fuerza_bruta.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "a320ecad",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "pd.set_option('display.max_columns', 100)\n",
    "sns.set_style()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e277f32f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>ult_actual</th>\n",
       "      <th>categoria</th>\n",
       "      <th>cupo_total_tc</th>\n",
       "      <th>ingreso_final</th>\n",
       "      <th>saldo_no_rot_mdo</th>\n",
       "      <th>cant_oblig_tot_sf</th>\n",
       "      <th>ind</th>\n",
       "      <th>gasto_familiar</th>\n",
       "      <th>cuotas_pagar</th>\n",
       "      <th>ingreso_poc</th>\n",
       "      <th>genre</th>\n",
       "      <th>pending_arrears</th>\n",
       "      <th>have_tc</th>\n",
       "      <th>age_cat_X</th>\n",
       "      <th>age_cat_Y</th>\n",
       "      <th>age_cat_Z</th>\n",
       "      <th>occupation_ama de casa</th>\n",
       "      <th>occupation_comerciante</th>\n",
       "      <th>occupation_desempleado con ingresos</th>\n",
       "      <th>occupation_desempleado sin ingresos</th>\n",
       "      <th>occupation_empleado</th>\n",
       "      <th>occupation_estudiante</th>\n",
       "      <th>occupation_ganadero</th>\n",
       "      <th>occupation_independiente</th>\n",
       "      <th>occupation_jubilado</th>\n",
       "      <th>occupation_pensionado</th>\n",
       "      <th>occupation_profesional independiente</th>\n",
       "      <th>occupation_rentista de capital</th>\n",
       "      <th>occupation_socio empleado - socio</th>\n",
       "      <th>civil_status_divorciado</th>\n",
       "      <th>civil_status_soltero</th>\n",
       "      <th>civil_status_union libre</th>\n",
       "      <th>civil_status_viudo</th>\n",
       "      <th>academic_level_especializacion</th>\n",
       "      <th>academic_level_ninguno</th>\n",
       "      <th>academic_level_primario</th>\n",
       "      <th>academic_level_tecnologo</th>\n",
       "      <th>academic_level_universitario</th>\n",
       "      <th>house_kind_familiar</th>\n",
       "      <th>house_kind_propia</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7</td>\n",
       "      <td>20190429</td>\n",
       "      <td>1</td>\n",
       "      <td>6000000.0</td>\n",
       "      <td>2629942.30</td>\n",
       "      <td>21423000.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1006462.495</td>\n",
       "      <td>84040.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8</td>\n",
       "      <td>20180628</td>\n",
       "      <td>1</td>\n",
       "      <td>6000000.0</td>\n",
       "      <td>2660041.28</td>\n",
       "      <td>21949000.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1417026.832</td>\n",
       "      <td>240906.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>9</td>\n",
       "      <td>20190429</td>\n",
       "      <td>1</td>\n",
       "      <td>6000000.0</td>\n",
       "      <td>2599610.30</td>\n",
       "      <td>22289000.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1196746.695</td>\n",
       "      <td>376766.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>20200618</td>\n",
       "      <td>2</td>\n",
       "      <td>6000000.0</td>\n",
       "      <td>3312905.18</td>\n",
       "      <td>20636000.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>512597.867</td>\n",
       "      <td>273565.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11</td>\n",
       "      <td>20200618</td>\n",
       "      <td>2</td>\n",
       "      <td>6000000.0</td>\n",
       "      <td>3312905.18</td>\n",
       "      <td>20636000.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>512438.857</td>\n",
       "      <td>150000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  ult_actual  categoria  cupo_total_tc  ingreso_final  \\\n",
       "0           7    20190429          1      6000000.0     2629942.30   \n",
       "1           8    20180628          1      6000000.0     2660041.28   \n",
       "2           9    20190429          1      6000000.0     2599610.30   \n",
       "3          10    20200618          2      6000000.0     3312905.18   \n",
       "4          11    20200618          2      6000000.0     3312905.18   \n",
       "\n",
       "   saldo_no_rot_mdo  cant_oblig_tot_sf          ind  gasto_familiar  \\\n",
       "0        21423000.0               10.0  1006462.495         84040.0   \n",
       "1        21949000.0                6.0  1417026.832        240906.0   \n",
       "2        22289000.0                9.0  1196746.695        376766.0   \n",
       "3        20636000.0               12.0   512597.867        273565.0   \n",
       "4        20636000.0               12.0   512438.857        150000.0   \n",
       "\n",
       "   cuotas_pagar  ingreso_poc  genre  pending_arrears  have_tc  age_cat_X  \\\n",
       "0           0.0          0.0      1                0        1          0   \n",
       "1           0.0          0.0      1                0        1          0   \n",
       "2           0.0          0.0      1                0        1          0   \n",
       "3           0.0          0.0      1                1        1          0   \n",
       "4           0.0          0.0      1                0        1          0   \n",
       "\n",
       "   age_cat_Y  age_cat_Z  occupation_ama de casa  occupation_comerciante  \\\n",
       "0          1          0                       0                       0   \n",
       "1          1          0                       0                       0   \n",
       "2          1          0                       0                       0   \n",
       "3          1          0                       0                       0   \n",
       "4          1          0                       0                       0   \n",
       "\n",
       "   occupation_desempleado con ingresos  occupation_desempleado sin ingresos  \\\n",
       "0                                    0                                    0   \n",
       "1                                    0                                    0   \n",
       "2                                    0                                    0   \n",
       "3                                    0                                    0   \n",
       "4                                    0                                    0   \n",
       "\n",
       "   occupation_empleado  occupation_estudiante  occupation_ganadero  \\\n",
       "0                    1                      0                    0   \n",
       "1                    1                      0                    0   \n",
       "2                    1                      0                    0   \n",
       "3                    1                      0                    0   \n",
       "4                    1                      0                    0   \n",
       "\n",
       "   occupation_independiente  occupation_jubilado  occupation_pensionado  \\\n",
       "0                         0                    0                      0   \n",
       "1                         0                    0                      0   \n",
       "2                         0                    0                      0   \n",
       "3                         0                    0                      0   \n",
       "4                         0                    0                      0   \n",
       "\n",
       "   occupation_profesional independiente  occupation_rentista de capital  \\\n",
       "0                                     0                               0   \n",
       "1                                     0                               0   \n",
       "2                                     0                               0   \n",
       "3                                     0                               0   \n",
       "4                                     0                               0   \n",
       "\n",
       "   occupation_socio empleado - socio  civil_status_divorciado  \\\n",
       "0                                  0                        0   \n",
       "1                                  0                        0   \n",
       "2                                  0                        0   \n",
       "3                                  0                        0   \n",
       "4                                  0                        0   \n",
       "\n",
       "   civil_status_soltero  civil_status_union libre  civil_status_viudo  \\\n",
       "0                     1                         0                   0   \n",
       "1                     1                         0                   0   \n",
       "2                     1                         0                   0   \n",
       "3                     1                         0                   0   \n",
       "4                     1                         0                   0   \n",
       "\n",
       "   academic_level_especializacion  academic_level_ninguno  \\\n",
       "0                               0                       0   \n",
       "1                               0                       0   \n",
       "2                               0                       0   \n",
       "3                               0                       0   \n",
       "4                               0                       0   \n",
       "\n",
       "   academic_level_primario  academic_level_tecnologo  \\\n",
       "0                        0                         1   \n",
       "1                        0                         1   \n",
       "2                        0                         1   \n",
       "3                        0                         1   \n",
       "4                        0                         1   \n",
       "\n",
       "   academic_level_universitario  house_kind_familiar  house_kind_propia  \n",
       "0                             0                    1                  0  \n",
       "1                             0                    1                  0  \n",
       "2                             0                    1                  0  \n",
       "3                             0                    1                  0  \n",
       "4                             0                    1                  0  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = ['ult_actual', 'categoria', 'cupo_total_tc', 'ingreso_final', 'saldo_no_rot_mdo',\n",
    "        'cant_oblig_tot_sf', 'ind', 'gasto_familiar', 'cuotas_pagar', 'ingreso_poc',\n",
    "        'age_cat', 'genre', 'occupation', 'civil_status', 'academic_level', 'house_kind',\n",
    "        'pending_arrears', 'have_tc']\n",
    "    \n",
    "df = pd.read_csv(\"dataF.csv\", sep = \",\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0b9e2dc6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(830857, 41)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "bfd56dc4",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.dropna(subset = [\"categoria\"])\n",
    "df[\"categoria\"] = df[\"categoria\"].astype(np.int64).astype(object)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "921a2cdd",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.fillna(0)\n",
    "del df[\"Unnamed: 0\"]\n",
    "del df[\"ult_actual\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "f45ac4c9-5e97-4c80-b183-a3ae8e1e7c02",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['categoria', 'cupo_total_tc', 'ingreso_final', 'saldo_no_rot_mdo',\n",
       "       'cant_oblig_tot_sf', 'ind', 'gasto_familiar', 'cuotas_pagar',\n",
       "       'ingreso_poc', 'genre', 'pending_arrears', 'have_tc', 'age_cat_X',\n",
       "       'age_cat_Y', 'age_cat_Z', 'occupation_ama de casa',\n",
       "       'occupation_comerciante', 'occupation_desempleado con ingresos',\n",
       "       'occupation_desempleado sin ingresos', 'occupation_empleado',\n",
       "       'occupation_estudiante', 'occupation_ganadero',\n",
       "       'occupation_independiente', 'occupation_jubilado',\n",
       "       'occupation_pensionado', 'occupation_profesional independiente',\n",
       "       'occupation_rentista de capital', 'occupation_socio empleado - socio',\n",
       "       'civil_status_divorciado', 'civil_status_soltero',\n",
       "       'civil_status_union libre', 'civil_status_viudo',\n",
       "       'academic_level_especializacion', 'academic_level_ninguno',\n",
       "       'academic_level_primario', 'academic_level_tecnologo',\n",
       "       'academic_level_universitario', 'house_kind_familiar',\n",
       "       'house_kind_propia'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7b8063e4",
   "metadata": {},
   "source": [
    "## Variables númericas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "ff998fcb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df[\"ingreso_final\"].hist();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "24d78025",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df[\"ind\"].hist();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "ae387633",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df[\"gasto_familiar\"].hist();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "207f37c6",
   "metadata": {},
   "source": [
    "### Eliminación de outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "9a8d09b8",
   "metadata": {},
   "outputs": [],
   "source": [
    "max_gf = df.gasto_familiar.quantile(0.95)\n",
    "min_gf = df.gasto_familiar.quantile(0.05)\n",
    "max_if = df.ingreso_final.quantile(0.95)\n",
    "min_if = df.ingreso_final.quantile(0.05)\n",
    "max_ind = df.ind.quantile(0.95)\n",
    "min_ind = df.ind.quantile(0.05)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "b0935dc1",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[(df['ingreso_final'] > min_if) & (df['ingreso_final']< max_if)]\n",
    "df = df[(df['ind'] > min_ind) & (df['ind']< max_ind)]\n",
    "df = df[(df['gasto_familiar'] > min_gf) & (df['gasto_familiar']< max_gf)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "913e8459",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df[\"ingreso_final\"].hist();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "aaba7557",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df[\"ind\"].hist();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "4117efbb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEFCAYAAAD9mKAdAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAAYvElEQVR4nO3df6zd9X3f8ecrmFIrLQmEcEcxm9PhtOOH0gbLoYtU3dYVsGUa0QSrK1acyps1RKtWYltN/xhqIibyR5oV1GSzihuT0RKP/sBKQplLctV24neb1AHC8IIVPLywxg7F3UJj9N4f53PL4eb4c4/vL9/r+3xIV+ec9/fz+Z73+dry635/nK9TVUiSdCJvOdUNSJKWN4NCktRlUEiSugwKSVKXQSFJ6lpzqhtYaOedd16tX79+zvP/+q//mre+9a0L19Ais9/FZb+Lb6X1fLr2+9RTT/1lVb1z5MKqOq1+rrjiipqPL37xi/Oav9Tsd3HZ7+JbaT2frv0CT9YJ/l310JMkqcugkCR1GRSSpC6DQpLUZVBIkroMCklSl0EhSeoyKCRJXQaFJKnrtLuFx3zt/1+v8KEdn1vy9z14xweW/D0laRzuUUiSugwKSVKXQSFJ6jIoJEldBoUkqcugkCR1GRSSpC6DQpLUZVBIkrrGCookb09yf5KvJnk2yY8lOTfJviTPt8dzhsbfmuRAkueSXD1UvyLJ/rbsziRp9bOSfKbVH0uyfmjO1vYezyfZuoCfXZI0hnH3KH4d+MOq+mHgPcCzwA7g4araADzcXpPkEmALcClwDfCJJGe09XwS2A5saD/XtPo24GhVXQx8HPhoW9e5wG3A+4BNwG3DgSRJWnyzBkWSs4EfB+4GqKq/qapvAdcCu9uw3cAH2/Nrgfuq6rWqegE4AGxKcgFwdlU9UlUF3DNjzvS67gc2t72Nq4F9VXWkqo4C+3gjXCRJS2CcPYofBP4P8FtJ/jzJbyZ5KzBRVYcB2uP5bfyFwItD8w+12oXt+cz6m+ZU1XHgFeAdnXVJkpbIOHePXQO8F/iFqnosya/TDjOdQEbUqlOf65w33jDZzuCQFhMTE0xNTXXa65tYC7dcfnzO8+dqrj0fO3ZsXp93qdnv4lpp/cLK63k19jtOUBwCDlXVY+31/QyC4htJLqiqw+2w0stD4y8amr8OeKnV142oD885lGQN8DbgSKtPzpgzNbPBqtoJ7ATYuHFjTU5OzhwytrvufYCP7V/6u68fvGFyTvOmpqaYz+ddava7uFZav7Dyel6N/c566Kmq/jfwYpIfaqXNwDPAXmD6KqStwAPt+V5gS7uS6V0MTlo/3g5PvZrkynb+4cYZc6bXdR3whXYe4yHgqiTntJPYV7WaJGmJjPur8y8A9yb5HuBrwM8xCJk9SbYBXweuB6iqp5PsYRAmx4Gbq+r1tp6bgE8Ba4EH2w8MTpR/OskBBnsSW9q6jiT5CPBEG/fhqjoyx88qSZqDsYKiqr4EbByxaPMJxt8O3D6i/iRw2Yj6t2lBM2LZLmDXOH1Kkhae38yWJHUZFJKkLoNCktRlUEiSugwKSVKXQSFJ6jIoJEldBoUkqcugkCR1GRSSpC6DQpLUZVBIkroMCklSl0EhSeoyKCRJXQaFJKnLoJAkdRkUkqQug0KS1GVQSJK6DApJUpdBIUnqMigkSV0GhSSpa6ygSHIwyf4kX0ryZKudm2Rfkufb4zlD429NciDJc0muHqpf0dZzIMmdSdLqZyX5TKs/lmT90Jyt7T2eT7J1wT65JGksa05i7E9U1V8Ovd4BPFxVdyTZ0V7/cpJLgC3ApcAPAH+U5N1V9TrwSWA78CjweeAa4EFgG3C0qi5OsgX4KPDTSc4FbgM2AgU8lWRvVR2dx2deltbv+Nyc5t1y+XE+NMe50w7e8YF5zZd0epvPoadrgd3t+W7gg0P1+6rqtap6ATgAbEpyAXB2VT1SVQXcM2PO9LruBza3vY2rgX1VdaSFwz4G4SJJWiLj7lEU8N+SFPCfq2onMFFVhwGq6nCS89vYCxnsMUw71Grfac9n1qfnvNjWdTzJK8A7husj5vytJNsZ7KkwMTHB1NTUmB/ru02sHfyWvlIsRL/z2V4n69ixY0v6fvNlv4tvpfW8GvsdNyjeX1UvtTDYl+SrnbEZUatOfa5z3igMgmsnwMaNG2tycrLTXt9d9z7Ax/afzBG5U+uWy4/Pu9+DN0wuTDNjmJqaYj5/PkvNfhffSut5NfY71qGnqnqpPb4M/D6wCfhGO5xEe3y5DT8EXDQ0fR3wUquvG1F/05wka4C3AUc665IkLZFZgyLJW5N8//Rz4CrgK8BeYPoqpK3AA+35XmBLu5LpXcAG4PF2mOrVJFe28w83zpgzva7rgC+08xgPAVclOaddVXVVq0mSlsg4xywmgN9vV7KuAX67qv4wyRPAniTbgK8D1wNU1dNJ9gDPAMeBm9sVTwA3AZ8C1jK42unBVr8b+HSSAwz2JLa0dR1J8hHgiTbuw1V1ZB6fV5J0kmYNiqr6GvCeEfVvAptPMOd24PYR9SeBy0bUv00LmhHLdgG7ZutTkrQ4/Ga2JKnLoJAkdRkUkqQug0KS1GVQSJK6DApJUpdBIUnqMigkSV0GhSSpy6CQJHUZFJKkLoNCktRlUEiSugwKSVKXQSFJ6jIoJEldBoUkqcugkCR1GRSSpC6DQpLUZVBIkroMCklSl0EhSeoaOyiSnJHkz5N8tr0+N8m+JM+3x3OGxt6a5ECS55JcPVS/Isn+tuzOJGn1s5J8ptUfS7J+aM7W9h7PJ9m6IJ9akjS2k9mj+EXg2aHXO4CHq2oD8HB7TZJLgC3ApcA1wCeSnNHmfBLYDmxoP9e0+jbgaFVdDHwc+Ghb17nAbcD7gE3AbcOBJElafGMFRZJ1wAeA3xwqXwvsbs93Ax8cqt9XVa9V1QvAAWBTkguAs6vqkaoq4J4Zc6bXdT+wue1tXA3sq6ojVXUU2Mcb4SJJWgJrxhz3H4F/B3z/UG2iqg4DVNXhJOe3+oXAo0PjDrXad9rzmfXpOS+2dR1P8grwjuH6iDl/K8l2BnsqTExMMDU1NebH+m4Ta+GWy4/Pef5SW4h+57O9TtaxY8eW9P3my34X30rreTX2O2tQJPknwMtV9VSSyTHWmRG16tTnOueNQtVOYCfAxo0ba3JynDZHu+veB/jY/nHz89S75fLj8+734A2TC9PMGKamppjPn89Ss9/Ft9J6Xo39jnPo6f3AP01yELgP+Mkk/wX4RjucRHt8uY0/BFw0NH8d8FKrrxtRf9OcJGuAtwFHOuuSJC2RWYOiqm6tqnVVtZ7BSeovVNW/APYC01chbQUeaM/3AlvalUzvYnDS+vF2mOrVJFe28w83zpgzva7r2nsU8BBwVZJz2knsq1pNkrRE5nPM4g5gT5JtwNeB6wGq6ukke4BngOPAzVX1eptzE/ApYC3wYPsBuBv4dJIDDPYktrR1HUnyEeCJNu7DVXVkHj1Lkk7SSQVFVU0BU+35N4HNJxh3O3D7iPqTwGUj6t+mBc2IZbuAXSfTpyRp4fjNbElSl0EhSeoyKCRJXQaFJKnLoJAkdRkUkqQug0KS1GVQSJK6DApJUpdBIUnqMigkSV0GhSSpy6CQJHUZFJKkLoNCktRlUEiSugwKSVKXQSFJ6jIoJEldBoUkqcugkCR1GRSSpC6DQpLUNWtQJPneJI8n+XKSp5P8aqufm2Rfkufb4zlDc25NciDJc0muHqpfkWR/W3ZnkrT6WUk+0+qPJVk/NGdre4/nk2xd0E8vSZrVmjHGvAb8ZFUdS3Im8KdJHgT+GfBwVd2RZAewA/jlJJcAW4BLgR8A/ijJu6vqdeCTwHbgUeDzwDXAg8A24GhVXZxkC/BR4KeTnAvcBmwECngqyd6qOrpgW0Cs3/G5JXuvWy4/zofa+x284wNL9r6S5m7WPYoaONZentl+CrgW2N3qu4EPtufXAvdV1WtV9QJwANiU5ALg7Kp6pKoKuGfGnOl13Q9sbnsbVwP7qupIC4d9DMJFkrRExtmjIMkZwFPAxcBvVNVjSSaq6jBAVR1Ocn4bfiGDPYZph1rtO+35zPr0nBfbuo4neQV4x3B9xJzh/rYz2FNhYmKCqampcT7WSBNrB7/1rhQrud/5/DktlWPHjq2IPqettH5h5fW8GvsdKyjaYaMfSfJ24PeTXNYZnlGr6NTnOme4v53AToCNGzfW5ORkp72+u+59gI/tH2uzLAu3XH58xfZ78IbJU9vMGKamppjP36elttL6hZXX82rs96SueqqqbwFTDA7/fKMdTqI9vtyGHQIuGpq2Dnip1deNqL9pTpI1wNuAI511SZKWyDhXPb2z7UmQZC3wU8BXgb3A9FVIW4EH2vO9wJZ2JdO7gA3A4+0w1atJrmznH26cMWd6XdcBX2jnMR4CrkpyTruq6qpWkyQtkXGOWVwA7G7nKd4C7KmqzyZ5BNiTZBvwdeB6gKp6Oske4BngOHBzO3QFcBPwKWAtg6udHmz1u4FPJznAYE9iS1vXkSQfAZ5o4z5cVUfm84ElSSdn1qCoqr8AfnRE/ZvA5hPMuR24fUT9SeC7zm9U1bdpQTNi2S5g12x9SpIWh9/MliR1GRSSpC6DQpLUZVBIkroMCklSl0EhSeoyKCRJXQaFJKnLoJAkdRkUkqQug0KS1GVQSJK6DApJUpdBIUnqMigkSV0GhSSpy6CQJHUZFJKkLoNCktRlUEiSugwKSVKXQSFJ6jIoJEldswZFkouSfDHJs0meTvKLrX5ukn1Jnm+P5wzNuTXJgSTPJbl6qH5Fkv1t2Z1J0upnJflMqz+WZP3QnK3tPZ5PsnVBP70kaVbj7FEcB26pqn8AXAncnOQSYAfwcFVtAB5ur2nLtgCXAtcAn0hyRlvXJ4HtwIb2c02rbwOOVtXFwMeBj7Z1nQvcBrwP2ATcNhxIkqTFN2tQVNXhqvqz9vxV4FngQuBaYHcbthv4YHt+LXBfVb1WVS8AB4BNSS4Azq6qR6qqgHtmzJle1/3A5ra3cTWwr6qOVNVRYB9vhIskaQmsOZnB7ZDQjwKPARNVdRgGYZLk/DbsQuDRoWmHWu077fnM+vScF9u6jid5BXjHcH3EnOG+tjPYU2FiYoKpqamT+VhvMrEWbrn8+JznL7WV3O98/pyWyrFjx1ZEn9NWWr+w8npejf2OHRRJvg/4XeCXquqv2umFkUNH1KpTn+ucNwpVO4GdABs3bqzJyckT9Taru+59gI/tP6n8PKVuufz4iu334A2Tp7aZMUxNTTGfv09LbaX1Cyuv59XY71hXPSU5k0FI3FtVv9fK32iHk2iPL7f6IeCioenrgJdafd2I+pvmJFkDvA040lmXJGmJzPqraDtXcDfwbFX92tCivcBW4I72+MBQ/beT/BrwAwxOWj9eVa8neTXJlQwOXd0I3DVjXY8A1wFfqKpK8hDwH4ZOYF8F3DrnT6tlZf2Oz52S9z14xwdOyftKK9U4xyzeD/wssD/Jl1rtVxgExJ4k24CvA9cDVNXTSfYAzzC4Yurmqnq9zbsJ+BSwFniw/cAgiD6d5ACDPYktbV1HknwEeKKN+3BVHZnbR5UkzcWsQVFVf8rocwUAm08w53bg9hH1J4HLRtS/TQuaEct2Abtm61OStDj8ZrYkqcugkCR1GRSSpC6DQpLUZVBIkroMCklSl0EhSeoyKCRJXQaFJKnLoJAkdRkUkqQug0KS1GVQSJK6DApJUpdBIUnqMigkSV0GhSSpy6CQJHUZFJKkrln/z2zpdLN+x+fGHnvL5cf50EmMn83BOz6wYOuSlop7FJKkLoNCktRlUEiSumYNiiS7kryc5CtDtXOT7EvyfHs8Z2jZrUkOJHkuydVD9SuS7G/L7kySVj8ryWda/bEk64fmbG3v8XySrQv2qSVJYxtnj+JTwDUzajuAh6tqA/Bwe02SS4AtwKVtzieSnNHmfBLYDmxoP9Pr3AYcraqLgY8DH23rOhe4DXgfsAm4bTiQJElLY9agqKo/Bo7MKF8L7G7PdwMfHKrfV1WvVdULwAFgU5ILgLOr6pGqKuCeGXOm13U/sLntbVwN7KuqI1V1FNjHdweWJGmRzfUcxURVHQZoj+e3+oXAi0PjDrXahe35zPqb5lTVceAV4B2ddUmSltBCf48iI2rVqc91zpvfNNnO4LAWExMTTE1NzdroiUysHVw7v1LY7+Ja6H7n83dzHMeOHVv091hoK63n1djvXIPiG0kuqKrD7bDSy61+CLhoaNw64KVWXzeiPjznUJI1wNsYHOo6BEzOmDM1qpmq2gnsBNi4cWNNTk6OGjaWu+59gI/tXznfQ7zl8uP2u4gWut+DN0wu2LpGmZqaYj5//0+Fldbzaux3roee9gLTVyFtBR4Yqm9pVzK9i8FJ68fb4alXk1zZzj/cOGPO9LquA77QzmM8BFyV5Jx2EvuqVpMkLaFZf1VK8jsMfrM/L8khBlci3QHsSbIN+DpwPUBVPZ1kD/AMcBy4uapeb6u6icEVVGuBB9sPwN3Ap5McYLAnsaWt60iSjwBPtHEfrqqZJ9UlSYts1qCoqp85waLNJxh/O3D7iPqTwGUj6t+mBc2IZbuAXbP1KElaPH4zW5LUtXLOKkqngZO5c+1cnOhut961VvPhHoUkqcugkCR1GRSSpC6DQpLUZVBIkroMCklSl0EhSeryexTSKrDY39/o8TscK597FJKkLoNCktRlUEiSugwKSVKXJ7MlLarZTqSf6EaG8+VJ9IXjHoUkqcugkCR1eehJ0mlpsb47Ms6hstPtsJd7FJKkLoNCktTloSdJWmCn6pYpi3XIyz0KSVKXQSFJ6jIoJEldKyIoklyT5LkkB5LsONX9SNJqsuyDIskZwG8A/wi4BPiZJJec2q4kafVY9kEBbAIOVNXXqupvgPuAa09xT5K0aqSqTnUPXUmuA66pqn/ZXv8s8L6q+vmhMduB7e3lDwHPzeMtzwP+ch7zl5r9Li77XXwrrefTtd+/V1XvHLVgJXyPIiNqb0q3qtoJ7FyQN0uerKqNC7GupWC/i8t+F99K63k19rsSDj0dAi4aer0OeOkU9SJJq85KCIongA1J3pXke4AtwN5T3JMkrRrL/tBTVR1P8vPAQ8AZwK6qenoR33JBDmEtIftdXPa7+FZaz6uu32V/MluSdGqthENPkqRTyKCQJHWtyqCY7ZYgGbizLf+LJO89FX0O9TNbv5NJXknypfbz709Fn0P97ErycpKvnGD5ctu+s/W73LbvRUm+mOTZJE8n+cURY5bNNh6z32WzjZN8b5LHk3y59furI8Ysm+3b+hmn57lv46paVT8MToj/T+AHge8BvgxcMmPMPwYeZPAdjiuBx5Z5v5PAZ0/1th3q58eB9wJfOcHyZbN9x+x3uW3fC4D3tuffD/yPZf53eJx+l802btvs+9rzM4HHgCuX6/Y9iZ7nvI1X4x7FOLcEuRa4pwYeBd6e5IKlbrRZcbcwqao/Bo50hiyn7TtOv8tKVR2uqj9rz18FngUunDFs2WzjMftdNto2O9Zentl+Zl71s2y2L4zd85ytxqC4EHhx6PUhvvsv7Thjlsq4vfxY2+18MMmlS9PanC2n7TuuZbl9k6wHfpTBb5DDluU27vQLy2gbJzkjyZeAl4F9VbXst+8YPcMct/FqDIpZbwky5pilMk4vf8bgPi3vAe4C/mCxm5qn5bR9x7Est2+S7wN+F/ilqvqrmYtHTDml23iWfpfVNq6q16vqRxjcCWJTkstmDFl223eMnue8jVdjUIxzS5DldNuQWXupqr+a3u2sqs8DZyY5b+laPGnLafvOajlu3yRnMvhH996q+r0RQ5bVNp6t3+W4jVsv3wKmgGtmLFpW23fYiXqezzZejUExzi1B9gI3tisbrgReqarDS91oM2u/Sf5OkrTnmxj8uX5zyTsd33LavrNabtu39XI38GxV/doJhi2bbTxOv8tpGyd5Z5K3t+drgZ8Cvjpj2LLZvjBez/PZxsv+Fh4LrU5wS5Ak/7ot/0/A5xlc1XAA+L/Azy3zfq8DbkpyHPh/wJZqlzmcCkl+h8EVFuclOQTcxuDk2rLbvjBWv8tq+wLvB34W2N+OSQP8CvB3YVlu43H6XU7b+AJgdwb/adpbgD1V9dnl+m9EM07Pc97G3sJDktS1Gg89SZJOgkEhSeoyKCRJXQaFJKnLoJCkFS6z3NhyxPh/nuSZdgPB3551vFc9SdLKluTHgWMM7j818xvZM8duAPYAP1lVR5OcX1Uv9+a4RyFJK9yoG1sm+ftJ/jDJU0n+JMkPt0X/CviNqjra5nZDAgwKSTpd7QR+oaquAP4N8IlWfzfw7iT/PcmjSWbenuS7rLpvZkvS6a7dgPEfAv+13bUD4Kz2uAbYwOBuBOuAP0lyWbtH1EgGhSSdft4CfKvdTXamQ8CjVfUd4IUkzzEIjid6K5MknUbabdxfSHI9/O1/3fqetvgPgJ9o9fMYHIr6Wm99BoUkrXDtxpaPAD+U5FCSbcANwLYkXwae5o3/GfMh4JtJngG+CPzbqureRdbLYyVJXe5RSJK6DApJUpdBIUnqMigkSV0GhSSpy6CQJHUZFJKkrv8PkEkp4uyepJsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df[\"gasto_familiar\"].hist();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "da42355e",
   "metadata": {},
   "source": [
    "### One Hot Encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "39a7a80e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['age_cat', 'occupation', 'civil_status', 'academic_level',\n",
       "       'house_kind'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "categorical = df.select_dtypes(include=['object'])\n",
    "names = categorical.columns\n",
    "names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "adc561b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.get_dummies(df, prefix=names)\n",
    "df = df.drop([\"age_cat_BB\", \"civil_status_casado\", \"academic_level_bachiller\", \n",
    "              \"occupation_agricultor\", \"house_kind_alquilada\"], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "99e7146a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1999652, 40)"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "d9da3e8c",
   "metadata": {},
   "outputs": [],
   "source": [
    "del df[\"ult_actual\"]\n",
    "#df.to_csv(\"dataF.csv\", sep = \",\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "b1608f9f-fc84-4777-bfdc-e771defe44e4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "categoria                                 int64\n",
       "cupo_total_tc                           float64\n",
       "ingreso_final                           float64\n",
       "saldo_no_rot_mdo                        float64\n",
       "cant_oblig_tot_sf                       float64\n",
       "ind                                     float64\n",
       "gasto_familiar                          float64\n",
       "cuotas_pagar                            float64\n",
       "ingreso_poc                             float64\n",
       "genre                                     int64\n",
       "pending_arrears                           int64\n",
       "have_tc                                   int64\n",
       "age_cat_X                                 int64\n",
       "age_cat_Y                                 int64\n",
       "age_cat_Z                                 int64\n",
       "occupation_ama de casa                    int64\n",
       "occupation_comerciante                    int64\n",
       "occupation_desempleado con ingresos       int64\n",
       "occupation_desempleado sin ingresos       int64\n",
       "occupation_empleado                       int64\n",
       "occupation_estudiante                     int64\n",
       "occupation_ganadero                       int64\n",
       "occupation_independiente                  int64\n",
       "occupation_jubilado                       int64\n",
       "occupation_pensionado                     int64\n",
       "occupation_profesional independiente      int64\n",
       "occupation_rentista de capital            int64\n",
       "occupation_socio empleado - socio         int64\n",
       "civil_status_divorciado                   int64\n",
       "civil_status_soltero                      int64\n",
       "civil_status_union libre                  int64\n",
       "civil_status_viudo                        int64\n",
       "academic_level_especializacion            int64\n",
       "academic_level_ninguno                    int64\n",
       "academic_level_primario                   int64\n",
       "academic_level_tecnologo                  int64\n",
       "academic_level_universitario              int64\n",
       "house_kind_familiar                       int64\n",
       "house_kind_propia                         int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "1b8f68f9-0090-4489-bcc2-f5d5aefb5a6a",
   "metadata": {},
   "outputs": [],
   "source": [
    "corr = df.select_dtypes(include=\"float64\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "7e7d760e-cb59-4bac-8694-3e59f2d2eba2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (15,10))\n",
    "corrMatrix = corr.corr()\n",
    "swarm_plot = sns.heatmap(corrMatrix, annot=True)\n",
    "fig = swarm_plot.get_figure()\n",
    "fig.savefig(\"correlograma.png\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b86a224-6c10-4368-906b-3f1eec6e7e25",
   "metadata": {},
   "source": [
    "### Modelos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "6c96dbbb-c1aa-466c-b332-dda14b844822",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.drop([\"gasto_familiar\"], axis = 1)\n",
    "Y = df[\"gasto_familiar\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "c32e68b6-5578-4953-9d95-f2d8d8301a3f",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.linear_model import ElasticNetCV\n",
    "from sklearn.neural_network import MLPRegressor\n",
    "\n",
    "scoring = [\"r2\", \"neg_mean_absolute_percentage_error\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "99f1a26b-538e-44b7-814b-6315c6387cd5",
   "metadata": {},
   "source": [
    "#### ElasticNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "6e81d5c8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Entrenamiento r2 score: 0.09101057299640614\n",
      "Test r2 score: 0.09077536862478137\n",
      "Entrenamiento nMAPE score: -220.16472665913452\n",
      "Test nMAPE score: -217.87883722117763\n"
     ]
    }
   ],
   "source": [
    "ENReg = ElasticNetCV(random_state = 0)\n",
    "\n",
    "scores = cross_validate(ENReg, X, Y,return_train_score = True, cv=5, scoring = scoring)\n",
    "print(f\"Entrenamiento r2 score: {np.mean(scores['train_r2'])}\")\n",
    "print(f\"Test r2 score: {np.mean(scores['test_r2'])}\")\n",
    "print(f\"Entrenamiento nMAPE score: {np.mean(scores['train_neg_mean_absolute_percentage_error'])}\")\n",
    "print(f\"Test nMAPE score: {np.mean(scores['test_neg_mean_absolute_percentage_error'])}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "8f3b9cd2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 478287475292.9548 14012804179.7533            2.56m\n",
      "         2 463317165814.5343 11444584324.3040            2.50m\n",
      "         3 452176152447.9870  9703408293.6000            2.45m\n",
      "         4 438878692914.0586  7922722349.9900            2.40m\n",
      "         5 429935965576.2411  6668389748.1906            2.35m\n",
      "         6 420686200713.0966  5627046639.3767            2.30m\n",
      "         7 412421565180.9791  4764724253.1730            2.25m\n",
      "         8 405572108995.9174  4015257689.6797            2.19m\n",
      "         9 400630512177.8563  3199498712.2465            2.14m\n",
      "        10 395722574208.6388  2743662053.2779            2.09m\n",
      "        11 388464710760.0272  2213302133.9666            2.04m\n",
      "        12 383747198776.5511  2351908329.1379            1.99m\n",
      "        13 378772551413.9266  1863619852.8784            1.94m\n",
      "        14 374538586019.1973  1783665121.1278            1.89m\n",
      "        15 372530057161.9822  1344920413.3699            1.83m\n",
      "        16 369136469884.4760  1270307966.6808            1.78m\n",
      "        17 366593958698.5478   951850168.3056            1.73m\n",
      "        18 363804741760.0946   745885478.6144            1.68m\n",
      "        19 361565200768.9305   717926407.0330            1.63m\n",
      "        20 357395810297.4117   838684994.9816            1.58m\n",
      "        21 357066852512.3509   603345869.9222            1.53m\n",
      "        22 353083242985.3458   716509655.7739            1.47m\n",
      "        23 353668051892.8795   546960463.9625            1.42m\n",
      "        24 351324049096.6597   390903411.8502            1.37m\n",
      "        25 347873446149.1729   510091973.8983            1.32m\n",
      "        26 346567280471.2323   517694506.9362            1.27m\n",
      "        27 344855326447.1261   604328953.2541            1.21m\n",
      "        28 342548862563.6988   512505520.6332            1.16m\n",
      "        29 339525122386.5035   472813562.0789            1.11m\n",
      "        30 337936197154.5009   557042982.4692            1.05m\n",
      "        31 337015457951.5815   332964136.1172            1.00m\n",
      "        32 336254905528.8165   313702764.7592           57.13s\n",
      "        33 334428574753.0713   253130404.3845           54.06s\n",
      "        34 333223721050.3238   220716169.6548           50.92s\n",
      "        35 333766127229.5554    99951378.7461           47.85s\n",
      "        36 331663908251.0878   210327848.1558           44.72s\n",
      "        37 330568408207.1390   216234715.5406           41.59s\n",
      "        38 329807138652.0582   283331555.5504           38.41s\n",
      "        39 329663036962.2814   104272549.1730           35.26s\n",
      "        40 328286191187.9810   111254722.3997           32.10s\n",
      "        41 326862747518.5029   260216763.1509           28.94s\n",
      "        42 326065232260.4745   139784489.9994           25.79s\n",
      "        43 324573130459.9319    59433424.8759           22.60s\n",
      "        44 323788348666.9333    33735804.8255           19.41s\n",
      "        45 325323025303.7849   165939123.4188           16.20s\n",
      "        46 323510736471.0856   149035542.9824           13.00s\n",
      "        47 321438072430.3658   197011548.9984            9.76s\n",
      "        48 320653553552.6157   119446087.8317            6.52s\n",
      "        49 319479436819.2911   165082806.8418            3.26s\n",
      "        50 317574673459.5387   119799212.1880            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 481536510624.5717 14289682198.6439            2.57m\n",
      "         2 467218949708.5623 11846121707.8776            2.51m\n",
      "         3 455596299989.7769  9828882857.3077            2.45m\n",
      "         4 441253992809.1226  8279080937.8767            2.40m\n",
      "         5 430598014770.4315  6741024226.7111            2.34m\n",
      "         6 422510634152.7817  5793676066.0145            2.29m\n",
      "         7 414391501716.9148  4837946304.5613            2.24m\n",
      "         8 408621057330.5236  4070017486.2325            2.18m\n",
      "         9 401896802245.1027  3411448150.2801            2.13m\n",
      "        10 397819466421.7412  2797381042.2820            2.08m\n",
      "        11 390480142762.2133  2585249292.4100            2.03m\n",
      "        12 384693789865.3812  2299297785.0427            1.98m\n",
      "        13 380501995139.4870  1943865992.1640            1.93m\n",
      "        14 377313974131.2811  1671984143.5549            1.88m\n",
      "        15 374483249316.1545  1226088890.5759            1.83m\n",
      "        16 370998666806.0070  1215926512.4843            1.78m\n",
      "        17 368044617064.4503  1033381153.0099            1.73m\n",
      "        18 365047276419.1915   925898100.7348            1.68m\n",
      "        19 362211358206.7183   734245667.1508            1.63m\n",
      "        20 358328302315.2106  1015690797.3555            1.57m\n",
      "        21 356693435102.1993   671233492.2679            1.52m\n",
      "        22 355129410182.5879   553084641.8760            1.47m\n",
      "        23 355212142758.9533   520165980.5394            1.42m\n",
      "        24 352777448523.0642   380254504.8016            1.37m\n",
      "        25 349624263298.5875   496924647.0723            1.32m\n",
      "        26 348468593875.5056   526548986.7573            1.27m\n",
      "        27 347368472944.0579   458616544.2685            1.21m\n",
      "        28 344733152996.3669   209386328.1760            1.16m\n",
      "        29 343608079018.7773   288694785.1665            1.11m\n",
      "        30 342182227803.9734   214589714.8872            1.06m\n",
      "        31 340772990535.8864   444400853.4238            1.00m\n",
      "        32 340558249498.7188   230215574.5765           57.21s\n",
      "        33 336422574643.8441   483549272.9216           54.00s\n",
      "        34 335706416589.9575   298957401.9365           50.91s\n",
      "        35 334998335834.4670   333861912.3064           47.78s\n",
      "        36 334490938265.0313    94080113.0584           44.73s\n",
      "        37 332717142210.0052   382767044.7599           41.61s\n",
      "        38 332460204832.4333   198691765.6047           38.48s\n",
      "        39 331898131722.9177    57203798.0148           35.36s\n",
      "        40 330863667846.5024   123929354.6917           32.22s\n",
      "        41 331163015041.3283    84264863.9140           29.08s\n",
      "        42 330535474469.7997    50794858.3760           25.90s\n",
      "        43 326865506118.6300   106243505.2274           22.70s\n",
      "        44 328215664931.6387     1743812.7568           19.51s\n",
      "        45 327602572929.1093    70681805.6492           16.30s\n",
      "        46 326817404346.4203   239518910.9318           13.05s\n",
      "        47 324724520914.5231    12107232.6262            9.81s\n",
      "        48 324617887977.3510     9995353.2083            6.56s\n",
      "        49 322928957572.2114   289140094.5254            3.28s\n",
      "        50 323308431235.9891    74027431.1985            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 480702832992.8176 13910248831.1580            2.57m\n",
      "         2 468280190844.3414 11637945018.1448            2.51m\n",
      "         3 452204291314.5748  9861751013.8608            2.44m\n",
      "         4 439395718059.8615  8138740008.0654            2.38m\n",
      "         5 428571886619.2152  6902038494.3912            2.32m\n",
      "         6 421483423255.9542  5929514001.0542            2.27m\n",
      "         7 413137638676.1107  4762993574.7662            2.22m\n",
      "         8 406579907585.7198  3897798362.0784            2.17m\n",
      "         9 400694205209.3011  3314925772.2933            2.11m\n",
      "        10 395815006506.2570  3080016912.2540            2.06m\n",
      "        11 389238433685.8546  2424430175.2981            2.01m\n",
      "        12 383321895581.7357  2202437400.8485            1.96m\n",
      "        13 381330770544.5302  1788043278.6893            1.91m\n",
      "        14 378279915374.4732  1492077729.5181            1.86m\n",
      "        15 372805314862.4309  1612537391.1747            1.81m\n",
      "        16 372874299167.4641  1163108603.6639            1.76m\n",
      "        17 365570674820.8119  1202383508.6160            1.70m\n",
      "        18 363858309572.1586   959138109.8323            1.65m\n",
      "        19 360111499876.8895   982495991.7258            1.60m\n",
      "        20 358022639528.8984   822819950.0033            1.55m\n",
      "        21 355107002033.1686   509293870.2426            1.50m\n",
      "        22 350522133663.8673   626915284.6194            1.45m\n",
      "        23 352279731141.9359   611881071.6351            1.40m\n",
      "        24 350441417297.3198   417870911.7687            1.35m\n",
      "        25 346355065065.7739   493561715.1227            1.30m\n",
      "        26 346162337980.5254   343515476.9571            1.25m\n",
      "        27 345931373296.8831   567470793.3484            1.20m\n",
      "        28 341926389163.1615   549193248.7154            1.15m\n",
      "        29 339444317494.7468   388401593.8768            1.10m\n",
      "        30 340010395331.6245   213105169.2417            1.05m\n",
      "        31 339885371250.3238   267821952.9780           59.98s\n",
      "        32 336168152444.0829   232611006.7217           56.91s\n",
      "        33 336937518148.2573   143976425.8172           53.81s\n",
      "        34 333645562366.3571   184523131.2749           50.77s\n",
      "        35 331656160758.7952   410441493.4293           47.61s\n",
      "        36 332771160057.6472    67704266.2970           44.58s\n",
      "        37 329802116619.6737   237714939.9525           41.41s\n",
      "        38 327984256627.7161   108694906.5624           38.39s\n",
      "        39 327071508809.6175   448117067.4567           35.30s\n",
      "        40 326594516336.7222   373776905.3333           32.11s\n",
      "        41 324470420585.1047   233279878.9742           28.88s\n",
      "        42 322359984490.8583   262898062.6372           25.70s\n",
      "        43 324335674547.7296   177251346.0854           22.52s\n",
      "        44 321246227507.5204    -9424937.8390           19.36s\n",
      "        45 319793763984.9980   529256825.8757           16.12s\n",
      "        46 319726130232.0364    95799258.5251           12.93s\n",
      "        47 317952313019.2704   111944893.7822            9.72s\n",
      "        48 317949052286.5468   103481742.7529            6.48s\n",
      "        49 317856660533.7119   118832899.1897            3.25s\n",
      "        50 315520265957.4649   -29423969.2958            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 480284072697.4990 14555918957.3922            2.65m\n",
      "         2 466940422155.2783 11677820550.1074            2.59m\n",
      "         3 451198309365.4883  9873917880.1657            2.53m\n",
      "         4 438712049733.8153  8215335437.2636            2.47m\n",
      "         5 427211741413.9808  6808504321.0950            2.41m\n",
      "         6 421230197559.1165  5694370943.1641            2.36m\n",
      "         7 411321457695.0018  4809710419.8149            2.30m\n",
      "         8 405039321798.1180  3985789208.3876            2.25m\n",
      "         9 398859220315.2183  3318285836.3561            2.19m\n",
      "        10 395059284002.4001  2628373887.4630            2.14m\n",
      "        11 390092860911.9483  2499922328.3827            2.08m\n",
      "        12 383097936718.3427  2071005894.7098            2.03m\n",
      "        13 379659868883.9277  2043661896.6423            1.98m\n",
      "        14 376521816466.2759  1691449862.5879            1.92m\n",
      "        15 371475253455.2861  1568136230.6111            1.87m\n",
      "        16 370779656466.4849  1249228778.3693            1.82m\n",
      "        17 364128729456.9824  1046919273.4949            1.77m\n",
      "        18 362522495867.0579  1129845940.2985            1.72m\n",
      "        19 359043972912.8171   856891956.0181            1.67m\n",
      "        20 356205296775.8096   807096865.7095            1.62m\n",
      "        21 354126974383.6076   675179979.8610            1.56m\n",
      "        22 351109872782.6718   351161884.9963            1.51m\n",
      "        23 353026235570.6660   502629804.7183            1.46m\n",
      "        24 351159356369.2582   364826310.1373            1.41m\n",
      "        25 346246548851.2040   633387603.3357            1.35m\n",
      "        26 345730764087.1810   313064555.9532            1.30m\n",
      "        27 346121273640.1469   422049633.4102            1.25m\n",
      "        28 339753240157.6732   654254211.4409            1.19m\n",
      "        29 338438646600.7570   218460725.3622            1.14m\n",
      "        30 340031518670.2435   319859579.4700            1.09m\n",
      "        31 337474205905.3466   527841245.4934            1.04m\n",
      "        32 334847847969.7609   405418731.6362           58.99s\n",
      "        33 333467798002.7012   401057554.7396           55.64s\n",
      "        34 332262981586.9078   197506608.9191           52.55s\n",
      "        35 330816280519.4507   188567564.4505           49.32s\n",
      "        36 330896975390.8450    94603324.6118           46.10s\n",
      "        37 327579386155.9863   318566236.5229           42.81s\n",
      "        38 327696368157.7300    86425313.2045           39.63s\n",
      "        39 327511524294.3755   249359113.1147           36.40s\n",
      "        40 326499601917.3703   217516698.7102           33.11s\n",
      "        41 324604119017.3472   240794721.0587           29.79s\n",
      "        42 323093672755.9075   202494124.5767           26.48s\n",
      "        43 324056374307.8223   119785014.1492           23.20s\n",
      "        44 321303320253.0287   113852846.9251           19.92s\n",
      "        45 321188013122.9520   151341858.0749           16.61s\n",
      "        46 322288134631.9695    42436740.0945           13.32s\n",
      "        47 319592612212.2078   203318987.8598           10.00s\n",
      "        48 318791594947.6017    59703728.3671            6.68s\n",
      "        49 317714815839.5962   205708602.0721            3.34s\n",
      "        50 315778417804.1844   166609196.9908            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 482308042987.2189 13926108298.5704            2.66m\n",
      "         2 466380057607.6215 11753329015.3958            2.62m\n",
      "         3 452516361466.2333  9773861080.1542            2.56m\n",
      "         4 440734413011.4563  8077884175.9782            2.51m\n",
      "         5 429714939923.1428  6640786818.3127            2.46m\n",
      "         6 421661145290.0915  5711939584.4688            2.40m\n",
      "         7 413840063561.7245  4618938119.7186            2.34m\n",
      "         8 408011767869.5187  3892455388.2866            2.29m\n",
      "         9 399866781916.7069  3646149351.5410            2.24m\n",
      "        10 394065856653.8480  2817274433.1543            2.19m\n",
      "        11 391503076963.7259  2680056201.9308            2.13m\n",
      "        12 384438446842.1322  2180893645.3328            2.07m\n",
      "        13 380682524432.6170  1807019337.4327            2.02m\n",
      "        14 377262944451.5597  1638439201.6915            1.97m\n",
      "        15 372867650329.4794  1529535969.3974            1.92m\n",
      "        16 372059383104.2717  1062138320.2876            1.86m\n",
      "        17 365048749856.7825  1133270492.6176            1.81m\n",
      "        18 365787808003.7560   721657694.5654            1.76m\n",
      "        19 361428156290.4591  1120929207.0788            1.70m\n",
      "        20 358227192948.8688   808519545.6322            1.65m\n",
      "        21 356013908679.8130   651738412.1882            1.59m\n",
      "        22 351985853847.2230   489926808.0796            1.54m\n",
      "        23 353035729368.5050   796327692.3464            1.48m\n",
      "        24 351634235251.6281   313582060.2233            1.43m\n",
      "        25 348206130830.9386   364610371.4213            1.38m\n",
      "        26 346459866528.8710   354348226.7408            1.32m\n",
      "        27 346368154242.0598   499077367.4603            1.27m\n",
      "        28 342045686145.4879   177971256.0851            1.22m\n",
      "        29 342311269827.0290   162738748.8807            1.16m\n",
      "        30 341705003987.5016   439531964.3231            1.11m\n",
      "        31 339433087175.2571   565656934.3105            1.05m\n",
      "        32 336998723928.9901   312359644.9090           59.98s\n",
      "        33 336699777577.6235   244300710.8147           56.81s\n",
      "        34 333776685009.0298   351141644.3952           53.55s\n",
      "        35 333521395278.0754   156817962.1711           50.26s\n",
      "        36 332696564292.1342   228486999.1086           47.01s\n",
      "        37 330710494247.5274   162364465.8972           43.77s\n",
      "        38 331559051778.2984   144805734.2385           40.51s\n",
      "        39 330508028552.4191    58662903.5838           37.27s\n",
      "        40 328651250574.0280    90802620.2589           33.97s\n",
      "        41 328368487682.6542    89922244.7228           30.60s\n",
      "        42 325434345240.7130   492188690.1359           27.18s\n",
      "        43 327534512429.3563    27250818.0411           23.87s\n",
      "        44 325554559855.7408    73315984.9908           20.52s\n",
      "        45 325184479873.7800    59655233.0817           17.13s\n",
      "        46 326095539249.0135    14772965.0726           13.76s\n",
      "        47 324883745668.7477    18568574.4529           10.35s\n",
      "        48 324139135970.2490    20300852.2733            6.91s\n",
      "        49 321449469814.9750   379995203.2509            3.45s\n",
      "        50 319606693515.2524   209305339.3810            0.00s\n",
      "Entrenamiento r2 score: 0.3646248712612062\n",
      "Test r2 score: 0.08585923318357935\n",
      "Entrenamiento nMAPE score: -143.03388819599851\n",
      "Test nMAPE score: -212.8745968783324\n"
     ]
    }
   ],
   "source": [
    "GBReg = GradientBoostingRegressor(n_estimators=50, \n",
    "                                  max_depth = 15,\n",
    "                                  random_state = 101,\n",
    "                                  subsample = 0.5,\n",
    "                                  verbose = 2)\n",
    "\n",
    "scores = cross_validate(GBReg, X, Y,return_train_score = True, cv=5, scoring = scoring)\n",
    "print(f\"Entrenamiento r2 score: {np.mean(scores['train_r2'])}\")\n",
    "print(f\"Test r2 score: {np.mean(scores['test_r2'])}\")\n",
    "print(f\"Entrenamiento nMAPE score: {np.mean(scores['train_neg_mean_absolute_percentage_error'])}\")\n",
    "print(f\"Test nMAPE score: {np.mean(scores['test_neg_mean_absolute_percentage_error'])}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "00d7df2d-6ade-409a-972f-67bdf93a351c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 425622947772.4495 24317321835.4986            4.07m\n",
      "         2 383010822094.4579 20586607692.3362            4.00m\n",
      "         3 350263555021.2340 16088927143.3942            3.92m\n",
      "         4 317328200185.4700 12558983378.8102            3.83m\n",
      "         5 291438018784.1011  9892118026.6293            3.75m\n",
      "         6 267282227216.0476  7628214057.5794            3.67m\n",
      "         7 247657990376.7360  5662181752.5552            3.60m\n",
      "         8 231260750222.1509  4088363906.6649            3.52m\n",
      "         9 215625884658.2798  2799353255.0235            3.44m\n",
      "        10 201597085383.7921  1886256836.9419            3.36m\n",
      "        11 189846686088.3466  1115889871.6492            3.29m\n",
      "        12 179990424210.9078   524223189.8716            3.21m\n",
      "        13 169976515652.4926    42755308.5072            3.14m\n",
      "        14 161657667326.4137  -401051536.7065            3.06m\n",
      "        15 155029889808.0010  -663796425.5184            2.98m\n",
      "        16 147997207239.9485 -1048572867.5745            2.90m\n",
      "        17 142405113507.9664 -1182594673.6180            2.82m\n",
      "        18 136807241920.2231 -1463066067.2455            2.74m\n",
      "        19 132312270857.9744 -1516429484.3217            2.67m\n",
      "        20 126855088332.2063 -1725443535.7315            2.59m\n",
      "        21 122908756170.4587 -1859524869.8954            2.50m\n",
      "        22 118932129016.4242 -1917281558.2548            2.42m\n",
      "        23 116578272847.4402 -1856519454.1126            2.34m\n",
      "        24 114261618209.9711 -1707892035.5284            2.26m\n",
      "        25 109984291390.3824 -1920555734.8042            2.17m\n",
      "        26 108069474446.4327 -1933286434.6341            2.09m\n",
      "        27 106861962059.7005 -1831763064.4065            2.01m\n",
      "        28 104414575593.2077 -1835330682.2937            1.92m\n",
      "        29 101901379020.9744 -1788423377.5350            1.84m\n",
      "        30 100548742517.2698 -1679987611.8459            1.76m\n",
      "        31 98552594047.4952 -1921729756.7143            1.67m\n",
      "        32 98495264550.6827 -1249642558.9623            1.60m\n",
      "        33 96216892275.1361 -1738019701.8298            1.51m\n",
      "        34 95529526484.8921 -1408976626.6093            1.43m\n",
      "        35 94107069110.5666 -1523123398.6608            1.34m\n",
      "        36 93369555644.6636 -1446254581.2711            1.26m\n",
      "        37 91056533107.3700 -1794993656.8970            1.17m\n",
      "        38 90067374284.4836 -1722022689.8996            1.08m\n",
      "        39 88767091587.7999 -1810998767.0907           59.73s\n",
      "        40 88351855888.2100 -1464529178.5779           54.54s\n",
      "        41 87924274143.4048 -1305464900.6537           49.32s\n",
      "        42 87288398892.7673 -1219779610.0976           44.02s\n",
      "        43 84436853339.0948 -2029006071.4658           38.50s\n",
      "        44 84989692262.1172 -1272125763.9826           33.10s\n",
      "        45 85056399759.3528 -1186953006.4836           27.70s\n",
      "        46 84333258813.9980 -1178441796.6597           22.24s\n",
      "        47 82972528787.9122 -1204323777.0017           16.73s\n",
      "        48 82197136149.1027 -1325786971.1534           11.18s\n",
      "        49 81948899493.8654 -1157542597.3521            5.62s\n",
      "        50 80944984955.4859 -1186712407.9933            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 427422683796.6341 25456558972.3635            4.02m\n",
      "         2 388232317294.9388 20173328864.4387            3.94m\n",
      "         3 353138767798.8794 16103249482.7762            3.85m\n",
      "         4 321744636677.2800 12591655002.4873            3.79m\n",
      "         5 293177156241.3475 10214984310.3153            3.70m\n",
      "         6 269467256802.7322  7962240931.8411            3.63m\n",
      "         7 249039392221.2010  5981591463.0472            3.55m\n",
      "         8 231390699646.6417  4147940545.5581            3.47m\n",
      "         9 215875407102.2001  3041753621.3692            3.39m\n",
      "        10 201612272596.3083  2045493184.4663            3.31m\n",
      "        11 188821986611.0667  1406683946.5073            3.23m\n",
      "        12 177239754385.5742   590025535.6204            3.16m\n",
      "        13 169317524988.1497   102156085.9344            3.08m\n",
      "        14 160324085822.9474  -490064566.5030            3.01m\n",
      "        15 153445289741.6805  -761030223.7349            2.93m\n",
      "        16 146211489864.2328 -1030387440.9953            2.85m\n",
      "        17 141362541708.4133 -1305324515.1675            2.77m\n",
      "        18 136343256151.2136 -1401569543.8961            2.70m\n",
      "        19 131006779222.6244 -1530769521.3557            2.62m\n",
      "        20 125989318819.2036 -1876421682.9369            2.55m\n",
      "        21 123391507837.7026 -1661670910.8145            2.47m\n",
      "        22 119085174156.2749 -1840121486.6323            2.39m\n",
      "        23 115597143470.2733 -2049255953.2208            2.30m\n",
      "        24 113335362393.2543 -1947724869.7998            2.23m\n",
      "        25 110086009774.8844 -1901991919.1099            2.15m\n",
      "        26 107453418539.6845 -1978045649.5899            2.07m\n",
      "        27 105855233386.3005 -1902967493.1303            1.99m\n",
      "        28 102433704715.3048 -2087203042.4062            1.90m\n",
      "        29 99944372166.1732 -2125783190.4447            1.82m\n",
      "        30 98921410404.8546 -1819058398.3202            1.74m\n",
      "        31 97024335693.1676 -1915628610.0197            1.65m\n",
      "        32 96301492727.7978 -1658470087.1260            1.57m\n",
      "        33 94101104244.9002 -1792756824.0695            1.49m\n",
      "        34 93198305559.9754 -1637855218.6431            1.41m\n",
      "        35 92603424170.4567 -1449248083.2616            1.32m\n",
      "        36 91066935998.8835 -1716105582.2008            1.24m\n",
      "        37 90578295141.2947 -1265696110.5571            1.16m\n",
      "        38 89046197197.3313 -1682781438.7901            1.07m\n",
      "        39 88291837093.3790 -1575443178.5888           59.21s\n",
      "        40 87166094356.5823 -1673484326.2722           54.01s\n",
      "        41 87520421480.9265 -1193586104.6802           48.86s\n",
      "        42 87105439538.7595  -956022624.0017           43.63s\n",
      "        43 85985727283.2019  -981220684.5677           38.33s\n",
      "        44 86454612217.0151  -779524490.6062           33.05s\n",
      "        45 84438335945.9905 -1376649361.0614           27.63s\n",
      "        46 84506361859.3649 -1100064107.4623           22.22s\n",
      "        47 83438904002.9804 -1163250961.3378           16.74s\n",
      "        48 82920490025.0632  -990606201.7053           11.20s\n",
      "        49 82910121745.4742  -801435509.5237            5.64s\n",
      "        50 82597184707.6255 -1015855276.4404            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 426281922802.4763 24805216122.4017            3.91m\n",
      "         2 388357902735.1093 19932790758.7629            3.84m\n",
      "         3 347707967609.8648 16540229394.2005            3.73m\n",
      "         4 318840331692.0635 12515105696.8585            3.67m\n",
      "         5 288930805186.5992 10440292223.7976            3.60m\n",
      "         6 266688411771.9815  7880907139.6858            3.52m\n",
      "         7 246509354123.6629  5887814045.3953            3.46m\n",
      "         8 227825115798.0803  4479488619.3690            3.38m\n",
      "         9 213401372877.5758  3073380288.0263            3.32m\n",
      "        10 199921201587.2769  1869685458.3690            3.25m\n",
      "        11 187420739208.3011  1054577544.8740            3.19m\n",
      "        12 177808674360.7476   454609454.1736            3.12m\n",
      "        13 167475271346.2113   -86014268.9835            3.05m\n",
      "        14 160388497860.1483  -612885271.6118            2.98m\n",
      "        15 153159258887.4569  -710348042.5116            2.92m\n",
      "        16 147807497510.7429  -963103672.2226            2.85m\n",
      "        17 139078295539.7713 -1318377575.1970            2.77m\n",
      "        18 133381173099.7490 -1580590675.5784            2.69m\n",
      "        19 129331296559.1170 -1606086552.9595            2.61m\n",
      "        20 124465645032.3232 -1828935078.2192            2.53m\n",
      "        21 120842393782.3402 -1803804348.7851            2.46m\n",
      "        22 117737361760.2578 -1758239034.7741            2.39m\n",
      "        23 115126440789.3201 -1963889030.6922            2.31m\n",
      "        24 111739278346.1249 -2069630684.3650            2.23m\n",
      "        25 108571710144.8562 -2020962311.1128            2.14m\n",
      "        26 105844185459.2701 -2037141630.8914            2.06m\n",
      "        27 104764281072.4538 -1840380906.1798            1.98m\n",
      "        28 100806602167.0507 -2008791540.9917            1.90m\n",
      "        29 99736723725.4621 -1736381317.1070            1.82m\n",
      "        30 97586910110.1466 -2044273845.4066            1.74m\n",
      "        31 97876419759.5130 -1509316959.6597            1.66m\n",
      "        32 95485681545.7934 -1691037901.8959            1.58m\n",
      "        33 94055860413.2654 -1849184883.1723            1.49m\n",
      "        34 91806087290.5281 -1990859204.7920            1.41m\n",
      "        35 90837055446.2894 -1777018989.6323            1.33m\n",
      "        36 89865572109.2971 -1775093111.5643            1.24m\n",
      "        37 89610696505.1628 -1290576260.2101            1.16m\n",
      "        38 88702964503.8389 -1331129214.1339            1.08m\n",
      "        39 88376930538.4666 -1220972154.7591           59.46s\n",
      "        40 85458781571.7188 -2100679561.8322           54.09s\n",
      "        41 85160995977.1767 -1670805351.7262           48.79s\n",
      "        42 84329631236.2338 -1505056433.5680           43.55s\n",
      "        43 84592455603.9660 -1199807275.5281           38.20s\n",
      "        44 83342661967.2500 -1371953908.4582           32.88s\n",
      "        45 82186489586.6179 -1528918898.9825           27.43s\n",
      "        46 80870446281.2526 -1802455909.6369           21.97s\n",
      "        47 81057763006.4138 -1243834636.3855           16.56s\n",
      "        48 81055808634.7391 -1053785350.3782           11.09s\n",
      "        49 80736867360.8567 -1065190085.6000            5.56s\n",
      "        50 79490347387.3233 -1374254013.3081            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 427249119410.8594 25035639080.6110            4.02m\n",
      "         2 387484996638.6804 20072205710.4896            3.95m\n",
      "         3 348794527313.5327 16378742905.6560            3.85m\n",
      "         4 318712903365.1167 12382803324.1523            3.74m\n",
      "         5 291835994856.4437  9947462858.1785            3.68m\n",
      "         6 269680412682.0074  7536501112.0066            3.60m\n",
      "         7 247567028448.2168  6156281004.2562            3.53m\n",
      "         8 230679233103.6975  4395168996.1857            3.46m\n",
      "         9 215068875138.3877  3029035950.8022            3.39m\n",
      "        10 199910845339.8481  2246007095.2696            3.31m\n",
      "        11 189297236224.4182  1210966089.9805            3.24m\n",
      "        12 177517306213.2465   501935982.4704            3.17m\n",
      "        13 168353330735.6973   -15955834.8611            3.09m\n",
      "        14 160103081083.9441  -394927510.9809            3.01m\n",
      "        15 152481850780.8096  -718963256.1947            2.93m\n",
      "        16 146553927400.7843 -1169398392.0998            2.85m\n",
      "        17 137488773494.0251 -1491885490.1693            2.77m\n",
      "        18 133474260203.5341 -1553399573.7783            2.69m\n",
      "        19 128794806235.8869 -1629200020.5494            2.61m\n",
      "        20 124206819273.2935 -1848550365.8973            2.54m\n",
      "        21 120479287512.1789 -1892154521.9942            2.46m\n",
      "        22 116521994220.1732 -1874674714.5151            2.38m\n",
      "        23 114695320512.3539 -2000666878.4155            2.31m\n",
      "        24 112044612442.9155 -1902695808.7985            2.23m\n",
      "        25 107614262155.9751 -2118886168.2340            2.15m\n",
      "        26 105663140704.2984 -2006478021.5776            2.07m\n",
      "        27 104872296029.2179 -1634335610.1722            1.99m\n",
      "        28 100750971224.4852 -2181657558.4668            1.91m\n",
      "        29 99374873732.2838 -1727631223.8218            1.83m\n",
      "        30 97293174976.0442 -2041200096.7343            1.75m\n",
      "        31 96348899660.5949 -1824532901.4824            1.67m\n",
      "        32 95775980853.6000 -1379754797.0482            1.59m\n",
      "        33 93275291248.1145 -1931720998.5225            1.50m\n",
      "        34 92430041550.4959 -1584095234.4412            1.42m\n",
      "        35 90782694230.5188 -1827595159.6715            1.34m\n",
      "        36 91885035067.7713  -994341160.3380            1.26m\n",
      "        37 89918775915.8304 -1295093543.5101            1.17m\n",
      "        38 89395403072.2740 -1411585916.7865            1.09m\n",
      "        39 87992257992.6420 -1589320985.4552           59.88s\n",
      "        40 87733800821.7982 -1427917564.5701           54.63s\n",
      "        41 85968076915.2776 -1594742513.4091           49.21s\n",
      "        42 84759453761.1056 -1588766756.5542           43.92s\n",
      "        43 84678034405.5806 -1418892152.2708           38.57s\n",
      "        44 83575240026.5554 -1496233320.2501           33.16s\n",
      "        45 82293661661.8333 -1862790266.9691           27.63s\n",
      "        46 82931793083.1804 -1123092214.1334           22.16s\n",
      "        47 81510269290.5556 -1397445301.3324           16.67s\n",
      "        48 80930846401.9747 -1401527297.4911           11.14s\n",
      "        49 81225409913.1839  -951712626.4731            5.59s\n",
      "        50 80552120624.8676 -1185534413.3717            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 428908933563.9658 25192517939.7335            4.03m\n",
      "         2 389663638021.1064 19875798263.5532            3.98m\n",
      "         3 350004384982.4893 16308936741.9643            3.91m\n",
      "         4 319737932627.9332 12754068308.2341            3.83m\n",
      "         5 291772136109.3255  9833506048.8485            3.76m\n",
      "         6 268737135325.4892  7803514015.0283            3.68m\n",
      "         7 246134027180.1836  6073555266.2869            3.59m\n",
      "         8 231036777873.3843  4273366399.7642            3.52m\n",
      "         9 213716412768.6347  3175415695.3784            3.44m\n",
      "        10 200131391914.4084  2048375863.9669            3.36m\n",
      "        11 189813774346.0798  1103018132.6215            3.28m\n",
      "        12 177370386202.3267   580219369.4139            3.20m\n",
      "        13 168722345896.8088   -33654454.4350            3.12m\n",
      "        14 160974888817.2084  -458413790.4841            3.04m\n",
      "        15 153647262736.3950  -745735063.6987            2.97m\n",
      "        16 147468547327.8619  -997597682.4981            2.89m\n",
      "        17 141579312627.2971 -1181025976.2626            2.81m\n",
      "        18 135348320123.6311 -1468283398.1176            2.74m\n",
      "        19 130659916638.3389 -1677367666.2366            2.65m\n",
      "        20 125700578760.8271 -1809707563.4567            2.57m\n",
      "        21 121006741224.7724 -1933343707.2676            2.49m\n",
      "        22 116458131674.2482 -1976371185.0521            2.41m\n",
      "        23 113740196597.3343 -2215513305.4708            2.33m\n",
      "        24 111317647077.9247 -2086191788.4871            2.25m\n",
      "        25 108476259531.1677 -1894088126.1191            2.17m\n",
      "        26 105785474000.2171 -2044604018.5547            2.09m\n",
      "        27 102858250056.4547 -2106129255.0140            2.00m\n",
      "        28 99570529494.8334 -2298014541.8274            1.92m\n",
      "        29 99164661177.2588 -1763829533.8431            1.84m\n",
      "        30 97016660074.3020 -1913836460.8958            1.76m\n",
      "        31 96093291579.6865 -1776793458.0800            1.68m\n",
      "        32 94861720190.8432 -1650372656.1415            1.59m\n",
      "        33 93018916613.0983 -1857254573.6342            1.51m\n",
      "        34 90239220773.5679 -2015674327.8806            1.42m\n",
      "        35 89251485241.4831 -2055704327.0303            1.34m\n",
      "        36 90081998170.5609 -1158130045.6684            1.25m\n",
      "        37 88479132259.2239 -1429276924.8154            1.17m\n",
      "        38 87176236713.3320 -1719264146.4293            1.08m\n",
      "        39 87903499113.3511 -1094664549.2542           59.77s\n",
      "        40 85550413675.8716 -1657669382.9141           54.45s\n",
      "        41 85042109542.3568 -1481956947.1595           49.14s\n",
      "        42 83841102449.7130 -1517982200.1527           43.84s\n",
      "        43 84116541232.3994 -1080538782.6053           38.53s\n",
      "        44 83316459825.1317 -1264581005.1967           33.17s\n",
      "        45 82926135511.2289 -1161415834.7716           27.75s\n",
      "        46 82108694882.4639 -1345509120.2108           22.28s\n",
      "        47 82736625422.9128  -630337040.0495           16.81s\n",
      "        48 81255200258.4328 -1244297744.2192           11.23s\n",
      "        49 81308366254.5157  -729262775.3016            5.65s\n",
      "        50 80841560898.2296  -899911873.5180            0.00s\n",
      "Entrenamiento r2 score: 0.8360666262700975\n",
      "Test r2 score: -0.005377906677231125\n",
      "Entrenamiento nMAPE score: -50.02766082964796\n",
      "Test nMAPE score: -323.05545094306297\n"
     ]
    }
   ],
   "source": [
    "GBReg = GradientBoostingRegressor(n_estimators=50, \n",
    "                                  max_depth = 30,\n",
    "                                  random_state = 101,\n",
    "                                  subsample = 0.5,\n",
    "                                  verbose = 2)\n",
    "\n",
    "scores = cross_validate(GBReg, X, Y,return_train_score = True, cv=5, scoring = scoring)\n",
    "print(f\"Entrenamiento r2 score: {np.mean(scores['train_r2'])}\")\n",
    "print(f\"Test r2 score: {np.mean(scores['test_r2'])}\")\n",
    "print(f\"Entrenamiento nMAPE score: {np.mean(scores['train_neg_mean_absolute_percentage_error'])}\")\n",
    "print(f\"Test nMAPE score: {np.mean(scores['test_neg_mean_absolute_percentage_error'])}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "47ac7f72-351e-4005-8512-a0ddb8762a6a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 425622947772.4495 24317321835.4986            4.03m\n",
      "         2 383010822094.4579 20586607692.3362            3.96m\n",
      "         3 350263555021.2340 16088927143.3942            3.87m\n",
      "         4 317328200185.4700 12558983378.8102            3.79m\n",
      "         5 291438018784.1011  9892118026.6293            3.71m\n",
      "         6 267282227216.0476  7628214057.5794            3.63m\n",
      "         7 247657990376.7360  5662181752.5552            3.54m\n",
      "         8 231260750222.1509  4088363906.6649            3.44m\n",
      "         9 215625884658.2798  2799353255.0235            3.36m\n",
      "        10 201597085383.7921  1886256836.9419            3.28m\n",
      "        11 189846686088.3466  1115889871.6492            3.21m\n",
      "        12 179990424210.9078   524223189.8716            3.13m\n",
      "        13 169976515652.4926    42755308.5072            3.05m\n",
      "        14 161657667326.4137  -401051536.7065            2.97m\n",
      "        15 155029889808.0010  -663796425.5184            2.89m\n",
      "        16 147997207239.9485 -1048572867.5745            2.81m\n",
      "        17 142405113507.9664 -1182594673.6180            2.73m\n",
      "        18 136807241920.2231 -1463066067.2455            2.66m\n",
      "        19 132312270857.9744 -1516429484.3217            2.58m\n",
      "        20 126855088332.2063 -1725443535.7315            2.50m\n",
      "        21 122908756170.4587 -1859524869.8954            2.42m\n",
      "        22 118932129016.4242 -1917281558.2548            2.33m\n",
      "        23 116578272847.4402 -1856519454.1126            2.25m\n",
      "        24 114261618209.9711 -1707892035.5284            2.17m\n",
      "        25 109984291390.3824 -1920555734.8042            2.09m\n",
      "        26 108069474446.4327 -1933286434.6341            2.01m\n",
      "        27 106861962059.7005 -1831763064.4065            1.93m\n",
      "        28 104414575593.2077 -1835330682.2937            1.85m\n",
      "        29 101901379020.9744 -1788423377.5350            1.77m\n",
      "        30 100548742517.2698 -1679987611.8459            1.69m\n",
      "        31 98552594047.4952 -1921729756.7143            1.61m\n",
      "        32 98495264550.6827 -1249642558.9623            1.53m\n",
      "        33 96216892275.1361 -1738019701.8298            1.45m\n",
      "        34 95529526484.8921 -1408976626.6093            1.37m\n",
      "        35 94107069110.5666 -1523123398.6608            1.29m\n",
      "        36 93369555644.6636 -1446254581.2711            1.21m\n",
      "        37 91056533107.3700 -1794993656.8970            1.12m\n",
      "        38 90067374284.4836 -1722022689.8996            1.04m\n",
      "        39 88767091587.7999 -1810998767.0907           57.32s\n",
      "        40 88351855888.2100 -1464529178.5779           52.30s\n",
      "        41 87924274143.4048 -1305464900.6537           47.29s\n",
      "        42 87288398892.7673 -1219779610.0976           42.21s\n",
      "        43 84436853339.0948 -2029006071.4658           36.92s\n",
      "        44 84989692262.1172 -1272125763.9826           31.73s\n",
      "        45 85056399759.3528 -1186953006.4836           26.54s\n",
      "        46 84333258813.9980 -1178441796.6597           21.29s\n",
      "        47 82972528787.9122 -1204323777.0017           16.01s\n",
      "        48 82197136149.1027 -1325786971.1534           10.70s\n",
      "        49 81948899493.8654 -1157542597.3521            5.37s\n",
      "        50 80944984955.4859 -1186712407.9933            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 427422683796.6341 25456558972.3635            3.87m\n",
      "         2 388232317294.9388 20173328864.4387            3.81m\n",
      "         3 353138767798.8794 16103249482.7762            3.72m\n",
      "         4 321744636677.2800 12591655002.4873            3.66m\n",
      "         5 293177156241.3475 10214984310.3153            3.57m\n",
      "         6 269467256802.7322  7962240931.8411            3.51m\n",
      "         7 249039392221.2010  5981591463.0472            3.43m\n",
      "         8 231390699646.6417  4147940545.5581            3.36m\n",
      "         9 215875407102.2001  3041753621.3692            3.29m\n",
      "        10 201612272596.3083  2045493184.4663            3.22m\n",
      "        11 188821986611.0667  1406683946.5073            3.14m\n",
      "        12 177239754385.5742   590025535.6204            3.06m\n",
      "        13 169317524988.1497   102156085.9344            2.99m\n",
      "        14 160324085822.9474  -490064566.5030            2.91m\n",
      "        15 153445289741.6805  -761030223.7349            2.84m\n",
      "        16 146211489864.2328 -1030387440.9953            2.76m\n",
      "        17 141362541708.4133 -1305324515.1675            2.69m\n",
      "        18 136343256151.2136 -1401569543.8961            2.61m\n",
      "        19 131006779222.6244 -1530769521.3557            2.54m\n",
      "        20 125989318819.2036 -1876421682.9369            2.46m\n",
      "        21 123391507837.7026 -1661670910.8145            2.39m\n",
      "        22 119085174156.2749 -1840121486.6323            2.31m\n",
      "        23 115597143470.2733 -2049255953.2208            2.23m\n",
      "        24 113335362393.2543 -1947724869.7998            2.15m\n",
      "        25 110086009774.8844 -1901991919.1099            2.07m\n",
      "        26 107453418539.6845 -1978045649.5899            1.99m\n",
      "        27 105855233386.3005 -1902967493.1303            1.91m\n",
      "        28 102433704715.3048 -2087203042.4062            1.83m\n",
      "        29 99944372166.1732 -2125783190.4447            1.75m\n",
      "        30 98921410404.8546 -1819058398.3202            1.67m\n",
      "        31 97024335693.1676 -1915628610.0197            1.59m\n",
      "        32 96301492727.7978 -1658470087.1260            1.52m\n",
      "        33 94101104244.9002 -1792756824.0695            1.43m\n",
      "        34 93198305559.9754 -1637855218.6431            1.35m\n",
      "        35 92603424170.4567 -1449248083.2616            1.27m\n",
      "        36 91066935998.8835 -1716105582.2008            1.19m\n",
      "        37 90578295141.2947 -1265696110.5571            1.11m\n",
      "        38 89046197197.3313 -1682781438.7901            1.03m\n",
      "        39 88291837093.3790 -1575443178.5888           56.88s\n",
      "        40 87166094356.5823 -1673484326.2722           51.84s\n",
      "        41 87520421480.9265 -1193586104.6802           46.87s\n",
      "        42 87105439538.7595  -956022624.0017           41.83s\n",
      "        43 85985727283.2019  -981220684.5677           36.72s\n",
      "        44 86454612217.0151  -779524490.6062           31.65s\n",
      "        45 84438335945.9905 -1376649361.0614           26.46s\n",
      "        46 84506361859.3649 -1100064107.4623           21.26s\n",
      "        47 83438904002.9804 -1163250961.3378           16.00s\n",
      "        48 82920490025.0632  -990606201.7053           10.70s\n",
      "        49 82910121745.4742  -801435509.5237            5.38s\n",
      "        50 82597184707.6255 -1015855276.4404            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 426281922802.4763 24805216122.4017            3.93m\n",
      "         2 388357902735.1093 19932790758.7629            3.83m\n",
      "         3 347707967609.8648 16540229394.2005            3.73m\n",
      "         4 318840331692.0635 12515105696.8585            3.67m\n",
      "         5 288930805186.5992 10440292223.7976            3.58m\n",
      "         6 266688411771.9815  7880907139.6858            3.50m\n",
      "         7 246509354123.6629  5887814045.3953            3.43m\n",
      "         8 227825115798.0803  4479488619.3690            3.35m\n",
      "         9 213401372877.5758  3073380288.0263            3.29m\n",
      "        10 199921201587.2769  1869685458.3690            3.21m\n",
      "        11 187420739208.3011  1054577544.8740            3.14m\n",
      "        12 177808674360.7476   454609454.1736            3.07m\n",
      "        13 167475271346.2113   -86014268.9835            2.99m\n",
      "        14 160388497860.1483  -612885271.6118            2.92m\n",
      "        15 153159258887.4569  -710348042.5116            2.84m\n",
      "        16 147807497510.7429  -963103672.2226            2.78m\n",
      "        17 139078295539.7713 -1318377575.1970            2.70m\n",
      "        18 133381173099.7490 -1580590675.5784            2.62m\n",
      "        19 129331296559.1170 -1606086552.9595            2.54m\n",
      "        20 124465645032.3232 -1828935078.2192            2.46m\n",
      "        21 120842393782.3402 -1803804348.7851            2.39m\n",
      "        22 117737361760.2578 -1758239034.7741            2.31m\n",
      "        23 115126440789.3201 -1963889030.6922            2.23m\n",
      "        24 111739278346.1249 -2069630684.3650            2.16m\n",
      "        25 108571710144.8562 -2020962311.1128            2.07m\n",
      "        26 105844185459.2701 -2037141630.8914            1.99m\n",
      "        27 104764281072.4538 -1840380906.1798            1.91m\n",
      "        28 100806602167.0507 -2008791540.9917            1.83m\n",
      "        29 99736723725.4621 -1736381317.1070            1.75m\n",
      "        30 97586910110.1466 -2044273845.4066            1.68m\n",
      "        31 97876419759.5130 -1509316959.6597            1.60m\n",
      "        32 95485681545.7934 -1691037901.8959            1.52m\n",
      "        33 94055860413.2654 -1849184883.1723            1.44m\n",
      "        34 91806087290.5281 -1990859204.7920            1.36m\n",
      "        35 90837055446.2894 -1777018989.6323            1.28m\n",
      "        36 89865572109.2971 -1775093111.5643            1.20m\n",
      "        37 89610696505.1628 -1290576260.2101            1.12m\n",
      "        38 88702964503.8389 -1331129214.1339            1.04m\n",
      "        39 88376930538.4666 -1220972154.7591           57.38s\n",
      "        40 85458781571.7188 -2100679561.8322           52.19s\n",
      "        41 85160995977.1767 -1670805351.7262           47.08s\n",
      "        42 84329631236.2338 -1505056433.5680           41.98s\n",
      "        43 84592455603.9660 -1199807275.5281           36.84s\n",
      "        44 83342661967.2500 -1371953908.4582           31.71s\n",
      "        45 82186489586.6179 -1528918898.9825           26.46s\n",
      "        46 80870446281.2526 -1802455909.6369           21.20s\n",
      "        47 81057763006.4138 -1243834636.3855           15.99s\n",
      "        48 81055808634.7391 -1053785350.3782           10.70s\n",
      "        49 80736867360.8567 -1065190085.6000            5.37s\n",
      "        50 79490347387.3233 -1374254013.3081            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 427249119410.8594 25035639080.6110            4.04m\n",
      "         2 387484996638.6804 20072205710.4896            4.01m\n",
      "         3 348794527313.5327 16378742905.6560            3.92m\n",
      "         4 318712903365.1167 12382803324.1523            3.85m\n",
      "         5 291835994856.4437  9947462858.1785            3.77m\n",
      "         6 269680412682.0074  7536501112.0066            3.70m\n",
      "         7 247567028448.2168  6156281004.2562            3.62m\n",
      "         8 230679233103.6975  4395168996.1857            3.54m\n",
      "         9 215068875138.3877  3029035950.8022            3.45m\n",
      "        10 199910845339.8481  2246007095.2696            3.37m\n",
      "        11 189297236224.4182  1210966089.9805            3.29m\n",
      "        12 177517306213.2465   501935982.4704            3.20m\n",
      "        13 168353330735.6973   -15955834.8611            3.12m\n",
      "        14 160103081083.9441  -394927510.9809            3.04m\n",
      "        15 152481850780.8096  -718963256.1947            2.96m\n",
      "        16 146553927400.7843 -1169398392.0998            2.88m\n",
      "        17 137488773494.0251 -1491885490.1693            2.80m\n",
      "        18 133474260203.5341 -1553399573.7783            2.72m\n",
      "        19 128794806235.8869 -1629200020.5494            2.64m\n",
      "        20 124206819273.2935 -1848550365.8973            2.57m\n",
      "        21 120479287512.1789 -1892154521.9942            2.49m\n",
      "        22 116521994220.1732 -1874674714.5151            2.41m\n",
      "        23 114695320512.3539 -2000666878.4155            2.33m\n",
      "        24 112044612442.9155 -1902695808.7985            2.25m\n",
      "        25 107614262155.9751 -2118886168.2340            2.16m\n",
      "        26 105663140704.2984 -2006478021.5776            2.08m\n",
      "        27 104872296029.2179 -1634335610.1722            2.00m\n",
      "        28 100750971224.4852 -2181657558.4668            1.92m\n",
      "        29 99374873732.2838 -1727631223.8218            1.84m\n",
      "        30 97293174976.0442 -2041200096.7343            1.75m\n",
      "        31 96348899660.5949 -1824532901.4824            1.67m\n",
      "        32 95775980853.6000 -1379754797.0482            1.59m\n",
      "        33 93275291248.1145 -1931720998.5225            1.51m\n",
      "        34 92430041550.4959 -1584095234.4412            1.42m\n",
      "        35 90782694230.5188 -1827595159.6715            1.34m\n",
      "        36 91885035067.7713  -994341160.3380            1.26m\n",
      "        37 89918775915.8304 -1295093543.5101            1.17m\n",
      "        38 89395403072.2740 -1411585916.7865            1.09m\n",
      "        39 87992257992.6420 -1589320985.4552           59.89s\n",
      "        40 87733800821.7982 -1427917564.5701           54.63s\n",
      "        41 85968076915.2776 -1594742513.4091           49.20s\n",
      "        42 84759453761.1056 -1588766756.5542           43.88s\n",
      "        43 84678034405.5806 -1418892152.2708           38.54s\n",
      "        44 83575240026.5554 -1496233320.2501           33.12s\n",
      "        45 82293661661.8333 -1862790266.9691           27.59s\n",
      "        46 82931793083.1804 -1123092214.1334           22.12s\n",
      "        47 81510269290.5556 -1397445301.3324           16.64s\n",
      "        48 80930846401.9747 -1401527297.4911           11.12s\n",
      "        49 81225409913.1839  -951712626.4731            5.58s\n",
      "        50 80552120624.8676 -1185534413.3717            0.00s\n",
      "      Iter       Train Loss      OOB Improve   Remaining Time \n",
      "         1 428908933563.9658 25192517939.7335            4.00m\n",
      "         2 389663638021.1064 19875798263.5532            3.96m\n",
      "         3 350004384982.4893 16308936741.9643            3.88m\n",
      "         4 319737932627.9332 12754068308.2341            3.80m\n",
      "         5 291772136109.3255  9833506048.8485            3.72m\n",
      "         6 268737135325.4892  7803514015.0283            3.65m\n",
      "         7 246134027180.1836  6073555266.2869            3.57m\n",
      "         8 231036777873.3843  4273366399.7642            3.51m\n",
      "         9 213716412768.6347  3175415695.3784            3.43m\n",
      "        10 200131391914.4084  2048375863.9669            3.35m\n",
      "        11 189813774346.0798  1103018132.6215            3.26m\n",
      "        12 177370386202.3267   580219369.4139            3.19m\n",
      "        13 168722345896.8088   -33654454.4350            3.10m\n",
      "        14 160974888817.2084  -458413790.4841            3.03m\n",
      "        15 153647262736.3950  -745735063.6987            2.95m\n",
      "        16 147468547327.8619  -997597682.4981            2.88m\n",
      "        17 141579312627.2971 -1181025976.2626            2.80m\n",
      "        18 135348320123.6311 -1468283398.1176            2.72m\n",
      "        19 130659916638.3389 -1677367666.2366            2.64m\n",
      "        20 125700578760.8271 -1809707563.4567            2.56m\n",
      "        21 121006741224.7724 -1933343707.2676            2.48m\n",
      "        22 116458131674.2482 -1976371185.0521            2.40m\n",
      "        23 113740196597.3343 -2215513305.4708            2.31m\n",
      "        24 111317647077.9247 -2086191788.4871            2.23m\n",
      "        25 108476259531.1677 -1894088126.1191            2.15m\n",
      "        26 105785474000.2171 -2044604018.5547            2.07m\n",
      "        27 102858250056.4547 -2106129255.0140            1.99m\n",
      "        28 99570529494.8334 -2298014541.8274            1.91m\n",
      "        29 99164661177.2588 -1763829533.8431            1.83m\n",
      "        30 97016660074.3020 -1913836460.8958            1.75m\n",
      "        31 96093291579.6865 -1776793458.0800            1.67m\n",
      "        32 94861720190.8432 -1650372656.1415            1.59m\n",
      "        33 93018916613.0983 -1857254573.6342            1.50m\n",
      "        34 90239220773.5679 -2015674327.8806            1.42m\n",
      "        35 89251485241.4831 -2055704327.0303            1.33m\n",
      "        36 90081998170.5609 -1158130045.6684            1.25m\n",
      "        37 88479132259.2239 -1429276924.8154            1.16m\n",
      "        38 87176236713.3320 -1719264146.4293            1.08m\n",
      "        39 87903499113.3511 -1094664549.2542           59.55s\n"
     ]
    }
   ],
   "source": [
    "MLPReg = MLPRegressor(random_state=1, max_iter=500, verbose = True)\n",
    "\n",
    "scores = cross_validate(GBReg, X, Y,return_train_score = True, cv=5, scoring = scoring)\n",
    "print(f\"Entrenamiento r2 score: {np.mean(scores['train_r2'])}\")\n",
    "print(f\"Test r2 score: {np.mean(scores['test_r2'])}\")\n",
    "print(f\"Entrenamiento nMAPE score: {np.mean(scores['train_neg_mean_absolute_percentage_error'])}\")\n",
    "print(f\"Test nMAPE score: {np.mean(scores['test_neg_mean_absolute_percentage_error'])}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4ea7d8ee-b12b-487b-b0fe-9c496436c8d9",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "environment": {
   "kernel": "python3",
   "name": "common-cpu.m86",
   "type": "gcloud",
   "uri": "gcr.io/deeplearning-platform-release/base-cpu:m86"
  },
  "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": "3.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}