186009, R2: 0.9652831047275834\n",
      "test set - MAE: 13.851890032128512, RMSE: 18.731877058463326, R2: 0.7814992718058092\n"
     ]
    }
   ],
   "source": [
    "bagging_reg = BaggingRegressor(estimator = knn_reg, n_estimators = 10, random_state = 42)\n",
    "bagging_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = bagging_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['bagging', 'knn regressor', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = bagging_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['bagging', 'knn regressor', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "08dbb48e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 17.56268310697441, RMSE: 21.46971263700095, R2: 0.7345703611244002\n",
      "test set - MAE: 16.569187686976576, RMSE: 20.829045143721473, R2: 0.7298350445480795\n"
     ]
    }
   ],
   "source": [
    "bagging_reg = BaggingRegressor(estimator = br_reg, n_estimators = 10, random_state = 42)\n",
    "bagging_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = bagging_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['bagging', 'bayesian ridge', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = bagging_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['bagging', 'bayesian ridge', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "53d9e55b",
   "metadata": {},
   "source": [
    "### Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "0442bd53",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 17.63353327564782, RMSE: 21.503338283140252, R2: 0.7337382836229128\n",
      "test set - MAE: 16.896787061848226, RMSE: 21.086276591929856, R2: 0.7231209545548627\n"
     ]
    }
   ],
   "source": [
    "boosting_reg = AdaBoostRegressor(estimator = linear_reg, n_estimators = 10, random_state = 42)\n",
    "boosting_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = boosting_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['boosting', 'linear regressor', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = boosting_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['boosting', 'linear regressor', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "4d63ecb2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 12.477362833006673, RMSE: 16.170369362451705, R2: 0.849430560672633\n",
      "test set - MAE: 13.317677136581256, RMSE: 17.825588424403016, R2: 0.8021308715219588\n"
     ]
    }
   ],
   "source": [
    "boosting_reg = AdaBoostRegressor(estimator = tree_reg, n_estimators = 10, random_state = 42)\n",
    "boosting_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = boosting_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['boosting', 'decision tree regressor', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = boosting_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['boosting', 'decision tree regressor', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "fc9fc4ac",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 1.1011059326108315, RMSE: 2.7212694630990044, R2: 0.9957357756145416\n",
      "test set - MAE: 16.2302559813445, RMSE: 22.07665552248048, R2: 0.6965012872741569\n"
     ]
    }
   ],
   "source": [
    "boosting_reg = AdaBoostRegressor(estimator = knn_reg, n_estimators = 10, random_state = 42)\n",
    "boosting_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = boosting_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['boosting', 'knn regressor', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = boosting_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['boosting', 'knn regressor', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "ddd8808c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 17.62605100585966, RMSE: 21.49857204946637, R2: 0.7338563048247568\n",
      "test set - MAE: 16.903581310967088, RMSE: 21.098437600942365, R2: 0.7228014956396722\n"
     ]
    }
   ],
   "source": [
    "boosting_reg = AdaBoostRegressor(estimator = br_reg, n_estimators = 10, random_state = 42)\n",
    "boosting_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = boosting_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['boosting', 'bayesian ridge', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = boosting_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['boosting', 'bayesian ridge', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "05a84ea5",
   "metadata": {},
   "source": [
    "### Voting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "ad164157",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 11.233624703621619, RMSE: 14.100570421290813, R2: 0.8855092720568583\n",
      "test set - MAE: 14.108728370578246, RMSE: 18.38136820053915, R2: 0.7895998910720892\n"
     ]
    }
   ],
   "source": [
    "vooting_reg = VotingRegressor(estimators = [('lr', linear_reg), ('dtr', tree_reg), ('knnr', knn_reg), ('brr', br_reg)])\n",
    "vooting_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = vooting_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['vooting', 'all', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = vooting_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['vooting', 'all', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7323666b",
   "metadata": {},
   "source": [
    "### Stacking"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "fffa3c17",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 6.879727135807473, RMSE: 9.058628855265649, R2: 0.9527478160825618\n",
      "test set - MAE: 13.099129864483064, RMSE: 17.73766214981567, R2: 0.8040780708613858\n"
     ]
    }
   ],
   "source": [
    "stacking_reg = StackingRegressor(estimators = [('lr', linear_reg), ('dtr', tree_reg), ('knnr', knn_reg), ('brr', br_reg)], final_estimator = None)\n",
    "stacking_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = stacking_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['stacking - no meta', 'all', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = stacking_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['stacking - no meta', 'all', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "4565b276",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 6.867982255756549, RMSE: 9.041130485743277, R2: 0.9529301919358604\n",
      "test set - MAE: 13.118247970145408, RMSE: 17.752703091672906, R2: 0.8037456595509327\n"
     ]
    }
   ],
   "source": [
    "stacking_reg = StackingRegressor(estimators = [('lr', linear_reg), ('dtr', tree_reg), ('knnr', knn_reg), ('brr', br_reg)], final_estimator = linear_reg)\n",
    "stacking_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = stacking_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['stacking - linear regression', 'all', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = stacking_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['stacking - linear regression', 'all', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "f0d5107a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 10.721901810282509, RMSE: 15.469305992246369, R2: 0.8622033646721645\n",
      "test set - MAE: 13.691454978623266, RMSE: 18.763251901120462, R2: 0.7807667059051256\n"
     ]
    }
   ],
   "source": [
    "stacking_reg = StackingRegressor(estimators = [('lr', linear_reg), ('dtr', tree_reg), ('knnr', knn_reg), ('brr', br_reg)], final_estimator = tree_reg)\n",
    "stacking_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = stacking_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['stacking - decision tree regression', 'all', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = stacking_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['stacking - decision tree regression', 'all', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "413d657d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 11.107148148875718, RMSE: 16.00614718952463, R2: 0.852473321192039\n",
      "test set - MAE: 13.704336048747429, RMSE: 18.308116044101663, R2: 0.7912734935576611\n"
     ]
    }
   ],
   "source": [
    "stacking_reg = StackingRegressor(estimators = [('lr', linear_reg), ('dtr', tree_reg), ('knnr', knn_reg), ('brr', br_reg)], final_estimator = knn_reg)\n",
    "stacking_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = stacking_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['stacking - knn regression', 'all', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = stacking_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['stacking - knn regression', 'all', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "415c5d4a",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 6.913483548979727, RMSE: 9.10688260825713, R2: 0.9522430668353004\n",
      "test set - MAE: 13.075894218774154, RMSE: 17.718136956838386, R2: 0.8045091657803136\n"
     ]
    }
   ],
   "source": [
    "stacking_reg = StackingRegressor(estimators = [('lr', linear_reg), ('dtr', tree_reg), ('knnr', knn_reg), ('brr', br_reg)], final_estimator = br_reg)\n",
    "stacking_reg.fit(x_train_scaled, y_train)\n",
    "\n",
    "y_pred_train = stacking_reg.predict(x_train_scaled)\n",
    "result = evaluate(y_train, y_pred_train, 'train')\n",
    "\n",
    "metric_df.loc[len(metric_df)] = ['stacking - bayesian ridge', 'all', result['label'], result['mae'], result['rmse'], result['r2']]\n",
    "\n",
    "y_pred_test = stacking_reg.predict(x_test_scaled)\n",
    "result = evaluate(y_test, y_pred_test, 'test')\n",
    "metric_df.loc[len(metric_df)] = ['stacking - bayesian ridge', 'all', result['label'], result['mae'], result['rmse'], result['r2']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "25172a43",
   "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>ensemble type</th>\n",
       "      <th>model</th>\n",
       "      <th>train/ test</th>\n",
       "      <th>MAE</th>\n",
       "      <th>RMSE</th>\n",
       "      <th>R2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>bagging</td>\n",
       "      <td>linear regressor</td>\n",
       "      <td>train</td>\n",
       "      <td>17.562471</td>\n",
       "      <td>21.469586</td>\n",
       "      <td>0.734573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bagging</td>\n",
       "      <td>linear regressor</td>\n",
       "      <td>test</td>\n",
       "      <td>16.571424</td>\n",
       "      <td>20.832436</td>\n",
       "      <td>0.729747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>bagging</td>\n",
       "      <td>decision tree regressor</td>\n",
       "      <td>train</td>\n",
       "      <td>11.435702</td>\n",
       "      <td>15.861465</td>\n",
       "      <td>0.855128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>bagging</td>\n",
       "      <td>decision tree regressor</td>\n",
       "      <td>test</td>\n",
       "      <td>12.541042</td>\n",
       "      <td>17.615177</td>\n",
       "      <td>0.806775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>bagging</td>\n",
       "      <td>knn regressor</td>\n",
       "      <td>train</td>\n",
       "      <td>5.170746</td>\n",
       "      <td>7.764651</td>\n",
       "      <td>0.965283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>bagging</td>\n",
       "      <td>knn regressor</td>\n",
       "      <td>test</td>\n",
       "      <td>13.851890</td>\n",
       "      <td>18.731877</td>\n",
       "      <td>0.781499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>bagging</td>\n",
       "      <td>bayesian ridge</td>\n",
       "      <td>train</td>\n",
       "      <td>17.562683</td>\n",
       "      <td>21.469713</td>\n",
       "      <td>0.734570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>bagging</td>\n",
       "      <td>bayesian ridge</td>\n",
       "      <td>test</td>\n",
       "      <td>16.569188</td>\n",
       "      <td>20.829045</td>\n",
       "      <td>0.729835</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>boosting</td>\n",
       "      <td>linear regressor</td>\n",
       "      <td>train</td>\n",
       "      <td>17.633533</td>\n",
       "      <td>21.503338</td>\n",
       "      <td>0.733738</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>boosting</td>\n",
       "      <td>linear regressor</td>\n",
       "      <td>test</td>\n",
       "      <td>16.896787</td>\n",
       "      <td>21.086277</td>\n",
       "      <td>0.723121</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>boosting</td>\n",
       "      <td>decision tree regressor</td>\n",
       "      <td>train</td>\n",
       "      <td>12.477363</td>\n",
       "      <td>16.170369</td>\n",
       "      <td>0.849431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>boosting</td>\n",
       "      <td>decision tree regressor</td>\n",
       "      <td>test</td>\n",
       "      <td>13.317677</td>\n",
       "      <td>17.825588</td>\n",
       "      <td>0.802131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>boosting</td>\n",
       "      <td>knn regressor</td>\n",
       "      <td>train</td>\n",
       "      <td>1.101106</td>\n",
       "      <td>2.721269</td>\n",
       "      <td>0.995736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>boosting</td>\n",
       "      <td>knn regressor</td>\n",
       "      <td>test</td>\n",
       "      <td>16.230256</td>\n",
       "      <td>22.076656</td>\n",
       "      <td>0.696501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>boosting</td>\n",
       "      <td>bayesian ridge</td>\n",
       "      <td>train</td>\n",
       "      <td>17.626051</td>\n",
       "      <td>21.498572</td>\n",
       "      <td>0.733856</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>boosting</td>\n",
       "      <td>bayesian ridge</td>\n",
       "      <td>test</td>\n",
       "      <td>16.903581</td>\n",
       "      <td>21.098438</td>\n",
       "      <td>0.722801</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>vooting</td>\n",
       "      <td>all</td>\n",
       "      <td>train</td>\n",
       "      <td>11.233625</td>\n",
       "      <td>14.100570</td>\n",
       "      <td>0.885509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>vooting</td>\n",
       "      <td>all</td>\n",
       "      <td>test</td>\n",
       "      <td>14.108728</td>\n",
       "      <td>18.381368</td>\n",
       "      <td>0.789600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>stacking - no meta</td>\n",
       "      <td>all</td>\n",
       "      <td>train</td>\n",
       "      <td>6.879727</td>\n",
       "      <td>9.058629</td>\n",
       "      <td>0.952748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>stacking - no meta</td>\n",
       "      <td>all</td>\n",
       "      <td>test</td>\n",
       "      <td>13.099130</td>\n",
       "      <td>17.737662</td>\n",
       "      <td>0.804078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>stacking - linear regression</td>\n",
       "      <td>all</td>\n",
       "      <td>train</td>\n",
       "      <td>6.867982</td>\n",
       "      <td>9.041130</td>\n",
       "      <td>0.952930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>stacking - linear regression</td>\n",
       "      <td>all</td>\n",
       "      <td>test</td>\n",
       "      <td>13.118248</td>\n",
       "      <td>17.752703</td>\n",
       "      <td>0.803746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>stacking - decision tree regression</td>\n",
       "      <td>all</td>\n",
       "      <td>train</td>\n",
       "      <td>10.721902</td>\n",
       "      <td>15.469306</td>\n",
       "      <td>0.862203</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>stacking - decision tree regression</td>\n",
       "      <td>all</td>\n",
       "      <td>test</td>\n",
       "      <td>13.691455</td>\n",
       "      <td>18.763252</td>\n",
       "      <td>0.780767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>stacking - knn regression</td>\n",
       "      <td>all</td>\n",
       "      <td>train</td>\n",
       "      <td>11.107148</td>\n",
       "      <td>16.006147</td>\n",
       "      <td>0.852473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>stacking - knn regression</td>\n",
       "      <td>all</td>\n",
       "      <td>test</td>\n",
       "      <td>13.704336</td>\n",
       "      <td>18.308116</td>\n",
       "      <td>0.791273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>stacking - bayesian ridge</td>\n",
       "      <td>all</td>\n",
       "      <td>train</td>\n",
       "      <td>6.913484</td>\n",
       "      <td>9.106883</td>\n",
       "      <td>0.952243</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>stacking - bayesian ridge</td>\n",
       "      <td>all</td>\n",
       "      <td>test</td>\n",
       "      <td>13.075894</td>\n",
       "      <td>17.718137</td>\n",
       "      <td>0.804509</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          ensemble type                    model train/ test  \\\n",
       "0                               bagging         linear regressor       train   \n",
       "1                               bagging         linear regressor        test   \n",
       "2                               bagging  decision tree regressor       train   \n",
       "3                               bagging  decision tree regressor        test   \n",
       "4                               bagging            knn regressor       train   \n",
       "5                               bagging            knn regressor        test   \n",
       "6                               bagging           bayesian ridge       train   \n",
       "7                               bagging           bayesian ridge        test   \n",
       "8                              boosting         linear regressor       train   \n",
       "9                              boosting         linear regressor        test   \n",
       "10                             boosting  decision tree regressor       train   \n",
       "11                             boosting  decision tree regressor        test   \n",
       "12                             boosting            knn regressor       train   \n",
       "13                             boosting            knn regressor        test   \n",
       "14                             boosting           bayesian ridge       train   \n",
       "15                             boosting           bayesian ridge        test   \n",
       "16                              vooting                      all       train   \n",
       "17                              vooting                      all        test   \n",
       "18                   stacking - no meta                      all       train   \n",
       "19                   stacking - no meta                      all        test   \n",
       "20         stacking - linear regression                      all       train   \n",
       "21         stacking - linear regression                      all        test   \n",
       "22  stacking - decision tree regression                      all       train   \n",
       "23  stacking - decision tree regression                      all        test   \n",
       "24            stacking - knn regression                      all       train   \n",
       "25            stacking - knn regression                      all        test   \n",
       "26            stacking - bayesian ridge                      all       train   \n",
       "27            stacking - bayesian ridge                      all        test   \n",
       "\n",
       "          MAE       RMSE        R2  \n",
       "0   17.562471  21.469586  0.734573  \n",
       "1   16.571424  20.832436  0.729747  \n",
       "2   11.435702  15.861465  0.855128  \n",
       "3   12.541042  17.615177  0.806775  \n",
       "4    5.170746   7.764651  0.965283  \n",
       "5   13.851890  18.731877  0.781499  \n",
       "6   17.562683  21.469713  0.734570  \n",
       "7   16.569188  20.829045  0.729835  \n",
       "8   17.633533  21.503338  0.733738  \n",
       "9   16.896787  21.086277  0.723121  \n",
       "10  12.477363  16.170369  0.849431  \n",
       "11  13.317677  17.825588  0.802131  \n",
       "12   1.101106   2.721269  0.995736  \n",
       "13  16.230256  22.076656  0.696501  \n",
       "14  17.626051  21.498572  0.733856  \n",
       "15  16.903581  21.098438  0.722801  \n",
       "16  11.233625  14.100570  0.885509  \n",
       "17  14.108728  18.381368  0.789600  \n",
       "18   6.879727   9.058629  0.952748  \n",
       "19  13.099130  17.737662  0.804078  \n",
       "20   6.867982   9.041130  0.952930  \n",
       "21  13.118248  17.752703  0.803746  \n",
       "22  10.721902  15.469306  0.862203  \n",
       "23  13.691455  18.763252  0.780767  \n",
       "24  11.107148  16.006147  0.852473  \n",
       "25  13.704336  18.308116  0.791273  \n",
       "26   6.913484   9.106883  0.952243  \n",
       "27  13.075894  17.718137  0.804509  "
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metric_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "646c9921",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12.541042247829475"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metric_df[metric_df['train/ test'] == 'test']['MAE'].min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "be2ebc41",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17.61517650147877"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metric_df[metric_df['train/ test'] == 'test']['RMSE'].min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "66cb1b6e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8067745669065163"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metric_df[metric_df['train/ test'] == 'test']['R2'].max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "44695c07",
   "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>ensemble type</th>\n",
       "      <th>model</th>\n",
       "      <th>train/ test</th>\n",
       "      <th>MAE</th>\n",
       "      <th>RMSE</th>\n",
       "      <th>R2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>bagging</td>\n",
       "      <td>decision tree regressor</td>\n",
       "      <td>test</td>\n",
       "      <td>12.541042</td>\n",
       "      <td>17.615177</td>\n",
       "      <td>0.806775</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  ensemble type                    model train/ test        MAE       RMSE  \\\n",
       "3       bagging  decision tree regressor        test  12.541042  17.615177   \n",
       "\n",
       "         R2  \n",
       "3  0.806775  "
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metric_df[metric_df['MAE'] == 12.541042247829475]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "48fd1fa1",
   "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>ensemble type</th>\n",
       "      <th>model</th>\n",
       "      <th>train/ test</th>\n",
       "      <th>MAE</th>\n",
       "      <th>RMSE</th>\n",
       "      <th>R2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>bagging</td>\n",
       "      <td>decision tree regressor</td>\n",
       "      <td>test</td>\n",
       "      <td>12.541042</td>\n",
       "      <td>17.615177</td>\n",
       "      <td>0.806775</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  ensemble type                    model train/ test        MAE       RMSE  \\\n",
       "3       bagging  decision tree regressor        test  12.541042  17.615177   \n",
       "\n",
       "         R2  \n",
       "3  0.806775  "
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metric_df[metric_df['RMSE'] == 17.61517650147877]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "a506294e",
   "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>ensemble type</th>\n",
       "      <th>model</th>\n",
       "      <th>train/ test</th>\n",
       "      <th>MAE</th>\n",
       "      <th>RMSE</th>\n",
       "      <th>R2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>bagging</td>\n",
       "      <td>decision tree regressor</td>\n",
       "      <td>test</td>\n",
       "      <td>12.541042</td>\n",
       "      <td>17.615177</td>\n",
       "      <td>0.806775</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  ensemble type                    model train/ test        MAE       RMSE  \\\n",
       "3       bagging  decision tree regressor        test  12.541042  17.615177   \n",
       "\n",
       "         R2  \n",
       "3  0.806775  "
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metric_df[metric_df['R2'] == 0.8067745669065163]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3518495e",
   "metadata": {},
   "source": [
    "### Experiment Set 4: Varying training sample size"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "02cd10de",
   "metadata": {},
   "source": [
    "• Goal: Examine if more training data will lead to a better performing model.<br>\n",
    "• Select the 3 best performing models from the three other experiment sets. Use the same training set for each model. You will train each model with increments of 10% following the steps below.<br>\n",
    "&emsp; i. Randomly subsample 10% of the training data and train each model (train-10P). Then, evaluate the performance of the models on the test set.<br>\n",
    "&emsp; ii. Randomly subsample another 10% from the remaining 90% of the training data. Add this second 10% sample to the first 10% sample in (i). Train another round of the models with the sample containing 20% of the training data (train-20P). Then, evaluate the performance of the models on the test set.<br>\n",
    "&emsp; iii. Repeat (i) and (ii) in 10% increments until you reach 100% of the training data. You will have a set of 10 results (10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% and 100%) for each model."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2d8b309a",
   "metadata": {},
   "source": [
    "#### EXP 1 : KNN Regressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "e8f2daa6",
   "metadata": {},
   "outputs": [],
   "source": [
    "column_names=['fraction','model','MAE(Train)','MAE(Test)','RMSE(Train)','RMSE(Test)','R2(Train)','R2(Test)']\n",
    "df_final = pd.DataFrame({}, columns = column_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "1228e996",
   "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>fraction</th>\n",
       "      <th>model</th>\n",
       "      <th>MAE(Train)</th>\n",
       "      <th>MAE(Test)</th>\n",
       "      <th>RMSE(Train)</th>\n",
       "      <th>RMSE(Test)</th>\n",
       "      <th>R2(Train)</th>\n",
       "      <th>R2(Test)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [fraction, model, MAE(Train), MAE(Test), RMSE(Train), RMSE(Test), R2(Train), R2(Test)]\n",
       "Index: []"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "043392f9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fraction:  0.1 Seed:  [267, 116, 251, 63, 318]\n",
      "train set - MAE: 11.842185444221313, RMSE: 16.85190184236456, R2: 0.8402360969928768\n",
      "test set - MAE: 14.778571428571427, RMSE: 20.77473445887419, R2: 0.7312420907473782\n",
      "train set - MAE: 12.888442628626827, RMSE: 17.608860769608306, R2: 0.8279992856587974\n",
      "test set - MAE: 14.41, RMSE: 20.325529337686213, R2: 0.7427389601837369\n",
      "train set - MAE: 12.770514507305588, RMSE: 17.628105860713145, R2: 0.8192052414550376\n",
      "test set - MAE: 14.761428571428567, RMSE: 20.631593565724014, R2: 0.7349328926195813\n",
      "train set - MAE: 12.532442351637698, RMSE: 17.649346480338593, R2: 0.8227052725766049\n",
      "test set - MAE: 15.497142857142858, RMSE: 21.13326037026826, R2: 0.7218857135613299\n",
      "train set - MAE: 12.577037601274151, RMSE: 17.42671442403997, R2: 0.8204113701548736\n",
      "test set - MAE: 13.83285714285714, RMSE: 19.283401977782205, R2: 0.7684431674240031\n",
      "Average Train MAE:  12.522124506613114 Average Train RMSE:  17.432985875412918 Average Train Variance:  0.826111453367638\n",
      "Average Test MAE:  14.655999999999997 Average Test RMSE:  20.429703942066975 Average Test Variance:  0.7398485649072059\n",
      "\n",
      "\n",
      "Fraction:  0.2 Seed:  [116, 407, 165, 201, 381]\n",
      "train set - MAE: 12.25199085935877, RMSE: 17.22574483960032, R2: 0.8321414187432139\n",
      "test set - MAE: 14.56857142857143, RMSE: 20.484698121838584, R2: 0.7386939724834973\n",
      "train set - MAE: 11.629942524755903, RMSE: 16.506184706591483, R2: 0.8440050057130998\n",
      "test set - MAE: 13.537142857142857, RMSE: 19.210328854391786, R2: 0.7701947796478367\n",
      "train set - MAE: 12.019527733536458, RMSE: 16.770749981523547, R2: 0.8369705220118536\n",
      "test set - MAE: 14.694285714285714, RMSE: 19.6634956511931, R2: 0.7592248045218613\n",
      "train set - MAE: 11.976248182258846, RMSE: 16.772912580018648, R2: 0.8408827848951831\n",
      "test set - MAE: 14.37, RMSE: 19.40116238890898, R2: 0.7656063765903374\n",
      "train set - MAE: 12.03725503773977, RMSE: 17.067901303881218, R2: 0.8336222068193443\n",
      "test set - MAE: 15.17428571428571, RMSE: 20.5269360120018, R2: 0.7376152752074401\n",
      "Average Train MAE:  11.98299286752995 Average Train RMSE:  16.868698682323046 Average Train Variance:  0.837524387636539\n",
      "Average Test MAE:  14.468857142857141 Average Test RMSE:  19.85732420566685 Average Test Variance:  0.7542670416901945\n",
      "\n",
      "\n",
      "Fraction:  0.3 Seed:  [3, 324, 454, 409, 321]\n",
      "train set - MAE: 11.552131662165593, RMSE: 16.450375473290862, R2: 0.846412831637115\n",
      "test set - MAE: 14.82, RMSE: 20.752939304973605, R2: 0.7318057126158177\n",
      "train set - MAE: 12.267640745100756, RMSE: 17.070757695020653, R2: 0.8306034125110261\n",
      "test set - MAE: 14.68142857142857, RMSE: 19.734322116245977, R2: 0.7574871716453847\n",
      "train set - MAE: 11.900030007155552, RMSE: 16.802195470381264, R2: 0.8356811576469819\n",
      "test set - MAE: 13.911428571428571, RMSE: 19.31402155356125, R2: 0.7677072183281444\n",
      "train set - MAE: 12.209703852457125, RMSE: 17.084107777462084, R2: 0.8294032832758287\n",
      "test set - MAE: 13.801428571428572, RMSE: 18.556197050570244, R2: 0.7855785430743247\n",
      "train set - MAE: 11.875401057175171, RMSE: 16.714064259616094, R2: 0.8366148365173801\n",
      "test set - MAE: 14.482857142857142, RMSE: 19.372882372754226, R2: 0.7662892041662395\n",
      "Average Train MAE:  11.960981464810839 Average Train RMSE:  16.824300135154193 Average Train Variance:  0.8357431043176664\n",
      "Average Test MAE:  14.339428571428574 Average Test RMSE:  19.546072479621063 Average Test Variance:  0.7617735699659821\n",
      "\n",
      "\n",
      "Fraction:  0.4 Seed:  [470, 338, 300, 428, 417]\n",
      "train set - MAE: 11.60852780278374, RMSE: 16.532960261719033, R2: 0.8478851240101242\n",
      "test set - MAE: 13.532857142857145, RMSE: 18.18884497797079, R2: 0.7939841975551202\n",
      "train set - MAE: 11.875060591371788, RMSE: 16.736035111385185, R2: 0.8390825269897715\n",
      "test set - MAE: 13.957142857142856, RMSE: 18.55283684782575, R2: 0.7856561919946441\n",
      "train set - MAE: 11.849127484246244, RMSE: 16.836383838825046, R2: 0.8331997421954115\n",
      "test set - MAE: 14.784285714285714, RMSE: 20.591542745823133, R2: 0.7359610100752016\n",
      "train set - MAE: 11.642181981857213, RMSE: 16.48660285621216, R2: 0.843885134313262\n",
      "test set - MAE: 14.559999999999997, RMSE: 19.540789364291093, R2: 0.762220451575493\n",
      "train set - MAE: 11.780312997714839, RMSE: 16.60900649105007, R2: 0.8426336815326488\n",
      "test set - MAE: 13.51, RMSE: 18.230155551409084, R2: 0.7930473271122485\n",
      "Average Train MAE:  11.751042171594765 Average Train RMSE:  16.640197711838297 Average Train Variance:  0.8413372418082437\n",
      "Average Test MAE:  14.068857142857144 Average Test RMSE:  19.02083389746397 Average Test Variance:  0.7741738356625415\n",
      "\n",
      "\n",
      "Fraction:  0.5 Seed:  [125, 266, 244, 277, 37]\n",
      "train set - MAE: 11.531075167562179, RMSE: 16.476075114861004, R2: 0.8436558307413943\n",
      "test set - MAE: 14.865714285714285, RMSE: 19.93778077020651, R2: 0.7524608337606831\n",
      "train set - MAE: 11.280008862792887, RMSE: 16.262234403148806, R2: 0.8525697052251104\n",
      "test set - MAE: 14.209999999999997, RMSE: 19.837600867122653, R2: 0.7549421679165527\n",
      "train set - MAE: 11.724048634575972, RMSE: 16.77136633972943, R2: 0.8382915144636119\n",
      "test set - MAE: 14.892857142857144, RMSE: 19.72509538012686, R2: 0.7577138912425203\n",
      "train set - MAE: 11.452708691076277, RMSE: 16.2967317747762, R2: 0.8479562200177718\n",
      "test set - MAE: 13.87, RMSE: 18.97086380398604, R2: 0.7758883136563959\n",
      "train set - MAE: 11.419279344153326, RMSE: 16.39557755861456, R2: 0.8455399751868519\n",
      "test set - MAE: 14.477142857142859, RMSE: 19.278167835716154, R2: 0.7685688544652403\n",
      "Average Train MAE:  11.481424140032129 Average Train RMSE:  16.440397038226003 Average Train Variance:  0.8456026491269482\n",
      "Average Test MAE:  14.463142857142856 Average Test RMSE:  19.54990173143164 Average Test Variance:  0.7619148122082784\n",
      "\n",
      "\n",
      "Fraction:  0.6 Seed:  [458, 306, 366, 390, 400]\n",
      "train set - MAE: 11.171442419766194, RMSE: 16.017565536623312, R2: 0.8519427184340027\n",
      "test set - MAE: 15.321428571428571, RMSE: 20.797012148039784, R2: 0.7306653791264135\n",
      "train set - MAE: 11.286325920625943, RMSE: 16.179025687027398, R2: 0.8497966475699558\n",
      "test set - MAE: 14.308571428571426, RMSE: 19.407667325580583, R2: 0.7654491724750591\n",
      "train set - MAE: 11.442777514915813, RMSE: 16.35082644538632, R2: 0.8463290487592208\n",
      "test set - MAE: 14.528571428571432, RMSE: 20.479925462236807, R2: 0.7388157198904794\n",
      "train set - MAE: 11.45638350662989, RMSE: 16.441168536024858, R2: 0.8419632719983531\n",
      "test set - MAE: 14.594285714285716, RMSE: 19.372524199947485, R2: 0.7662978459446057\n",
      "train set - MAE: 11.361487773071909, RMSE: 16.35719608597406, R2: 0.845922109223342\n",
      "test set - MAE: 14.41142857142857, RMSE: 20.227986270793952, R2: 0.7452022482730184\n",
      "Average Train MAE:  11.343683427001949 Average Train RMSE:  16.269156458207192 Average Train Variance:  0.8471907591969747\n",
      "Average Test MAE:  14.632857142857143 Average Test RMSE:  20.05702308131972 Average Test Variance:  0.7492860731419152\n",
      "\n",
      "\n",
      "Fraction:  0.7 Seed:  [66, 54, 385, 96, 263]\n",
      "train set - MAE: 11.099996043286446, RMSE: 16.038882532518706, R2: 0.8516739730514217\n",
      "test set - MAE: 14.678571428571429, RMSE: 20.464359261897258, R2: 0.7392126062704426\n",
      "train set - MAE: 11.107523690822404, RMSE: 15.985553151367037, R2: 0.8533557007703988\n",
      "test set - MAE: 15.014285714285716, RMSE: 21.04439428000352, R2: 0.7242197562300471\n",
      "train set - MAE: 11.179763388529489, RMSE: 16.060426746722932, R2: 0.8512484227636467\n",
      "test set - MAE: 14.104285714285716, RMSE: 19.660646490086823, R2: 0.7592945741736705\n",
      "train set - MAE: 11.205432567709261, RMSE: 16.23291381442729, R2: 0.8487639533452618\n",
      "test set - MAE: 14.13857142857143, RMSE: 19.434836174839685, R2: 0.7647920160643035\n",
      "train set - MAE: 11.003709418956614, RMSE: 15.930336526261092, R2: 0.8537398045668299\n",
      "test set - MAE: 15.147142857142859, RMSE: 21.107001449852646, R2: 0.7225764204057423\n",
      "Average Train MAE:  11.119285021860842 Average Train RMSE:  16.049622554259408 Average Train Variance:  0.8517563708995117\n",
      "Average Test MAE:  14.61657142857143 Average Test RMSE:  20.342247531335985 Average Test Variance:  0.7420190746288412\n",
      "\n",
      "\n",
      "Fraction:  0.8 Seed:  [268, 125, 140, 34, 379]\n",
      "train set - MAE: 10.83708832821223, RMSE: 15.874862617670466, R2: 0.855701826546301\n",
      "test set - MAE: 13.877142857142857, RMSE: 18.60535362935812, R2: 0.7844410054543919\n",
      "train set - MAE: 10.934920154065868, RMSE: 15.878257493008721, R2: 0.8556846803435503\n",
      "test set - MAE: 14.598571428571427, RMSE: 19.946658458836396, R2: 0.7522403413273702\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 10.894516813086943, RMSE: 15.730231309947863, R2: 0.8570273648398634\n",
      "test set - MAE: 13.9, RMSE: 18.962325699089956, R2: 0.7760899975132647\n",
      "train set - MAE: 11.090907517202579, RMSE: 16.012971097547823, R2: 0.8515235378434625\n",
      "test set - MAE: 14.232857142857148, RMSE: 19.46052519374353, R2: 0.7641698080219403\n",
      "train set - MAE: 11.06785822478037, RMSE: 16.036301431567775, R2: 0.8530366003061302\n",
      "test set - MAE: 14.768571428571427, RMSE: 19.97316567143966, R2: 0.751581405726951\n",
      "Average Train MAE:  10.9650582074696 Average Train RMSE:  15.90652478994853 Average Train Variance:  0.8545948019758616\n",
      "Average Test MAE:  14.275428571428572 Average Test RMSE:  19.389605730493532 Average Test Variance:  0.7657045116087836\n",
      "\n",
      "\n",
      "Fraction:  0.9 Seed:  [137, 11, 106, 156, 15]\n",
      "train set - MAE: 10.831815104326951, RMSE: 15.838884069698944, R2: 0.856205790429008\n",
      "test set - MAE: 14.460000000000004, RMSE: 18.820993489143113, R2: 0.77941530299457\n",
      "train set - MAE: 10.761463654828583, RMSE: 15.6884527924333, R2: 0.8562111295904453\n",
      "test set - MAE: 16.03, RMSE: 21.153226221828117, R2: 0.7213599630157304\n",
      "train set - MAE: 10.93068720379147, RMSE: 16.03460910956315, R2: 0.8516611360749584\n",
      "test set - MAE: 13.052857142857144, RMSE: 17.519126574072335, R2: 0.8088760150396173\n",
      "train set - MAE: 10.933626207915307, RMSE: 15.96077673046465, R2: 0.8516385130356547\n",
      "test set - MAE: 13.931428571428574, RMSE: 18.60109718071842, R2: 0.7845396233957467\n",
      "train set - MAE: 10.899058287683879, RMSE: 15.87126024755487, R2: 0.8549325812356194\n",
      "test set - MAE: 13.515714285714287, RMSE: 18.419715480086733, R2: 0.7887211003601811\n",
      "Average Train MAE:  10.871330091709238 Average Train RMSE:  15.878796589942983 Average Train Variance:  0.8541298300731371\n",
      "Average Test MAE:  14.198000000000002 Average Test RMSE:  18.902831789169742 Average Test Variance:  0.7765824009611692\n",
      "\n",
      "\n",
      "Fraction:  1 Seed:  [48, 277, 427, 430, 252]\n",
      "train set - MAE: 10.641399558223753, RMSE: 15.551030943388005, R2: 0.8589850013240014\n",
      "test set - MAE: 14.331428571428571, RMSE: 19.286825364819926, R2: 0.7683609434445486\n",
      "train set - MAE: 10.628243212364195, RMSE: 15.528269880451026, R2: 0.8599204153469362\n",
      "test set - MAE: 14.317142857142859, RMSE: 18.901689806761205, R2: 0.7775197034929592\n",
      "train set - MAE: 10.56938587562406, RMSE: 15.62418654125, R2: 0.8602620452168684\n",
      "test set - MAE: 14.890000000000002, RMSE: 20.156373378043533, R2: 0.7470031694675295\n",
      "train set - MAE: 10.628374775822792, RMSE: 15.57201570437973, R2: 0.8611722249297475\n",
      "test set - MAE: 14.37, RMSE: 19.937161487114093, R2: 0.7524762110427757\n",
      "train set - MAE: 10.620764868401919, RMSE: 15.494935088177336, R2: 0.8615343642670392\n",
      "test set - MAE: 15.471428571428572, RMSE: 20.69769779172435, R2: 0.7332316060462329\n",
      "Average Train MAE:  10.617633658087346 Average Train RMSE:  15.55408763152922 Average Train Variance:  0.8603748102169186\n",
      "Average Test MAE:  14.675999999999998 Average Test RMSE:  19.79594956569262 Average Test Variance:  0.7557183266988092\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import random\n",
    "neigh = KNeighborsRegressor(n_neighbors = 7, p = 1, weights = 'uniform')\n",
    "#neigh = KNeighborsRegressor()\n",
    "\n",
    "for fraction in [.1,.2,.3,.4,.5,.6,.7,.8,.9,1]:\n",
    "    sumTrainMAE, sumTrainRMSE, sumTrainVariance, sumTestMAE, sumTestRMSE, sumTestVariance = \\\n",
    "    [],[],[],[],[],[]\n",
    "    seed = [random.randint(1, 500) for _ in range(5)]\n",
    "    print('Fraction: ', fraction, 'Seed: ', seed)\n",
    "    \n",
    "    for x in range(5):  \n",
    "        ###Setting fraction and seeds for each run\n",
    "        sub_x_train = df_x_train_scaled.sample(frac = fraction, random_state = seed[x], replace=True)#using scaled x_train\n",
    "        sub_y_train = y_train.sample(frac = fraction, random_state = seed[x], replace=True)\n",
    "\n",
    "        ### KNN regression\n",
    "        neigh.fit(sub_x_train, sub_y_train)\n",
    "        \n",
    "        y_pred_train = neigh.predict(sub_x_train)\n",
    "        evaluateTrainValue = evaluate(sub_y_train, y_pred_train, 'train')   \n",
    "        y_pred_test = neigh.predict(df_x_test_scaled)\n",
    "        evaluateTestValue = evaluate(y_test, y_pred_test, 'test')\n",
    "        \n",
    "        ###Adding results from each run into list to be processed later\n",
    "        sumTrainMAE.append(evaluateTrainValue['mae'])\n",
    "        sumTrainRMSE.append(evaluateTrainValue['rmse'])\n",
    "        sumTrainVariance.append(evaluateTrainValue['r2'])\n",
    "        \n",
    "        sumTestMAE.append(evaluateTestValue['mae'])\n",
    "        sumTestRMSE.append(evaluateTestValue['rmse'])\n",
    "        sumTestVariance.append(evaluateTestValue['r2'])\n",
    "        \n",
    "    print(\"Average Train MAE: \", np.mean(sumTrainMAE), \"Average Train RMSE: \", np.mean(sumTrainRMSE), \\\n",
    "          \"Average Train Variance: \", np.mean(sumTrainVariance))\n",
    "    print(\"Average Test MAE: \", np.mean(sumTestMAE), \"Average Test RMSE: \", np.mean(sumTestRMSE), \\\n",
    "          \"Average Test Variance: \", np.mean(sumTestVariance))\n",
    "    \n",
    "    ###Finalizing and tabulate the results for each experiment\n",
    "    df_final.loc[len(df_final)] = [fraction, 'KNN Regressor',  \\\n",
    "                                   round(np.mean(sumTrainMAE),2), round(np.mean(sumTestMAE),2), \\\n",
    "                                   round(np.mean(sumTrainRMSE),2), round(np.mean(sumTestRMSE),2), \\\n",
    "                                   round(np.mean(sumTrainVariance),2), round(np.mean(sumTestVariance),2)]\n",
    "    print('\\n')\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6874870a",
   "metadata": {},
   "source": [
    "#### EXP 2: Feature Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "a7366cf5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selected features:  ['sensor_2', 'sensor_3', 'sensor_4', 'sensor_6', 'sensor_7', 'sensor_8', 'sensor_9', 'sensor_11', 'sensor_12', 'sensor_13', 'sensor_14', 'sensor_15', 'sensor_17', 'sensor_20', 'sensor_21']\n",
      "Fraction:  0.1 Seed:  [500, 299, 97, 120, 327]\n",
      "train set - MAE: 12.137940585832006, RMSE: 16.940783252510585, R2: 0.8266568133618928\n",
      "test set - MAE: 14.594285714285716, RMSE: 19.098968879101772, R2: 0.772851363985573\n",
      "train set - MAE: 12.461048403850148, RMSE: 17.501098948124895, R2: 0.8282551211634839\n",
      "test set - MAE: 13.23, RMSE: 17.932520908183793, R2: 0.799749788745056\n",
      "train set - MAE: 11.99141333702652, RMSE: 16.73486346014887, R2: 0.8328607054678188\n",
      "test set - MAE: 14.395714285714284, RMSE: 19.111440104218804, R2: 0.7725546205666763\n",
      "train set - MAE: 12.238626133924242, RMSE: 17.052658395937392, R2: 0.8314902805380212\n",
      "test set - MAE: 12.958571428571428, RMSE: 18.108150155132257, R2: 0.7958081211302823\n",
      "train set - MAE: 11.52039332456201, RMSE: 16.485968386477452, R2: 0.8442353618926444\n",
      "test set - MAE: 13.995714285714284, RMSE: 18.268848248494777, R2: 0.7921678990785164\n",
      "Average Train MAE:  12.069884357038985 Average Train RMSE:  16.943074488639837 Average Train Variance:  0.8326996564847722\n",
      "Average Test MAE:  13.834857142857143 Average Test RMSE:  18.503985659026277 Average Test Variance:  0.7866263587012208\n",
      "\n",
      "\n",
      "Fraction:  0.2 Seed:  [104, 114, 322, 295, 462]\n",
      "train set - MAE: 12.107402534450522, RMSE: 17.087597325885827, R2: 0.8276444836111434\n",
      "test set - MAE: 14.05571428571429, RMSE: 19.38741965338158, R2: 0.7659383225475785\n",
      "train set - MAE: 12.104563395886712, RMSE: 16.672296937618867, R2: 0.8357862325544152\n",
      "test set - MAE: 12.655714285714284, RMSE: 17.5676592937484, R2: 0.8078156180000999\n",
      "train set - MAE: 11.855134685963575, RMSE: 17.09274986638477, R2: 0.8329228807609499\n",
      "test set - MAE: 13.688571428571429, RMSE: 19.206344614980427, R2: 0.7702900933798164\n",
      "train set - MAE: 12.130669621217367, RMSE: 17.27086838963485, R2: 0.8270432341967553\n",
      "test set - MAE: 13.742857142857142, RMSE: 18.742846254352237, R2: 0.7812432932889634\n",
      "train set - MAE: 12.208157329824806, RMSE: 17.045919922205506, R2: 0.8333575038847375\n",
      "test set - MAE: 13.508571428571424, RMSE: 17.99440956270041, R2: 0.7983651979318318\n",
      "Average Train MAE:  12.081185513468597 Average Train RMSE:  17.033886488345964 Average Train Variance:  0.8313508670016002\n",
      "Average Test MAE:  13.530285714285714 Average Test RMSE:  18.579735875832608 Average Test Variance:  0.784730505029658\n",
      "\n",
      "\n",
      "Fraction:  0.3 Seed:  [213, 364, 493, 249, 371]\n",
      "train set - MAE: 11.595872861990168, RMSE: 16.573398619095517, R2: 0.8425569188894038\n",
      "test set - MAE: 12.707142857142857, RMSE: 18.390553253765688, R2: 0.7893895673337983\n",
      "train set - MAE: 11.461094568704846, RMSE: 16.336465648124573, R2: 0.8467497412102666\n",
      "test set - MAE: 14.02142857142857, RMSE: 19.669965709690704, R2: 0.7590663295568232\n",
      "train set - MAE: 11.556240334233548, RMSE: 16.48660733144098, R2: 0.8444557040607877\n",
      "test set - MAE: 13.917142857142858, RMSE: 18.1951664359701, R2: 0.7938409727871989\n",
      "train set - MAE: 11.561364633104816, RMSE: 16.496033517563067, R2: 0.8410158997549446\n",
      "test set - MAE: 13.838571428571427, RMSE: 19.149716255157966, R2: 0.7716426587790952\n",
      "train set - MAE: 11.793897006209175, RMSE: 16.88457346753051, R2: 0.8352295885626344\n",
      "test set - MAE: 14.245714285714289, RMSE: 19.443712587458023, R2: 0.7645771153699333\n",
      "Average Train MAE:  11.59369388084851 Average Train RMSE:  16.55541571675093 Average Train Variance:  0.8420015704956073\n",
      "Average Test MAE:  13.746 Average Test RMSE:  18.969822848408494 Average Test Variance:  0.7757033287653698\n",
      "\n",
      "\n",
      "Fraction:  0.4 Seed:  [44, 94, 458, 103, 467]\n",
      "train set - MAE: 11.559102555224706, RMSE: 16.603160456171, R2: 0.840946845981302\n",
      "test set - MAE: 13.728571428571426, RMSE: 19.45286296541001, R2: 0.7643554791718368\n",
      "train set - MAE: 11.400872515753756, RMSE: 16.52315453863709, R2: 0.8440148680161305\n",
      "test set - MAE: 14.064285714285713, RMSE: 19.173041235132253, R2: 0.7710860265843342\n",
      "train set - MAE: 11.360778339450178, RMSE: 16.217617096734624, R2: 0.8477998815510783\n",
      "test set - MAE: 14.894285714285719, RMSE: 19.43543996236289, R2: 0.7647774012920666\n",
      "train set - MAE: 11.218405927567343, RMSE: 16.209737406009715, R2: 0.8496849655250259\n",
      "test set - MAE: 13.755714285714284, RMSE: 19.725726493954173, R2: 0.7576983868754517\n",
      "train set - MAE: 11.46953119590056, RMSE: 16.517952933093262, R2: 0.8405285100007163\n",
      "test set - MAE: 14.597142857142858, RMSE: 20.10188335170897, R2: 0.7483692058742615\n",
      "Average Train MAE:  11.401738106779309 Average Train RMSE:  16.414324486129136 Average Train Variance:  0.8445950142148506\n",
      "Average Test MAE:  14.208000000000002 Average Test RMSE:  19.57779080171366 Average Test Variance:  0.7612572999595901\n",
      "\n",
      "\n",
      "Fraction:  0.5 Seed:  [224, 87, 382, 115, 96]\n",
      "train set - MAE: 11.25814269096549, RMSE: 16.261204906829015, R2: 0.8509482169565367\n",
      "test set - MAE: 14.405714285714284, RMSE: 19.92022866846868, R2: 0.752896481058318\n",
      "train set - MAE: 11.208843405528166, RMSE: 16.161834387165474, R2: 0.8486657446582329\n",
      "test set - MAE: 13.557142857142855, RMSE: 18.83679638259074, R2: 0.7790447232046332\n",
      "train set - MAE: 11.192710352849941, RMSE: 16.162174365650333, R2: 0.8471638981176153\n",
      "test set - MAE: 13.775714285714285, RMSE: 19.682806129788943, R2: 0.7587516671563145\n",
      "train set - MAE: 11.484407023763364, RMSE: 16.331376258229106, R2: 0.8447578757647649\n",
      "test set - MAE: 13.66142857142857, RMSE: 18.480949926994597, R2: 0.7873140155062098\n",
      "train set - MAE: 11.174056943444302, RMSE: 16.197376830292928, R2: 0.8498582279321942\n",
      "test set - MAE: 13.415714285714287, RMSE: 18.745540966385818, R2: 0.7811803862258567\n",
      "Average Train MAE:  11.263632083310252 Average Train RMSE:  16.22279334963337 Average Train Variance:  0.8482787926858688\n",
      "Average Test MAE:  13.763142857142856 Average Test RMSE:  19.13326441484576 Average Test Variance:  0.7718374546302664\n",
      "\n",
      "\n",
      "Fraction:  0.6 Seed:  [370, 458, 215, 362, 262]\n",
      "train set - MAE: 11.0692070672683, RMSE: 16.104751042607752, R2: 0.850184369856417\n",
      "test set - MAE: 14.017142857142858, RMSE: 20.369454923427654, R2: 0.7416258228791908\n",
      "train set - MAE: 10.983381994853037, RMSE: 15.87599595491467, R2: 0.8545483302775162\n",
      "test set - MAE: 14.542857142857144, RMSE: 19.189123194653575, R2: 0.7707018487019683\n",
      "train set - MAE: 11.055151004581493, RMSE: 16.23176197022675, R2: 0.8490320273986572\n",
      "test set - MAE: 14.87714285714286, RMSE: 20.061487116017158, R2: 0.7493795314332456\n",
      "train set - MAE: 11.188487415323186, RMSE: 16.174455014505035, R2: 0.8488959536169732\n",
      "test set - MAE: 13.572857142857144, RMSE: 18.19957949727276, R2: 0.7937409569111081\n",
      "train set - MAE: 11.155147542497085, RMSE: 16.212120686675856, R2: 0.8491534220641455\n",
      "test set - MAE: 13.751428571428573, RMSE: 19.05714285714286, R2: 0.7738451684976805\n",
      "Average Train MAE:  11.09027500490462 Average Train RMSE:  16.119816933786012 Average Train Variance:  0.8503628206427418\n",
      "Average Test MAE:  14.152285714285716 Average Test RMSE:  19.375357517702803 Average Test Variance:  0.7658586656846387\n",
      "\n",
      "\n",
      "Fraction:  0.7 Seed:  [171, 126, 372, 341, 298]\n",
      "train set - MAE: 10.867004965675509, RMSE: 15.873807059172044, R2: 0.8541735758167244\n",
      "test set - MAE: 14.947142857142858, RMSE: 19.815562838643896, R2: 0.7554863457836684\n",
      "train set - MAE: 11.017468890339682, RMSE: 15.981945256485943, R2: 0.8541364426359204\n",
      "test set - MAE: 13.227142857142859, RMSE: 18.619821564720773, R2: 0.7841056282027996\n",
      "train set - MAE: 10.745682236334499, RMSE: 15.882252385471888, R2: 0.854583764672608\n",
      "test set - MAE: 15.321428571428571, RMSE: 20.49008222013165, R2: 0.7385565936244708\n",
      "train set - MAE: 10.79852414584446, RMSE: 15.873009628001753, R2: 0.8563342800671843\n",
      "test set - MAE: 13.975714285714284, RMSE: 19.6845064941534, R2: 0.7587099832841954\n",
      "train set - MAE: 10.872851009951132, RMSE: 16.020955445887985, R2: 0.8527881453125239\n",
      "test set - MAE: 14.348571428571429, RMSE: 20.086029259875417, R2: 0.7487659651692488\n",
      "Average Train MAE:  10.860306249629058 Average Train RMSE:  15.926393955003922 Average Train Variance:  0.8544032417009921\n",
      "Average Test MAE:  14.363999999999999 Average Test RMSE:  19.739200475505026 Average Test Variance:  0.7571249032128766\n",
      "\n",
      "\n",
      "Fraction:  0.8 Seed:  [297, 42, 386, 83, 264]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 10.835989094213875, RMSE: 15.885999739641802, R2: 0.8554966216589149\n",
      "test set - MAE: 14.76142857142857, RMSE: 20.565777993788515, R2: 0.7366213436103564\n",
      "train set - MAE: 10.853005582723851, RMSE: 15.899697974325804, R2: 0.8558059601912421\n",
      "test set - MAE: 13.942857142857143, RMSE: 18.93669367516248, R2: 0.7766949219988954\n",
      "train set - MAE: 10.759484139005496, RMSE: 15.716933134455159, R2: 0.8574283197433482\n",
      "test set - MAE: 14.052857142857144, RMSE: 19.584766049776796, R2: 0.7611489981430661\n",
      "train set - MAE: 10.819734279655513, RMSE: 15.76200813371892, R2: 0.8558426704795967\n",
      "test set - MAE: 14.877142857142855, RMSE: 20.87281211802295, R2: 0.7286984849532818\n",
      "train set - MAE: 10.878634180118578, RMSE: 16.0477012521707, R2: 0.8515594863315592\n",
      "test set - MAE: 15.32285714285714, RMSE: 21.03455854210457, R2: 0.72447748456132\n",
      "Average Train MAE:  10.82936945514346 Average Train RMSE:  15.862468046862478 Average Train Variance:  0.8552266116809323\n",
      "Average Test MAE:  14.59142857142857 Average Test RMSE:  20.19892167577106 Average Test Variance:  0.7455282466533839\n",
      "\n",
      "\n",
      "Fraction:  0.9 Seed:  [304, 159, 464, 129, 206]\n",
      "train set - MAE: 10.437526928048255, RMSE: 15.408616951293133, R2: 0.863309916925419\n",
      "test set - MAE: 13.241428571428571, RMSE: 18.257673812981338, R2: 0.7924220690304621\n",
      "train set - MAE: 10.591232227488153, RMSE: 15.517609697663287, R2: 0.8613646636148435\n",
      "test set - MAE: 14.252857142857147, RMSE: 19.300919931549423, R2: 0.768022261983581\n",
      "train set - MAE: 10.669377423524343, RMSE: 15.630437029401246, R2: 0.8583531780524223\n",
      "test set - MAE: 13.718571428571432, RMSE: 18.80970765431553, R2: 0.7796797668295695\n",
      "train set - MAE: 10.609012433064565, RMSE: 15.634604584399051, R2: 0.8580642245740714\n",
      "test set - MAE: 13.517142857142858, RMSE: 19.028067346390582, R2: 0.7745347315773091\n",
      "train set - MAE: 10.386579060749677, RMSE: 15.477303477945132, R2: 0.8641390621053291\n",
      "test set - MAE: 14.581428571428571, RMSE: 19.97114755575847, R2: 0.7516316042924602\n",
      "Average Train MAE:  10.538745614574998 Average Train RMSE:  15.53371434814037 Average Train Variance:  0.8610462090544171\n",
      "Average Test MAE:  13.862285714285715 Average Test RMSE:  19.073503260199068 Average Test Variance:  0.7732580867426764\n",
      "\n",
      "\n",
      "Fraction:  1 Seed:  [443, 430, 500, 406, 309]\n",
      "train set - MAE: 10.462251673971899, RMSE: 15.423578037320357, R2: 0.8629263133872878\n",
      "test set - MAE: 14.91142857142857, RMSE: 20.845534930306112, R2: 0.7294071107793063\n",
      "train set - MAE: 10.414147918873816, RMSE: 15.463737848000983, R2: 0.8630961521591617\n",
      "test set - MAE: 14.294285714285714, RMSE: 19.434001342061944, R2: 0.7648122225754831\n",
      "train set - MAE: 10.435177299071437, RMSE: 15.391087274412813, R2: 0.863013399105344\n",
      "test set - MAE: 14.117142857142856, RMSE: 19.83009463066741, R2: 0.7551275848964971\n",
      "train set - MAE: 10.596217896785005, RMSE: 15.715383003348567, R2: 0.8586768414666156\n",
      "test set - MAE: 15.014285714285712, RMSE: 21.373806208458397, R2: 0.7155185020951388\n",
      "train set - MAE: 10.497593773586212, RMSE: 15.527007169956146, R2: 0.8604398384663903\n",
      "test set - MAE: 14.871428571428572, RMSE: 20.166247817093, R2: 0.7467552266794064\n",
      "Average Train MAE:  10.481077712457672 Average Train RMSE:  15.504158666607774 Average Train Variance:  0.8616305089169598\n",
      "Average Test MAE:  14.641714285714283 Average Test RMSE:  20.32993698571737 Average Test Variance:  0.7423241294051663\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import random\n",
    "neigh = KNeighborsRegressor(n_neighbors = 7, p = 1, weights = 'uniform')\n",
    "#neigh = KNeighborsRegressor()\n",
    "selected_features = x_train_correlated_features_2\n",
    "\n",
    "print('Selected features: ', selected_features)\n",
    "for fraction in [.1,.2,.3,.4,.5,.6,.7,.8,.9,1]:\n",
    "    sumTrainMAE, sumTrainRMSE, sumTrainVariance, sumTestMAE, sumTestRMSE, sumTestVariance = \\\n",
    "    [],[],[],[],[],[]\n",
    "    seed = [random.randint(1, 500) for _ in range(5)]\n",
    "\n",
    "    print('Fraction: ', fraction, 'Seed: ', seed)\n",
    "    for x in range(5):\n",
    "        ### Include only selected features in the evaluation\n",
    "        x_train_filtered = df_x_train_scaled.loc[:, selected_features]\n",
    "        df_x_test_scaled_filtered = df_x_test_scaled.loc[:, selected_features]\n",
    "        ###Setting fraction and seeds for each run\n",
    "        sub_x_train = x_train_filtered.sample(frac = fraction, random_state = seed[x], replace=True)\n",
    "        sub_y_train = y_train.sample(frac = fraction, random_state = seed[x], replace=True)\n",
    "\n",
    "        ### KNN regression\n",
    "        neigh.fit(sub_x_train, sub_y_train)\n",
    "        \n",
    "        y_pred_train = neigh.predict(sub_x_train)\n",
    "        evaluateTrainValue = evaluate(sub_y_train, y_pred_train, 'train')    \n",
    "        y_pred_test = neigh.predict(df_x_test_scaled_filtered)\n",
    "        evaluateTestValue = evaluate(y_test, y_pred_test, 'test')\n",
    "        \n",
    "        \n",
    "        ###Adding results from each run into list to be processed later\n",
    "        sumTrainMAE.append(evaluateTrainValue['mae'])\n",
    "        sumTrainRMSE.append(evaluateTrainValue['rmse'])\n",
    "        sumTrainVariance.append(evaluateTrainValue['r2'])\n",
    "        \n",
    "        sumTestMAE.append(evaluateTestValue['mae'])\n",
    "        sumTestRMSE.append(evaluateTestValue['rmse'])\n",
    "        sumTestVariance.append(evaluateTestValue['r2'])\n",
    "        \n",
    "    print(\"Average Train MAE: \", np.mean(sumTrainMAE), \"Average Train RMSE: \", np.mean(sumTrainRMSE), \\\n",
    "          \"Average Train Variance: \", np.mean(sumTrainVariance))\n",
    "    print(\"Average Test MAE: \", np.mean(sumTestMAE), \"Average Test RMSE: \", np.mean(sumTestRMSE), \\\n",
    "          \"Average Test Variance: \", np.mean(sumTestVariance))\n",
    "    \n",
    "    ###Finalizing and tabulate the results for each experiment\n",
    "    df_final.loc[len(df_final)] = [fraction, 'KNN Regressor(Feature Selection)',  \\\n",
    "                                   round(np.mean(sumTrainMAE),2), round(np.mean(sumTestMAE),2), \\\n",
    "                                   round(np.mean(sumTrainRMSE),2), round(np.mean(sumTestRMSE),2), \\\n",
    "                                   round(np.mean(sumTrainVariance),2), round(np.mean(sumTestVariance),2)]\n",
    "    \n",
    "    print('\\n')\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6917c1d4",
   "metadata": {},
   "source": [
    "#### Ensemble Learning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "603fcb1d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fraction:  0.1 Seed:  [399, 63, 332, 174, 497]\n",
      "train set - MAE: 9.379108779661557, RMSE: 13.095583340124403, R2: 0.9037349903287555\n",
      "test set - MAE: 12.152067136332715, RMSE: 17.060069369308668, R2: 0.8187609084276106\n",
      "train set - MAE: 9.251355783595178, RMSE: 13.050707662797516, R2: 0.9030591807998956\n",
      "test set - MAE: 11.738254430122042, RMSE: 15.87047180538121, R2: 0.8431552570025859\n",
      "train set - MAE: 9.444561768690352, RMSE: 13.186231873299464, R2: 0.8996951643752648\n",
      "test set - MAE: 12.76900991430701, RMSE: 17.493854404658297, R2: 0.809427028113227\n",
      "train set - MAE: 9.173615324296392, RMSE: 12.930391976350494, R2: 0.905521145718791\n",
      "test set - MAE: 13.812783486239026, RMSE: 18.912027472351937, R2: 0.7772762801943532\n",
      "train set - MAE: 9.66788841432214, RMSE: 13.627421523717528, R2: 0.8927680289252452\n",
      "test set - MAE: 12.560543495387936, RMSE: 18.128137531450548, R2: 0.7953571073832767\n",
      "Average Train MAE:  9.383306014113124 Average Train RMSE:  13.17806727525788 Average Train Variance:  0.9009557020295904\n",
      "Average Test MAE:  12.606531692477745 Average Test RMSE:  17.49291211663013 Average Test Variance:  0.8087953162242106\n",
      "\n",
      "\n",
      "Fraction:  0.2 Seed:  [477, 387, 187, 25, 393]\n",
      "train set - MAE: 9.48663329657012, RMSE: 13.382203059383933, R2: 0.8985995197370964\n",
      "test set - MAE: 13.054436897047292, RMSE: 18.350130226077688, R2: 0.7903144068150019\n",
      "train set - MAE: 9.28906351862789, RMSE: 13.10692312730769, R2: 0.8999739604023352\n",
      "test set - MAE: 12.568984337633836, RMSE: 17.39888610726609, R2: 0.8114905259782534\n",
      "train set - MAE: 9.609265353019405, RMSE: 13.38113769106764, R2: 0.8954507962250011\n",
      "test set - MAE: 13.185219739958477, RMSE: 18.193770568130127, R2: 0.7938726031345936\n",
      "train set - MAE: 9.804719083894557, RMSE: 13.785439439368771, R2: 0.891001374261986\n",
      "test set - MAE: 12.424400600202503, RMSE: 17.689804849128766, R2: 0.8051338634101135\n",
      "train set - MAE: 9.747822006553282, RMSE: 13.69504127226603, R2: 0.8914383120613721\n",
      "test set - MAE: 12.701452294290783, RMSE: 17.218052992072757, R2: 0.8153886613697421\n",
      "Average Train MAE:  9.587500651733048 Average Train RMSE:  13.470148917878811 Average Train Variance:  0.8952927925375581\n",
      "Average Test MAE:  12.786898773826577 Average Test RMSE:  17.770128948535085 Average Test Variance:  0.8032400121415411\n",
      "\n",
      "\n",
      "Fraction:  0.3 Seed:  [136, 389, 461, 94, 21]\n",
      "train set - MAE: 9.90079550304928, RMSE: 13.839940610982547, R2: 0.8873532766380984\n",
      "test set - MAE: 12.870362293294823, RMSE: 18.002542464886076, R2: 0.798182891676912\n",
      "train set - MAE: 9.674466554466774, RMSE: 13.646005751191465, R2: 0.8908074812425015\n",
      "test set - MAE: 12.575422977887378, RMSE: 17.710069526877994, R2: 0.8046871471981016\n",
      "train set - MAE: 9.591480426117355, RMSE: 13.447854471286403, R2: 0.895829737408973\n",
      "test set - MAE: 11.955498289845677, RMSE: 16.97844857622467, R2: 0.820490970733572\n",
      "train set - MAE: 9.524531123874388, RMSE: 13.635599224398277, R2: 0.8939844412286307\n",
      "test set - MAE: 12.801844097217568, RMSE: 17.88915202420698, R2: 0.8007172072757007\n",
      "train set - MAE: 9.642096897391218, RMSE: 13.632697691082315, R2: 0.8940456360292812\n",
      "test set - MAE: 12.785318493225647, RMSE: 17.795375489709308, R2: 0.8028010475212016\n",
      "Average Train MAE:  9.666674100979803 Average Train RMSE:  13.640419549788202 Average Train Variance:  0.892404114509497\n",
      "Average Test MAE:  12.597689230294218 Average Test RMSE:  17.675117616381005 Average Test Variance:  0.8053758528810976\n",
      "\n",
      "\n",
      "Fraction:  0.4 Seed:  [53, 402, 95, 342, 16]\n",
      "train set - MAE: 9.74852051949251, RMSE: 13.769185028884301, R2: 0.8895459358171259\n",
      "test set - MAE: 13.121750457722902, RMSE: 18.890824390235274, R2: 0.7777754103979887\n",
      "train set - MAE: 9.620530456059052, RMSE: 13.603762381149753, R2: 0.8951960303302442\n",
      "test set - MAE: 13.269593392876564, RMSE: 18.20309745453638, R2: 0.7936612099445635\n",
      "train set - MAE: 9.67233741158145, RMSE: 13.764944817964526, R2: 0.8904765926577398\n",
      "test set - MAE: 12.214820881678206, RMSE: 17.111049759894467, R2: 0.8176761009948925\n",
      "train set - MAE: 10.007337899288506, RMSE: 14.110520833757958, R2: 0.8862773012989567\n",
      "test set - MAE: 12.846779171277522, RMSE: 18.425295005552947, R2: 0.7885930837745615\n",
      "train set - MAE: 10.073803784681136, RMSE: 14.268933243009561, R2: 0.8824485183228885\n",
      "test set - MAE: 11.355057970405504, RMSE: 16.884209521047204, R2: 0.8224781738526864\n",
      "Average Train MAE:  9.824506014220532 Average Train RMSE:  13.90346926095322 Average Train Variance:  0.888788875685391\n",
      "Average Test MAE:  12.561600374792139 Average Test RMSE:  17.902895226253257 Average Test Variance:  0.8000367957929386\n",
      "\n",
      "\n",
      "Fraction:  0.5 Seed:  [193, 296, 54, 415, 228]\n",
      "train set - MAE: 9.978872105983708, RMSE: 14.066866090433994, R2: 0.8869838052048911\n",
      "test set - MAE: 13.049421648197939, RMSE: 18.45658284922733, R2: 0.7878744974473969\n",
      "train set - MAE: 10.224363356141993, RMSE: 14.384497042962861, R2: 0.8790622594992286\n",
      "test set - MAE: 12.647195913964941, RMSE: 18.13195328861794, R2: 0.7952709485305463\n",
      "train set - MAE: 9.987602258483289, RMSE: 14.179737430727496, R2: 0.8842559993288164\n",
      "test set - MAE: 12.72826539942237, RMSE: 17.362848139407372, R2: 0.8122706291072445\n",
      "train set - MAE: 10.133824071329608, RMSE: 14.259888742966135, R2: 0.8818587019679918\n",
      "test set - MAE: 13.04110602407476, RMSE: 18.274380372279943, R2: 0.7920420096983279\n",
      "train set - MAE: 10.172525397012238, RMSE: 14.2897486262093, R2: 0.8814636982742178\n",
      "test set - MAE: 13.012284331551955, RMSE: 18.164301045304708, R2: 0.7945398157292194\n",
      "Average Train MAE:  10.099437437790167 Average Train RMSE:  14.236147586659957 Average Train Variance:  0.8827248928550292\n",
      "Average Test MAE:  12.895654663442391 Average Test RMSE:  18.078013138967457 Average Test Variance:  0.7963995801025469\n",
      "\n",
      "\n",
      "Fraction:  0.6 Seed:  [487, 168, 452, 228, 446]\n",
      "train set - MAE: 10.085695946434846, RMSE: 14.171918118872725, R2: 0.8852596034446253\n",
      "test set - MAE: 12.816380858293641, RMSE: 18.328504264944442, R2: 0.7908083521273788\n",
      "train set - MAE: 10.231643964037431, RMSE: 14.410480595750947, R2: 0.8791489847901522\n",
      "test set - MAE: 12.929048559686686, RMSE: 18.134718823751715, R2: 0.7952084920975517\n",
      "train set - MAE: 9.813462638621578, RMSE: 14.005826723604942, R2: 0.8873656105145461\n",
      "test set - MAE: 12.413913592119464, RMSE: 17.463098522738754, R2: 0.8100965303707308\n",
      "train set - MAE: 10.088977716467303, RMSE: 14.246357886171388, R2: 0.8829055856843522\n",
      "test set - MAE: 12.940694882416272, RMSE: 18.221896303237322, R2: 0.7932348061805012\n",
      "train set - MAE: 10.092312433305333, RMSE: 14.255675205706693, R2: 0.884503612764569\n",
      "test set - MAE: 12.843874191068176, RMSE: 17.299388733872586, R2: 0.8136403840505909\n",
      "Average Train MAE:  10.062418539773299 Average Train RMSE:  14.218051706021338 Average Train Variance:  0.883836679439649\n",
      "Average Test MAE:  12.788782416716847 Average Test RMSE:  17.889521329708963 Average Test Variance:  0.8005977129653508\n",
      "\n",
      "\n",
      "Fraction:  0.7 Seed:  [378, 328, 332, 160, 254]\n",
      "train set - MAE: 10.205944858489579, RMSE: 14.41552437865324, R2: 0.879851160693591\n",
      "test set - MAE: 12.62370509184818, RMSE: 17.6751387536871, R2: 0.8054568449999195\n",
      "train set - MAE: 10.079961722736009, RMSE: 14.27936432540985, R2: 0.8827385855277105\n",
      "test set - MAE: 12.746924716984614, RMSE: 17.61164549145668, R2: 0.8068520242692835\n",
      "train set - MAE: 10.345359722914292, RMSE: 14.640216202199776, R2: 0.8768438298008252\n",
      "test set - MAE: 12.558603658466058, RMSE: 18.04111296665627, R2: 0.7973171777399735\n",
      "train set - MAE: 10.232111825369087, RMSE: 14.462014093065772, R2: 0.8794975870630543\n",
      "test set - MAE: 12.218318540712541, RMSE: 17.652186408995572, R2: 0.8059617714301286\n",
      "train set - MAE: 10.338804722402212, RMSE: 14.43975915092956, R2: 0.879933803598949\n",
      "test set - MAE: 11.943788256561065, RMSE: 17.32409380537444, R2: 0.8131077276441468\n",
      "Average Train MAE:  10.240436570382235 Average Train RMSE:  14.447375630051638 Average Train Variance:  0.8797729933368259\n",
      "Average Test MAE:  12.418268052914494 Average Test RMSE:  17.660835485234013 Average Test Variance:  0.8057391092166905\n",
      "\n",
      "\n",
      "Fraction:  0.8 Seed:  [246, 468, 201, 244, 119]\n",
      "train set - MAE: 9.95480248410183, RMSE: 13.99195852454664, R2: 0.886250314496984\n",
      "test set - MAE: 12.459682244555752, RMSE: 17.556791960972696, R2: 0.8080533145101476\n",
      "train set - MAE: 10.192572960234344, RMSE: 14.395009411916837, R2: 0.8803490952366668\n",
      "test set - MAE: 12.06402433035524, RMSE: 17.416746645430987, R2: 0.8111033047800447\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train set - MAE: 10.184902071272553, RMSE: 14.380102773556118, R2: 0.8825075112852274\n",
      "test set - MAE: 12.40997551595787, RMSE: 17.727388389018685, R2: 0.8043049634573747\n",
      "train set - MAE: 10.25076939666365, RMSE: 14.501848393555717, R2: 0.8779977146627419\n",
      "test set - MAE: 12.389000237470295, RMSE: 17.638428803351125, R2: 0.8062641091803111\n",
      "train set - MAE: 10.152461212819102, RMSE: 14.448255541860195, R2: 0.8806826994039432\n",
      "test set - MAE: 13.300605851499427, RMSE: 18.53566628439691, R2: 0.7860527567769269\n",
      "Average Train MAE:  10.147101625018298 Average Train RMSE:  14.3434349290871 Average Train Variance:  0.8815574670171127\n",
      "Average Test MAE:  12.524657635967717 Average Test RMSE:  17.77500441663408 Average Test Variance:  0.803155689740961\n",
      "\n",
      "\n",
      "Fraction:  0.9 Seed:  [76, 458, 402, 419, 444]\n",
      "train set - MAE: 10.039136037177187, RMSE: 14.222310891323824, R2: 0.8831103166436197\n",
      "test set - MAE: 12.705753804858784, RMSE: 17.95360891659039, R2: 0.7992785375319732\n",
      "train set - MAE: 10.17470729774247, RMSE: 14.359391409249964, R2: 0.8811813025012243\n",
      "test set - MAE: 12.4721811160426, RMSE: 17.482952825263567, R2: 0.8096644713898306\n",
      "train set - MAE: 10.295831049439387, RMSE: 14.56337957867901, R2: 0.8781928685649902\n",
      "test set - MAE: 13.063967980884515, RMSE: 18.242821880186405, R2: 0.7927596453927813\n",
      "train set - MAE: 10.010559234685319, RMSE: 14.21866849042508, R2: 0.8828023402894735\n",
      "test set - MAE: 13.029129621785152, RMSE: 18.602982849722274, R2: 0.7844959369891593\n",
      "train set - MAE: 10.402676589594163, RMSE: 14.710663581874178, R2: 0.8746015523317868\n",
      "test set - MAE: 12.129777600525912, RMSE: 16.87582666543039, R2: 0.8226544060193937\n",
      "Average Train MAE:  10.184582041727705 Average Train RMSE:  14.414882790310411 Average Train Variance:  0.8799776760662189\n",
      "Average Test MAE:  12.680162024819392 Average Test RMSE:  17.831638627438604 Average Test Variance:  0.8017705994646276\n",
      "\n",
      "\n",
      "Fraction:  1 Seed:  [384, 373, 471, 263, 145]\n",
      "train set - MAE: 10.095217822015956, RMSE: 14.369780555109934, R2: 0.8819356781952382\n",
      "test set - MAE: 12.854112301798969, RMSE: 18.32643551628246, R2: 0.7908555726219884\n",
      "train set - MAE: 10.13985143116981, RMSE: 14.38798644806682, R2: 0.8824917737679856\n",
      "test set - MAE: 12.452105636646216, RMSE: 18.341074876540986, R2: 0.7905213053836116\n",
      "train set - MAE: 10.133631082795462, RMSE: 14.34957990586985, R2: 0.8803006817660707\n",
      "test set - MAE: 12.530886610880682, RMSE: 18.18471231046943, R2: 0.7940778041684395\n",
      "train set - MAE: 10.518937104751881, RMSE: 14.808381726711822, R2: 0.8742401493354932\n",
      "test set - MAE: 12.531164653080697, RMSE: 17.405433911725673, R2: 0.8113486139700504\n",
      "train set - MAE: 10.108237315958968, RMSE: 14.341955155306593, R2: 0.8815422958687719\n",
      "test set - MAE: 13.401959262567717, RMSE: 18.840172807251314, R2: 0.7789655053065762\n",
      "Average Train MAE:  10.199174951338414 Average Train RMSE:  14.451536758213004 Average Train Variance:  0.8801021157867119\n",
      "Average Test MAE:  12.754045692994856 Average Test RMSE:  18.21956588445397 Average Test Variance:  0.7931537602901332\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import random\n",
    "bagging_reg = BaggingRegressor(estimator = tree_reg, n_estimators = 10, random_state = 42)\n",
    "\n",
    "for fraction in [.1,.2,.3,.4,.5,.6,.7,.8,.9,1]:\n",
    "    sumTrainMAE, sumTrainRMSE, sumTrainVariance, sumTestMAE, sumTestRMSE, sumTestVariance = \\\n",
    "    [],[],[],[],[],[]\n",
    "    seed = [random.randint(1, 500) for _ in range(5)]\n",
    "\n",
    "    print('Fraction: ', fraction, 'Seed: ', seed)\n",
    "    for x in range(5):\n",
    "        ###Setting fraction and seeds for each run\n",
    "        sub_x_train = df_x_train_scaled.sample(frac = fraction, random_state = seed[x], replace=True)#using scaled x_train\n",
    "        sub_y_train = y_train.sample(frac = fraction, random_state = seed[x], replace=True)\n",
    "\n",
    "        ### Bagging - decision tree regressor\n",
    "        bagging_reg.fit(sub_x_train, sub_y_train)\n",
    "\n",
    "        y_pred_train = bagging_reg.predict(sub_x_train)\n",
    "        evaluateTrainValue = evaluate(sub_y_train, y_pred_train, 'train')\n",
    "        y_pred_test = bagging_reg.predict(df_x_test_scaled)\n",
    "        evaluateTestValue = evaluate(y_test, y_pred_test, 'test')\n",
    "        \n",
    "        \n",
    "        ###Adding results from each run into list to be processed later\n",
    "        sumTrainMAE.append(evaluateTrainValue['mae'])\n",
    "        sumTrainRMSE.append(evaluateTrainValue['rmse'])\n",
    "        sumTrainVariance.append(evaluateTrainValue['r2'])\n",
    "        \n",
    "        sumTestMAE.append(evaluateTestValue['mae'])\n",
    "        sumTestRMSE.append(evaluateTestValue['rmse'])\n",
    "        sumTestVariance.append(evaluateTestValue['r2'])\n",
    "        \n",
    "    print(\"Average Train MAE: \", np.mean(sumTrainMAE), \"Average Train RMSE: \", np.mean(sumTrainRMSE), \\\n",
    "          \"Average Train Variance: \", np.mean(sumTrainVariance))\n",
    "    print(\"Average Test MAE: \", np.mean(sumTestMAE), \"Average Test RMSE: \", np.mean(sumTestRMSE), \\\n",
    "          \"Average Test Variance: \", np.mean(sumTestVariance))\n",
    "    \n",
    "    ###Finalizing and tabulate the results for each experiment\n",
    "    df_final.loc[len(df_final)] = [fraction, 'Decision Tree Regressor(Bagging)', \\\n",
    "                                   round(np.mean(sumTrainMAE),2), round(np.mean(sumTestMAE),2), \\\n",
    "                                   round(np.mean(sumTrainRMSE),2), round(np.mean(sumTestRMSE),2), \\\n",
    "                                   round(np.mean(sumTrainVariance),2), round(np.mean(sumTestVariance),2)]\n",
    "    \n",
    "    print('\\n')\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "bbbbc954",
   "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>fraction</th>\n",
       "      <th>model</th>\n",
       "      <th>MAE(Train)</th>\n",
       "      <th>MAE(Test)</th>\n",
       "      <th>RMSE(Train)</th>\n",
       "      <th>RMSE(Test)</th>\n",
       "      <th>R2(Train)</th>\n",
       "      <th>R2(Test)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.1</td>\n",
       "      <td>KNN Regressor</td>\n",
       "      <td>12.52</td>\n",
       "      <td>14.66</td>\n",
       "      <td>17.43</td>\n",
       "      <td>20.43</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.2</td>\n",
       "      <td>KNN Regressor</td>\n",
       "      <td>11.98</td>\n",
       "      <td>14.47</td>\n",
       "      <td>16.87</td>\n",
       "      <td>19.86</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.3</td>\n",
       "      <td>KNN Regressor</td>\n",
       "      <td>11.96</td>\n",
       "      <td>14.34</td>\n",
       "      <td>16.82</td>\n",
       "      <td>19.55</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.4</td>\n",
       "      <td>KNN Regressor</td>\n",
       "      <td>11.75</td>\n",
       "      <td>14.07</td>\n",
       "      <td>16.64</td>\n",
       "      <td>19.02</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.5</td>\n",
       "      <td>KNN Regressor</td>\n",
       "      <td>11.48</td>\n",
       "      <td>14.46</td>\n",
       "      <td>16.44</td>\n",
       "      <td>19.55</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.6</td>\n",
       "      <td>KNN Regressor</td>\n",
       "      <td>11.34</td>\n",
       "      <td>14.63</td>\n",
       "      <td>16.27</td>\n",
       "      <td>20.06</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.7</td>\n",
       "      <td>KNN Regressor</td>\n",
       "      <td>11.12</td>\n",
       "      <td>14.62</td>\n",
       "      <td>16.05</td>\n",
       "      <td>20.34</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.8</td>\n",
       "      <td>KNN Regressor</td>\n",
       "      <td>10.97</td>\n",
       "      <td>14.28</td>\n",
       "      <td>15.91</td>\n",
       "      <td>19.39</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.9</td>\n",
       "      <td>KNN Regressor</td>\n",
       "      <td>10.87</td>\n",
       "      <td>14.20</td>\n",
       "      <td>15.88</td>\n",
       "      <td>18.90</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.0</td>\n",
       "      <td>KNN Regressor</td>\n",
       "      <td>10.62</td>\n",
       "      <td>14.68</td>\n",
       "      <td>15.55</td>\n",
       "      <td>19.80</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.1</td>\n",
       "      <td>KNN Regressor(Feature Selection)</td>\n",
       "      <td>12.07</td>\n",
       "      <td>13.83</td>\n",
       "      <td>16.94</td>\n",
       "      <td>18.50</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.2</td>\n",
       "      <td>KNN Regressor(Feature Selection)</td>\n",
       "      <td>12.08</td>\n",
       "      <td>13.53</td>\n",
       "      <td>17.03</td>\n",
       "      <td>18.58</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.3</td>\n",
       "      <td>KNN Regressor(Feature Selection)</td>\n",
       "      <td>11.59</td>\n",
       "      <td>13.75</td>\n",
       "      <td>16.56</td>\n",
       "      <td>18.97</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.4</td>\n",
       "      <td>KNN Regressor(Feature Selection)</td>\n",
       "      <td>11.40</td>\n",
       "      <td>14.21</td>\n",
       "      <td>16.41</td>\n",
       "      <td>19.58</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.5</td>\n",
       "      <td>KNN Regressor(Feature Selection)</td>\n",
       "      <td>11.26</td>\n",
       "      <td>13.76</td>\n",
       "      <td>16.22</td>\n",
       "      <td>19.13</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.6</td>\n",
       "      <td>KNN Regressor(Feature Selection)</td>\n",
       "      <td>11.09</td>\n",
       "      <td>14.15</td>\n",
       "      <td>16.12</td>\n",
       "      <td>19.38</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.7</td>\n",
       "      <td>KNN Regressor(Feature Selection)</td>\n",
       "      <td>10.86</td>\n",
       "      <td>14.36</td>\n",
       "      <td>15.93</td>\n",
       "      <td>19.74</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.8</td>\n",
       "      <td>KNN Regressor(Feature Selection)</td>\n",
       "      <td>10.83</td>\n",
       "      <td>14.59</td>\n",
       "      <td>15.86</td>\n",
       "      <td>20.20</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.9</td>\n",
       "      <td>KNN Regressor(Feature Selection)</td>\n",
       "      <td>10.54</td>\n",
       "      <td>13.86</td>\n",
       "      <td>15.53</td>\n",
       "      <td>19.07</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1.0</td>\n",
       "      <td>KNN Regressor(Feature Selection)</td>\n",
       "      <td>10.48</td>\n",
       "      <td>14.64</td>\n",
       "      <td>15.50</td>\n",
       "      <td>20.33</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.1</td>\n",
       "      <td>Decision Tree Regressor(Bagging)</td>\n",
       "      <td>9.38</td>\n",
       "      <td>12.61</td>\n",
       "      <td>13.18</td>\n",
       "      <td>17.49</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.2</td>\n",
       "      <td>Decision Tree Regressor(Bagging)</td>\n",
       "      <td>9.59</td>\n",
       "      <td>12.79</td>\n",
       "      <td>13.47</td>\n",
       "      <td>17.77</td>\n",
       "      <td>0.90</td>\n",
       "      <td>0.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.3</td>\n",
       "      <td>Decision Tree Regressor(Bagging)</td>\n",
       "      <td>9.67</td>\n",
       "      <td>12.60</td>\n",
       "      <td>13.64</td>\n",
       "      <td>17.68</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.4</td>\n",
       "      <td>Decision Tree Regressor(Bagging)</td>\n",
       "      <td>9.82</td>\n",
       "      <td>12.56</td>\n",
       "      <td>13.90</td>\n",
       "      <td>17.90</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.5</td>\n",
       "      <td>Decision Tree Regressor(Bagging)</td>\n",
       "      <td>10.10</td>\n",
       "      <td>12.90</td>\n",
       "      <td>14.24</td>\n",
       "      <td>18.08</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.6</td>\n",
       "      <td>Decision Tree Regressor(Bagging)</td>\n",
       "      <td>10.06</td>\n",
       "      <td>12.79</td>\n",
       "      <td>14.22</td>\n",
       "      <td>17.89</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0.7</td>\n",
       "      <td>Decision Tree Regressor(Bagging)</td>\n",
       "      <td>10.24</td>\n",
       "      <td>12.42</td>\n",
       "      <td>14.45</td>\n",
       "      <td>17.66</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.8</td>\n",
       "      <td>Decision Tree Regressor(Bagging)</td>\n",
       "      <td>10.15</td>\n",
       "      <td>12.52</td>\n",
       "      <td>14.34</td>\n",
       "      <td>17.78</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0.9</td>\n",
       "      <td>Decision Tree Regressor(Bagging)</td>\n",
       "      <td>10.18</td>\n",
       "      <td>12.68</td>\n",
       "      <td>14.41</td>\n",
       "      <td>17.83</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1.0</td>\n",
       "      <td>Decision Tree Regressor(Bagging)</td>\n",
       "      <td>10.20</td>\n",
       "      <td>12.75</td>\n",
       "      <td>14.45</td>\n",
       "      <td>18.22</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.79</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    fraction                             model  MAE(Train)  MAE(Test)  \\\n",
       "0        0.1                     KNN Regressor       12.52      14.66   \n",
       "1        0.2                     KNN Regressor       11.98      14.47   \n",
       "2        0.3                     KNN Regressor       11.96      14.34   \n",
       "3        0.4                     KNN Regressor       11.75      14.07   \n",
       "4        0.5                     KNN Regressor       11.48      14.46   \n",
       "5        0.6                     KNN Regressor       11.34      14.63   \n",
       "6        0.7                     KNN Regressor       11.12      14.62   \n",
       "7        0.8                     KNN Regressor       10.97      14.28   \n",
       "8        0.9                     KNN Regressor       10.87      14.20   \n",
       "9        1.0                     KNN Regressor       10.62      14.68   \n",
       "10       0.1  KNN Regressor(Feature Selection)       12.07      13.83   \n",
       "11       0.2  KNN Regressor(Feature Selection)       12.08      13.53   \n",
       "12       0.3  KNN Regressor(Feature Selection)       11.59      13.75   \n",
       "13       0.4  KNN Regressor(Feature Selection)       11.40      14.21   \n",
       "14       0.5  KNN Regressor(Feature Selection)       11.26      13.76   \n",
       "15       0.6  KNN Regressor(Feature Selection)       11.09      14.15   \n",
       "16       0.7  KNN Regressor(Feature Selection)       10.86      14.36   \n",
       "17       0.8  KNN Regressor(Feature Selection)       10.83      14.59   \n",
       "18       0.9  KNN Regressor(Feature Selection)       10.54      13.86   \n",
       "19       1.0  KNN Regressor(Feature Selection)       10.48      14.64   \n",
       "20       0.1  Decision Tree Regressor(Bagging)        9.38      12.61   \n",
       "21       0.2  Decision Tree Regressor(Bagging)        9.59      12.79   \n",
       "22       0.3  Decision Tree Regressor(Bagging)        9.67      12.60   \n",
       "23       0.4  Decision Tree Regressor(Bagging)        9.82      12.56   \n",
       "24       0.5  Decision Tree Regressor(Bagging)       10.10      12.90   \n",
       "25       0.6  Decision Tree Regressor(Bagging)       10.06      12.79   \n",
       "26       0.7  Decision Tree Regressor(Bagging)       10.24      12.42   \n",
       "27       0.8  Decision Tree Regressor(Bagging)       10.15      12.52   \n",
       "28       0.9  Decision Tree Regressor(Bagging)       10.18      12.68   \n",
       "29       1.0  Decision Tree Regressor(Bagging)       10.20      12.75   \n",
       "\n",
       "    RMSE(Train)  RMSE(Test)  R2(Train)  R2(Test)  \n",
       "0         17.43       20.43       0.83      0.74  \n",
       "1         16.87       19.86       0.84      0.75  \n",
       "2         16.82       19.55       0.84      0.76  \n",
       "3         16.64       19.02       0.84      0.77  \n",
       "4         16.44       19.55       0.85      0.76  \n",
       "5         16.27       20.06       0.85      0.75  \n",
       "6         16.05       20.34       0.85      0.74  \n",
       "7         15.91       19.39       0.85      0.77  \n",
       "8         15.88       18.90       0.85      0.78  \n",
       "9         15.55       19.80       0.86      0.76  \n",
       "10        16.94       18.50       0.83      0.79  \n",
       "11        17.03       18.58       0.83      0.78  \n",
       "12        16.56       18.97       0.84      0.78  \n",
       "13        16.41       19.58       0.84      0.76  \n",
       "14        16.22       19.13       0.85      0.77  \n",
       "15        16.12       19.38       0.85      0.77  \n",
       "16        15.93       19.74       0.85      0.76  \n",
       "17        15.86       20.20       0.86      0.75  \n",
       "18        15.53       19.07       0.86      0.77  \n",
       "19        15.50       20.33       0.86      0.74  \n",
       "20        13.18       17.49       0.90      0.81  \n",
       "21        13.47       17.77       0.90      0.80  \n",
       "22        13.64       17.68       0.89      0.81  \n",
       "23        13.90       17.90       0.89      0.80  \n",
       "24        14.24       18.08       0.88      0.80  \n",
       "25        14.22       17.89       0.88      0.80  \n",
       "26        14.45       17.66       0.88      0.81  \n",
       "27        14.34       17.78       0.88      0.80  \n",
       "28        14.41       17.83       0.88      0.80  \n",
       "29        14.45       18.22       0.88      0.79  "
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_final"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "26c9f285",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_final.to_csv('Final Evaluation.csv', index = False)"
   ]
  }
 ],
 "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": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}