{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Search for Optimal Parameters of Regression Models\n",
    "\n",
    "Hyperparameter Tuning of *4.2 Decision Tree and 4.3 Random Forest* in [Main Notebook](https://github.com/czarinagluna/regression-analysis-for-estimating-prices/blob/main/main.ipynb)\n",
    "\n",
    "*** \n",
    "**By [Czarina Luna](https://www.linkedin.com/in/czarinagluna/)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%store -r X_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%store -r y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "def grid_search(regressor, params):\n",
    "    '''Performs grid search on given regression model, returns best cross validation scores and parameters'''\n",
    "    gridsearch = GridSearchCV(estimator=regressor, param_grid=params, cv=5)\n",
    "    gridsearch.fit(X_train, y_train)\n",
    "    \n",
    "    print(f'Best cv score: {gridsearch.best_score_ :.2%}')\n",
    "    print(f'Best parameters: {gridsearch.best_params_}')\n",
    "    \n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Decision Tree**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "%store -r dt_pipe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cv score: 73.24%\n",
      "Best parameters: {'decision_tree__criterion': 'friedman_mse', 'decision_tree__max_depth': None, 'decision_tree__min_samples_split': 5}\n"
     ]
    }
   ],
   "source": [
    "grid = {'decision_tree__criterion': ['squared_error', 'friedman_mse', 'absolute_error'], \n",
    "        'decision_tree__max_depth': [None, 1, 2], \n",
    "        'decision_tree__min_samples_split': [2, 3, 5]}\n",
    "\n",
    "grid_search(dt_pipe, grid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cv score: 76.86%\n",
      "Best parameters: {'decision_tree__criterion': 'friedman_mse', 'decision_tree__max_depth': None, 'decision_tree__min_samples_leaf': 5, 'decision_tree__min_samples_split': 5}\n"
     ]
    }
   ],
   "source": [
    "grid = {'decision_tree__criterion': ['friedman_mse'], \n",
    "        'decision_tree__max_depth': [None], \n",
    "        'decision_tree__min_samples_split': [5, 10], \n",
    "        'decision_tree__min_samples_leaf': [1, 2, 5]}\n",
    "\n",
    "grid_search(dt_pipe, grid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cv score: 78.44%\n",
      "Best parameters: {'decision_tree__criterion': 'friedman_mse', 'decision_tree__max_depth': None, 'decision_tree__min_samples_leaf': 5, 'decision_tree__min_samples_split': 20}\n"
     ]
    }
   ],
   "source": [
    "grid = {'decision_tree__criterion': ['friedman_mse'], \n",
    "        'decision_tree__max_depth': [None], \n",
    "        'decision_tree__min_samples_split': [10, 15, 20], \n",
    "        'decision_tree__min_samples_leaf': [5, 10]}\n",
    "\n",
    "grid_search(dt_pipe, grid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cv score: 78.95%\n",
      "Best parameters: {'decision_tree__criterion': 'friedman_mse', 'decision_tree__max_depth': None, 'decision_tree__min_samples_leaf': 5, 'decision_tree__min_samples_split': 30}\n"
     ]
    }
   ],
   "source": [
    "grid = {'decision_tree__criterion': ['friedman_mse'], \n",
    "        'decision_tree__max_depth': [None], \n",
    "        'decision_tree__min_samples_split': [20, 30, 50], \n",
    "        'decision_tree__min_samples_leaf': [5]}\n",
    "\n",
    "grid_search(dt_pipe, grid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cv score: 79.00%\n",
      "Best parameters: {'decision_tree__criterion': 'friedman_mse', 'decision_tree__max_depth': None, 'decision_tree__min_samples_leaf': 5, 'decision_tree__min_samples_split': 45}\n"
     ]
    }
   ],
   "source": [
    "grid = {'decision_tree__criterion': ['friedman_mse'], \n",
    "        'decision_tree__max_depth': [None], \n",
    "        'decision_tree__min_samples_split': [25, 30, 45], \n",
    "        'decision_tree__min_samples_leaf': [5]}\n",
    "\n",
    "grid_search(dt_pipe, grid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Random Forest**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "%store -r rf_pipe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cv score: 86.22%\n",
      "Best parameters: {'random_forest__criterion': 'friedman_mse', 'random_forest__max_depth': None, 'random_forest__min_samples_split': 2}\n"
     ]
    }
   ],
   "source": [
    "grid = {'random_forest__criterion': ['squared_error', 'friedman_mse', 'absolute_error'], \n",
    "        'random_forest__max_depth': [None, 1, 2], \n",
    "        'random_forest__min_samples_split': [2, 3, 5]}\n",
    "\n",
    "grid_search(rf_pipe, grid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cv score: 86.14%\n",
      "Best parameters: {'random_forest__criterion': 'friedman_mse', 'random_forest__max_depth': None, 'random_forest__min_samples_leaf': 1, 'random_forest__min_samples_split': 3}\n"
     ]
    }
   ],
   "source": [
    "grid = {'random_forest__criterion': ['friedman_mse'], \n",
    "        'random_forest__max_depth': [None], \n",
    "        'random_forest__min_samples_split': [2, 3, 10], \n",
    "        'random_forest__min_samples_leaf': [1, 2, 5]}\n",
    "\n",
    "grid_search(rf_pipe, grid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cv score: 86.25%\n",
      "Best parameters: {'random_forest__criterion': 'friedman_mse', 'random_forest__max_depth': None, 'random_forest__min_samples_leaf': 1, 'random_forest__min_samples_split': 2, 'random_forest__n_estimators': 150}\n"
     ]
    }
   ],
   "source": [
    "grid = {'random_forest__criterion': ['friedman_mse'], \n",
    "        'random_forest__max_depth': [None], \n",
    "        'random_forest__min_samples_split': [1,2,3], \n",
    "        'random_forest__min_samples_leaf': [1], \n",
    "        'random_forest__n_estimators': [100, 150, 200]}\n",
    "\n",
    "grid_search(rf_pipe, grid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best cv score: 86.29%\n",
      "Best parameters: {'random_forest__criterion': 'friedman_mse', 'random_forest__max_depth': None, 'random_forest__max_features': 'sqrt', 'random_forest__min_samples_leaf': 1, 'random_forest__min_samples_split': 2, 'random_forest__n_estimators': 150, 'random_forest__n_jobs': 2}\n"
     ]
    }
   ],
   "source": [
    "grid = {'random_forest__criterion': ['friedman_mse'], \n",
    "        'random_forest__max_depth': [None], \n",
    "        'random_forest__min_samples_split': [2], \n",
    "        'random_forest__min_samples_leaf': [1], \n",
    "        'random_forest__n_estimators': [150], \n",
    "        'random_forest__max_features': ['auto', 'sqrt', 'log2'], \n",
    "        'random_forest__n_jobs': [1, 2, 3]}\n",
    "\n",
    "grid_search(rf_pipe, grid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Contact\n",
    "\n",
    "Feel free to contact me for any questions and connect with me on [Linkedin](https://www.linkedin.com/in/czarinagluna/)."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}