hyperopt.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import xgboost as xgb\n",
    "\n",
    "from hyperopt import fmin, hp, tpe, Trials, space_eval, STATUS_OK\n",
    "from hyperopt.pyll import scope as ho_scope\n",
    "from hyperopt.pyll.stochastic import sample as ho_sample\n",
    "\n",
    "from sklearn.model_selection import train_test_split \n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# pon los datos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test=train_test_split(X, y, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# espacio de hiperparametros\n",
    "\n",
    "espacio ={'n_estimators': hp.quniform('n_estimators', 10, 1000, 25),\n",
    "        'learning_rate': hp.uniform('learning_rate', 0.0001, 1.0),\n",
    "        'max_depth': hp.quniform('x_max_depth', 4, 16, 1),\n",
    "        'min_child_weight': hp.quniform ('x_min_child', 1, 10, 1),\n",
    "        'subsample': hp.uniform ('x_subsample', 0.7, 1),\n",
    "        'gamma' : hp.uniform ('x_gamma', 0.1,0.5),\n",
    "        'reg_lambda' : hp.uniform ('x_reg_lambda', 0,1)\n",
    "    }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# funcion objetivo a minimizar\n",
    "\n",
    "def objetivo(space):\n",
    "    #print(space)\n",
    "    clf = xgb.XGBRegressor(n_estimators =int(space['n_estimators']),\n",
    "                           learning_rate = space['learning_rate'],\n",
    "                           max_depth = int(space['max_depth']),\n",
    "                           min_child_weight = space['min_child_weight'],\n",
    "                           subsample = space['subsample'],\n",
    "                           gamma = space['gamma'],\n",
    "                           reg_lambda = space['reg_lambda'],\n",
    "                           objective='reg:squarederror')\n",
    "\n",
    "    eval_set=[(X_train, y_train), (X_test, y_test)]\n",
    "\n",
    "    clf.fit(X_train, y_train,\n",
    "            eval_set=eval_set, eval_metric=\"rmse\", verbose=False)\n",
    "\n",
    "    y_pred = clf.predict(X_test)\n",
    "    rmse = mean_squared_error(y_test, y_pred)**(0.5)\n",
    "    #print (\"SCORE:\", np.sqrt(mse_scr))\n",
    "    #change the metric if you like\n",
    "    return {'loss':rmse, 'status': STATUS_OK }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# minimiza\n",
    "\n",
    "trials_reg = Trials()\n",
    "best = fmin(fn=objetivo,\n",
    "            space=space,\n",
    "            algo=tpe.suggest,\n",
    "            max_evals=10,\n",
    "            trials=trials_reg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# mejores parametros\n",
    "\n",
    "print (best)"
   ]
  }
 ],
 "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": "3.7.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}