ann_regresion_pr_viv.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Importaciones"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Creación del dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Lectura de los datos históricos utilizando pandas\n",
    "df = pd.read_excel('./datasets/viviendas_Madrid.xlsx', sheet_name='Datos_Históricos')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 21742 entries, 0 to 21741\n",
      "Data columns (total 57 columns):\n",
      " #   Column                        Non-Null Count  Dtype  \n",
      "---  ------                        --------------  -----  \n",
      " 0   id                            21742 non-null  int64  \n",
      " 1   title                         21742 non-null  object \n",
      " 2   subtitle                      21742 non-null  object \n",
      " 3   sq_mt_built                   21616 non-null  float64\n",
      " 4   sq_mt_useful                  8228 non-null   float64\n",
      " 5   n_rooms                       21742 non-null  int64  \n",
      " 6   n_bathrooms                   21726 non-null  float64\n",
      " 7   n_floors                      1437 non-null   float64\n",
      " 8   sq_mt_allotment               1432 non-null   float64\n",
      " 9   latitude                      0 non-null      float64\n",
      " 10  longitude                     0 non-null      float64\n",
      " 11  raw_address                   16277 non-null  object \n",
      " 12  is_exact_address_hidden       21742 non-null  bool   \n",
      " 13  street_name                   15837 non-null  object \n",
      " 14  street_number                 6300 non-null   object \n",
      " 15  portal                        0 non-null      float64\n",
      " 16  floor                         19135 non-null  object \n",
      " 17  is_floor_under                20572 non-null  float64\n",
      " 18  door                          0 non-null      float64\n",
      " 19  neighborhood_id               21742 non-null  object \n",
      " 20  operation                     21742 non-null  object \n",
      " 21  rent_price                    21742 non-null  float64\n",
      " 22  rent_price_by_area            0 non-null      float64\n",
      " 23  is_rent_price_known           21742 non-null  bool   \n",
      " 24  buy_price                     21742 non-null  int64  \n",
      " 25  buy_price_by_area             21742 non-null  int64  \n",
      " 26  is_buy_price_known            21742 non-null  bool   \n",
      " 27  house_type_id                 21351 non-null  object \n",
      " 28  is_renewal_needed             21742 non-null  bool   \n",
      " 29  is_new_development            20750 non-null  float64\n",
      " 30  built_year                    10000 non-null  float64\n",
      " 31  has_central_heating           13608 non-null  float64\n",
      " 32  has_individual_heating        13608 non-null  float64\n",
      " 33  are_pets_allowed              0 non-null      float64\n",
      " 34  has_ac                        11211 non-null  float64\n",
      " 35  has_fitted_wardrobes          13399 non-null  float64\n",
      " 36  has_lift                      19356 non-null  float64\n",
      " 37  is_exterior                   18699 non-null  float64\n",
      " 38  has_garden                    1556 non-null   float64\n",
      " 39  has_pool                      5171 non-null   float64\n",
      " 40  has_terrace                   9548 non-null   float64\n",
      " 41  has_balcony                   3321 non-null   float64\n",
      " 42  has_storage_room              7698 non-null   float64\n",
      " 43  is_furnished                  0 non-null      float64\n",
      " 44  is_kitchen_equipped           0 non-null      float64\n",
      " 45  is_accessible                 4074 non-null   float64\n",
      " 46  has_green_zones               4057 non-null   float64\n",
      " 47  energy_certificate            21742 non-null  object \n",
      " 48  has_parking                   21742 non-null  bool   \n",
      " 49  has_private_parking           0 non-null      float64\n",
      " 50  has_public_parking            0 non-null      float64\n",
      " 51  is_parking_included_in_price  7719 non-null   float64\n",
      " 52  parking_price                 7719 non-null   float64\n",
      " 53  is_orientation_north          11358 non-null  float64\n",
      " 54  is_orientation_west           11358 non-null  float64\n",
      " 55  is_orientation_south          11358 non-null  float64\n",
      " 56  is_orientation_east           11358 non-null  float64\n",
      "dtypes: bool(5), float64(38), int64(4), object(10)\n",
      "memory usage: 8.7+ MB\n"
     ]
    }
   ],
   "source": [
    "# Con el método .info() mostramos la información del dataframe\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>title</th>\n",
       "      <th>subtitle</th>\n",
       "      <th>sq_mt_built</th>\n",
       "      <th>sq_mt_useful</th>\n",
       "      <th>n_rooms</th>\n",
       "      <th>n_bathrooms</th>\n",
       "      <th>n_floors</th>\n",
       "      <th>sq_mt_allotment</th>\n",
       "      <th>latitude</th>\n",
       "      <th>...</th>\n",
       "      <th>energy_certificate</th>\n",
       "      <th>has_parking</th>\n",
       "      <th>has_private_parking</th>\n",
       "      <th>has_public_parking</th>\n",
       "      <th>is_parking_included_in_price</th>\n",
       "      <th>parking_price</th>\n",
       "      <th>is_orientation_north</th>\n",
       "      <th>is_orientation_west</th>\n",
       "      <th>is_orientation_south</th>\n",
       "      <th>is_orientation_east</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>21742</td>\n",
       "      <td>Piso en venta en calle de Godella, 64</td>\n",
       "      <td>San CristÃ³bal, Madrid</td>\n",
       "      <td>64.0</td>\n",
       "      <td>60.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>21741</td>\n",
       "      <td>Piso en venta en calle de la del Manojo de Rosas</td>\n",
       "      <td>Los Ãngeles, Madrid</td>\n",
       "      <td>70.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>en trÃ¡mite</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>21740</td>\n",
       "      <td>Piso en venta en calle del Talco, 68</td>\n",
       "      <td>San AndrÃ©s, Madrid</td>\n",
       "      <td>94.0</td>\n",
       "      <td>54.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>no indicado</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>21739</td>\n",
       "      <td>Piso en venta en calle Pedro JimÃ©nez</td>\n",
       "      <td>San AndrÃ©s, Madrid</td>\n",
       "      <td>64.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>en trÃ¡mite</td>\n",
       "      <td>False</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>21738</td>\n",
       "      <td>Piso en venta en carretera de Villaverde a Val...</td>\n",
       "      <td>Los Rosales, Madrid</td>\n",
       "      <td>108.0</td>\n",
       "      <td>90.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>en trÃ¡mite</td>\n",
       "      <td>True</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 57 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      id                                              title  \\\n",
       "0  21742              Piso en venta en calle de Godella, 64   \n",
       "1  21741   Piso en venta en calle de la del Manojo de Rosas   \n",
       "2  21740               Piso en venta en calle del Talco, 68   \n",
       "3  21739              Piso en venta en calle Pedro JimÃ©nez   \n",
       "4  21738  Piso en venta en carretera de Villaverde a Val...   \n",
       "\n",
       "                 subtitle  sq_mt_built  sq_mt_useful  n_rooms  n_bathrooms  \\\n",
       "0  San CristÃ³bal, Madrid         64.0          60.0        2          1.0   \n",
       "1    Los Ãngeles, Madrid         70.0           NaN        3          1.0   \n",
       "2     San AndrÃ©s, Madrid         94.0          54.0        2          2.0   \n",
       "3     San AndrÃ©s, Madrid         64.0           NaN        2          1.0   \n",
       "4     Los Rosales, Madrid        108.0          90.0        2          2.0   \n",
       "\n",
       "   n_floors  sq_mt_allotment  latitude  ...  energy_certificate has_parking  \\\n",
       "0       NaN              NaN       NaN  ...                   D       False   \n",
       "1       NaN              NaN       NaN  ...         en trÃ¡mite       False   \n",
       "2       NaN              NaN       NaN  ...         no indicado       False   \n",
       "3       NaN              NaN       NaN  ...         en trÃ¡mite       False   \n",
       "4       NaN              NaN       NaN  ...         en trÃ¡mite        True   \n",
       "\n",
       "   has_private_parking has_public_parking is_parking_included_in_price  \\\n",
       "0                  NaN                NaN                          NaN   \n",
       "1                  NaN                NaN                          NaN   \n",
       "2                  NaN                NaN                          NaN   \n",
       "3                  NaN                NaN                          NaN   \n",
       "4                  NaN                NaN                          1.0   \n",
       "\n",
       "   parking_price is_orientation_north  is_orientation_west  \\\n",
       "0            NaN                  0.0                  1.0   \n",
       "1            NaN                  NaN                  NaN   \n",
       "2            NaN                  NaN                  NaN   \n",
       "3            NaN                  0.0                  0.0   \n",
       "4            0.0                  1.0                  1.0   \n",
       "\n",
       "   is_orientation_south is_orientation_east  \n",
       "0                   0.0                 0.0  \n",
       "1                   NaN                 NaN  \n",
       "2                   NaN                 NaN  \n",
       "3                   1.0                 0.0  \n",
       "4                   1.0                 1.0  \n",
       "\n",
       "[5 rows x 57 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Con el método .head(n) mostramos los primeros n renglones del dataframe\n",
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "id                                  0\n",
       "title                               0\n",
       "subtitle                            0\n",
       "sq_mt_built                       126\n",
       "sq_mt_useful                    13514\n",
       "n_rooms                             0\n",
       "n_bathrooms                        16\n",
       "n_floors                        20305\n",
       "sq_mt_allotment                 20310\n",
       "latitude                        21742\n",
       "longitude                       21742\n",
       "raw_address                      5465\n",
       "is_exact_address_hidden             0\n",
       "street_name                      5905\n",
       "street_number                   15442\n",
       "portal                          21742\n",
       "floor                            2607\n",
       "is_floor_under                   1170\n",
       "door                            21742\n",
       "neighborhood_id                     0\n",
       "operation                           0\n",
       "rent_price                          0\n",
       "rent_price_by_area              21742\n",
       "is_rent_price_known                 0\n",
       "buy_price                           0\n",
       "buy_price_by_area                   0\n",
       "is_buy_price_known                  0\n",
       "house_type_id                     391\n",
       "is_renewal_needed                   0\n",
       "is_new_development                992\n",
       "built_year                      11742\n",
       "has_central_heating              8134\n",
       "has_individual_heating           8134\n",
       "are_pets_allowed                21742\n",
       "has_ac                          10531\n",
       "has_fitted_wardrobes             8343\n",
       "has_lift                         2386\n",
       "is_exterior                      3043\n",
       "has_garden                      20186\n",
       "has_pool                        16571\n",
       "has_terrace                     12194\n",
       "has_balcony                     18421\n",
       "has_storage_room                14044\n",
       "is_furnished                    21742\n",
       "is_kitchen_equipped             21742\n",
       "is_accessible                   17668\n",
       "has_green_zones                 17685\n",
       "energy_certificate                  0\n",
       "has_parking                         0\n",
       "has_private_parking             21742\n",
       "has_public_parking              21742\n",
       "is_parking_included_in_price    14023\n",
       "parking_price                   14023\n",
       "is_orientation_north            10384\n",
       "is_orientation_west             10384\n",
       "is_orientation_south            10384\n",
       "is_orientation_east             10384\n",
       "dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Con los métodos .isnull().sum() verificamos cuantos valores nulos tenemos en cada columna\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Limpieza **inicial** del dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>title</th>\n",
       "      <th>subtitle</th>\n",
       "      <th>sq_mt_built</th>\n",
       "      <th>n_rooms</th>\n",
       "      <th>n_bathrooms</th>\n",
       "      <th>is_exact_address_hidden</th>\n",
       "      <th>is_floor_under</th>\n",
       "      <th>neighborhood_id</th>\n",
       "      <th>operation</th>\n",
       "      <th>rent_price</th>\n",
       "      <th>is_rent_price_known</th>\n",
       "      <th>buy_price</th>\n",
       "      <th>buy_price_by_area</th>\n",
       "      <th>is_buy_price_known</th>\n",
       "      <th>house_type_id</th>\n",
       "      <th>is_renewal_needed</th>\n",
       "      <th>is_new_development</th>\n",
       "      <th>energy_certificate</th>\n",
       "      <th>has_parking</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>21742</td>\n",
       "      <td>Piso en venta en calle de Godella, 64</td>\n",
       "      <td>San CristÃ³bal, Madrid</td>\n",
       "      <td>64.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Neighborhood 135: San CristÃ³bal (1308.89 â‚¬/...</td>\n",
       "      <td>sale</td>\n",
       "      <td>471.0</td>\n",
       "      <td>False</td>\n",
       "      <td>85000</td>\n",
       "      <td>1328</td>\n",
       "      <td>True</td>\n",
       "      <td>HouseType 1: Pisos</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>D</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>21741</td>\n",
       "      <td>Piso en venta en calle de la del Manojo de Rosas</td>\n",
       "      <td>Los Ãngeles, Madrid</td>\n",
       "      <td>70.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>True</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Neighborhood 132: Los Ãngeles (1796.68 â‚¬/m2...</td>\n",
       "      <td>sale</td>\n",
       "      <td>666.0</td>\n",
       "      <td>False</td>\n",
       "      <td>129900</td>\n",
       "      <td>1856</td>\n",
       "      <td>True</td>\n",
       "      <td>HouseType 1: Pisos</td>\n",
       "      <td>True</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en trÃ¡mite</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>21740</td>\n",
       "      <td>Piso en venta en calle del Talco, 68</td>\n",
       "      <td>San AndrÃ©s, Madrid</td>\n",
       "      <td>94.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Neighborhood 134: San AndrÃ©s (1617.18 â‚¬/m2)...</td>\n",
       "      <td>sale</td>\n",
       "      <td>722.0</td>\n",
       "      <td>False</td>\n",
       "      <td>144247</td>\n",
       "      <td>1535</td>\n",
       "      <td>True</td>\n",
       "      <td>HouseType 1: Pisos</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>no indicado</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>21739</td>\n",
       "      <td>Piso en venta en calle Pedro JimÃ©nez</td>\n",
       "      <td>San AndrÃ©s, Madrid</td>\n",
       "      <td>64.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>True</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Neighborhood 134: San AndrÃ©s (1617.18 â‚¬/m2)...</td>\n",
       "      <td>sale</td>\n",
       "      <td>583.0</td>\n",
       "      <td>False</td>\n",
       "      <td>109900</td>\n",
       "      <td>1717</td>\n",
       "      <td>True</td>\n",
       "      <td>HouseType 1: Pisos</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en trÃ¡mite</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>21738</td>\n",
       "      <td>Piso en venta en carretera de Villaverde a Val...</td>\n",
       "      <td>Los Rosales, Madrid</td>\n",
       "      <td>108.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>True</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Neighborhood 133: Los Rosales (1827.79 â‚¬/m2)...</td>\n",
       "      <td>sale</td>\n",
       "      <td>1094.0</td>\n",
       "      <td>False</td>\n",
       "      <td>260000</td>\n",
       "      <td>2407</td>\n",
       "      <td>True</td>\n",
       "      <td>HouseType 1: Pisos</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en trÃ¡mite</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id                                              title  \\\n",
       "0  21742              Piso en venta en calle de Godella, 64   \n",
       "1  21741   Piso en venta en calle de la del Manojo de Rosas   \n",
       "2  21740               Piso en venta en calle del Talco, 68   \n",
       "3  21739              Piso en venta en calle Pedro JimÃ©nez   \n",
       "4  21738  Piso en venta en carretera de Villaverde a Val...   \n",
       "\n",
       "                 subtitle  sq_mt_built  n_rooms  n_bathrooms  \\\n",
       "0  San CristÃ³bal, Madrid         64.0        2          1.0   \n",
       "1    Los Ãngeles, Madrid         70.0        3          1.0   \n",
       "2     San AndrÃ©s, Madrid         94.0        2          2.0   \n",
       "3     San AndrÃ©s, Madrid         64.0        2          1.0   \n",
       "4     Los Rosales, Madrid        108.0        2          2.0   \n",
       "\n",
       "   is_exact_address_hidden  is_floor_under  \\\n",
       "0                    False             0.0   \n",
       "1                     True             0.0   \n",
       "2                    False             0.0   \n",
       "3                     True             1.0   \n",
       "4                     True             0.0   \n",
       "\n",
       "                                     neighborhood_id operation  rent_price  \\\n",
       "0  Neighborhood 135: San CristÃ³bal (1308.89 â‚¬/...      sale       471.0   \n",
       "1  Neighborhood 132: Los Ãngeles (1796.68 â‚¬/m2...      sale       666.0   \n",
       "2  Neighborhood 134: San AndrÃ©s (1617.18 â‚¬/m2)...      sale       722.0   \n",
       "3  Neighborhood 134: San AndrÃ©s (1617.18 â‚¬/m2)...      sale       583.0   \n",
       "4  Neighborhood 133: Los Rosales (1827.79 â‚¬/m2)...      sale      1094.0   \n",
       "\n",
       "   is_rent_price_known  buy_price  buy_price_by_area  is_buy_price_known  \\\n",
       "0                False      85000               1328                True   \n",
       "1                False     129900               1856                True   \n",
       "2                False     144247               1535                True   \n",
       "3                False     109900               1717                True   \n",
       "4                False     260000               2407                True   \n",
       "\n",
       "        house_type_id  is_renewal_needed  is_new_development  \\\n",
       "0  HouseType 1: Pisos              False                 0.0   \n",
       "1  HouseType 1: Pisos               True                 0.0   \n",
       "2  HouseType 1: Pisos              False                 0.0   \n",
       "3  HouseType 1: Pisos              False                 0.0   \n",
       "4  HouseType 1: Pisos              False                 0.0   \n",
       "\n",
       "  energy_certificate  has_parking  \n",
       "0                  D        False  \n",
       "1        en trÃ¡mite        False  \n",
       "2        no indicado        False  \n",
       "3        en trÃ¡mite        False  \n",
       "4        en trÃ¡mite         True  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "Si una columna tiene un alto porcentaje de valores nulos, no conviene utilizarla como variable\n",
    "\"\"\"\n",
    "# Eliminamos las variables con alto porcentaje de valores nulos\n",
    "\n",
    "#   Creamos una serie de pandas con el número de valores nulos de cada variable\n",
    "\"\"\"\n",
    "    Una serie es un dataframe de una sola columna. En este caso creamos las cabeceras con su respectiva cantidad de valores nulos.\n",
    "\"\"\"\n",
    "df_var = df.isnull().sum()\n",
    "\n",
    "#   Establecemos un porcentaje como límite para valores nulos (10%)\n",
    "pcje_eliminacion = .1\n",
    "\n",
    "#   Dejamos unicamente las variables que tienen un porcentaje de valores nulos menor al límite establecido\n",
    "df_var = df_var[df_var<pcje_eliminacion*len(df)]\n",
    "\n",
    "#   Creamos una lista con los nombres de las variables que nos interesan (los índices de de la serie)\n",
    "lista_variables_OK = df_var.index\n",
    "\n",
    "#   Filtramos el dataframe original y nos quedamos unicamente con las columnas que cumplen nuestros requisitos\n",
    "df = df[lista_variables_OK]\n",
    "\n",
    "# Vemos como va quedando el dataframe\n",
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 21742 entries, 0 to 21741\n",
      "Data columns (total 20 columns):\n",
      " #   Column                   Non-Null Count  Dtype  \n",
      "---  ------                   --------------  -----  \n",
      " 0   id                       21742 non-null  int64  \n",
      " 1   title                    21742 non-null  object \n",
      " 2   subtitle                 21742 non-null  object \n",
      " 3   sq_mt_built              21616 non-null  float64\n",
      " 4   n_rooms                  21742 non-null  int64  \n",
      " 5   n_bathrooms              21726 non-null  float64\n",
      " 6   is_exact_address_hidden  21742 non-null  bool   \n",
      " 7   is_floor_under           20572 non-null  float64\n",
      " 8   neighborhood_id          21742 non-null  object \n",
      " 9   operation                21742 non-null  object \n",
      " 10  rent_price               21742 non-null  float64\n",
      " 11  is_rent_price_known      21742 non-null  bool   \n",
      " 12  buy_price                21742 non-null  int64  \n",
      " 13  buy_price_by_area        21742 non-null  int64  \n",
      " 14  is_buy_price_known       21742 non-null  bool   \n",
      " 15  house_type_id            21351 non-null  object \n",
      " 16  is_renewal_needed        21742 non-null  bool   \n",
      " 17  is_new_development       20750 non-null  float64\n",
      " 18  energy_certificate       21742 non-null  object \n",
      " 19  has_parking              21742 non-null  bool   \n",
      "dtypes: bool(5), float64(5), int64(4), object(6)\n",
      "memory usage: 2.6+ MB\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "id                            0\n",
       "title                         0\n",
       "subtitle                      0\n",
       "sq_mt_built                 126\n",
       "n_rooms                       0\n",
       "n_bathrooms                  16\n",
       "is_exact_address_hidden       0\n",
       "is_floor_under             1170\n",
       "neighborhood_id               0\n",
       "operation                     0\n",
       "rent_price                    0\n",
       "is_rent_price_known           0\n",
       "buy_price                     0\n",
       "buy_price_by_area             0\n",
       "is_buy_price_known            0\n",
       "house_type_id               391\n",
       "is_renewal_needed             0\n",
       "is_new_development          992\n",
       "energy_certificate            0\n",
       "has_parking                   0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Si revisamos el dataframe sólo quedan las columnas con una cantidad inferior al 10% de valores nulos\n",
    "df.info()\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 19310 entries, 0 to 21740\n",
      "Data columns (total 20 columns):\n",
      " #   Column                   Non-Null Count  Dtype  \n",
      "---  ------                   --------------  -----  \n",
      " 0   id                       19310 non-null  int64  \n",
      " 1   title                    19310 non-null  object \n",
      " 2   subtitle                 19310 non-null  object \n",
      " 3   sq_mt_built              19310 non-null  float64\n",
      " 4   n_rooms                  19310 non-null  int64  \n",
      " 5   n_bathrooms              19310 non-null  float64\n",
      " 6   is_exact_address_hidden  19310 non-null  bool   \n",
      " 7   is_floor_under           19310 non-null  float64\n",
      " 8   neighborhood_id          19310 non-null  object \n",
      " 9   operation                19310 non-null  object \n",
      " 10  rent_price               19310 non-null  float64\n",
      " 11  is_rent_price_known      19310 non-null  bool   \n",
      " 12  buy_price                19310 non-null  int64  \n",
      " 13  buy_price_by_area        19310 non-null  int64  \n",
      " 14  is_buy_price_known       19310 non-null  bool   \n",
      " 15  house_type_id            19310 non-null  object \n",
      " 16  is_renewal_needed        19310 non-null  bool   \n",
      " 17  is_new_development       19310 non-null  float64\n",
      " 18  energy_certificate       19310 non-null  object \n",
      " 19  has_parking              19310 non-null  bool   \n",
      "dtypes: bool(5), float64(5), int64(4), object(6)\n",
      "memory usage: 2.4+ MB\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "id                         0\n",
       "title                      0\n",
       "subtitle                   0\n",
       "sq_mt_built                0\n",
       "n_rooms                    0\n",
       "n_bathrooms                0\n",
       "is_exact_address_hidden    0\n",
       "is_floor_under             0\n",
       "neighborhood_id            0\n",
       "operation                  0\n",
       "rent_price                 0\n",
       "is_rent_price_known        0\n",
       "buy_price                  0\n",
       "buy_price_by_area          0\n",
       "is_buy_price_known         0\n",
       "house_type_id              0\n",
       "is_renewal_needed          0\n",
       "is_new_development         0\n",
       "energy_certificate         0\n",
       "has_parking                0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Para lograr un dataframe limpio eliminamos todos los registros que tengan algún valor nulo (Método más sencillo, algo mas complejo sería interpolar)\n",
    "df.dropna(inplace=True)\n",
    "\n",
    "# Verificamos que no hayan quedado valores nulos en el dataframe\n",
    "df.info()\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <td>19310.0</td>\n",
       "      <td>10751.771621</td>\n",
       "      <td>6272.883675</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5347.25</td>\n",
       "      <td>10575.5</td>\n",
       "      <td>16243.75</td>\n",
       "      <td>21742.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sq_mt_built</th>\n",
       "      <td>19310.0</td>\n",
       "      <td>144.126359</td>\n",
       "      <td>125.418617</td>\n",
       "      <td>16.0</td>\n",
       "      <td>71.00</td>\n",
       "      <td>101.0</td>\n",
       "      <td>161.00</td>\n",
       "      <td>999.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_rooms</th>\n",
       "      <td>19310.0</td>\n",
       "      <td>3.013361</td>\n",
       "      <td>1.388427</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.00</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.00</td>\n",
       "      <td>24.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>n_bathrooms</th>\n",
       "      <td>19310.0</td>\n",
       "      <td>2.039617</td>\n",
       "      <td>1.271573</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.00</td>\n",
       "      <td>14.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>is_floor_under</th>\n",
       "      <td>19310.0</td>\n",
       "      <td>0.118747</td>\n",
       "      <td>0.323499</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rent_price</th>\n",
       "      <td>19310.0</td>\n",
       "      <td>1573.749995</td>\n",
       "      <td>931.244378</td>\n",
       "      <td>18.0</td>\n",
       "      <td>913.00</td>\n",
       "      <td>1368.0</td>\n",
       "      <td>2073.50</td>\n",
       "      <td>12584.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>buy_price</th>\n",
       "      <td>19310.0</td>\n",
       "      <td>632592.345106</td>\n",
       "      <td>720196.584637</td>\n",
       "      <td>36000.0</td>\n",
       "      <td>200000.00</td>\n",
       "      <td>380000.0</td>\n",
       "      <td>750000.00</td>\n",
       "      <td>8800000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>buy_price_by_area</th>\n",
       "      <td>19310.0</td>\n",
       "      <td>4061.935008</td>\n",
       "      <td>1917.230783</td>\n",
       "      <td>447.0</td>\n",
       "      <td>2586.00</td>\n",
       "      <td>3783.5</td>\n",
       "      <td>5047.00</td>\n",
       "      <td>18889.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>is_new_development</th>\n",
       "      <td>19310.0</td>\n",
       "      <td>0.060124</td>\n",
       "      <td>0.237723</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      count           mean            std      min        25%  \\\n",
       "id                  19310.0   10751.771621    6272.883675      2.0    5347.25   \n",
       "sq_mt_built         19310.0     144.126359     125.418617     16.0      71.00   \n",
       "n_rooms             19310.0       3.013361       1.388427      0.0       2.00   \n",
       "n_bathrooms         19310.0       2.039617       1.271573      1.0       1.00   \n",
       "is_floor_under      19310.0       0.118747       0.323499      0.0       0.00   \n",
       "rent_price          19310.0    1573.749995     931.244378     18.0     913.00   \n",
       "buy_price           19310.0  632592.345106  720196.584637  36000.0  200000.00   \n",
       "buy_price_by_area   19310.0    4061.935008    1917.230783    447.0    2586.00   \n",
       "is_new_development  19310.0       0.060124       0.237723      0.0       0.00   \n",
       "\n",
       "                         50%        75%        max  \n",
       "id                   10575.5   16243.75    21742.0  \n",
       "sq_mt_built            101.0     161.00      999.0  \n",
       "n_rooms                  3.0       4.00       24.0  \n",
       "n_bathrooms              2.0       2.00       14.0  \n",
       "is_floor_under           0.0       0.00        1.0  \n",
       "rent_price            1368.0    2073.50    12584.0  \n",
       "buy_price           380000.0  750000.00  8800000.0  \n",
       "buy_price_by_area     3783.5    5047.00    18889.0  \n",
       "is_new_development       0.0       0.00        1.0  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Vemos los datos estadísticos principales del dataframe con los métodos .describe().transpose() (Buena práctica)\n",
    "df.describe().transpose()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualización de estadísticas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plots básicos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\pety2\\AppData\\Local\\Temp\\ipykernel_2920\\487558491.py:7: UserWarning: \n",
      "\n",
      "`distplot` is a deprecated function and will be removed in seaborn v0.14.0.\n",
      "\n",
      "Please adapt your code to use either `displot` (a figure-level function with\n",
      "similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "\n",
      "For a guide to updating your code to use the new functions, please see\n",
      "https://gist.github.com/mwaskom/de44147ed2974457ad6372750bbe5751\n",
      "\n",
      "  sns.distplot(df['buy_price'])\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<AxesSubplot: xlabel='buy_price', ylabel='Density'>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1700x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot de distribución de la variable objetivo para ver los outliers y definir si es necesario descartarlos\n",
    "#   Creamos una figura con matplotlib\n",
    "plt.figure(figsize=(17,8))\n",
    "\n",
    "#   Insertamos un distplot de seaborn de la variable objetivo en la figura creada\n",
    "sns.set(style='darkgrid')\n",
    "sns.distplot(df['buy_price'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot: ylabel='count'>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Hacemos un countplot sobre la variable de cantidad de habitaciones\n",
    "sns.countplot(df['n_rooms'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "id                        -0.195828\n",
       "is_floor_under            -0.149768\n",
       "is_new_development        -0.013903\n",
       "is_renewal_needed          0.052132\n",
       "is_exact_address_hidden    0.167964\n",
       "has_parking                0.315350\n",
       "n_rooms                    0.578291\n",
       "buy_price_by_area          0.581698\n",
       "n_bathrooms                0.753692\n",
       "sq_mt_built                0.831700\n",
       "rent_price                 0.920350\n",
       "buy_price                  1.000000\n",
       "is_rent_price_known             NaN\n",
       "is_buy_price_known              NaN\n",
       "Name: buy_price, dtype: float64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Usamos el método .corr() para verificar qué variables tienen correlación con la variable objetivo y la ordenamos de menor a mayor\n",
    "#   Los datos negativos tienen una correlación inversamente proporcional\n",
    "df.corr()['buy_price'].sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot: xlabel='buy_price', ylabel='sq_mt_built'>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualizamos las variables altamente correlacionadas con la variable objetivo\n",
    "\n",
    "#   Relación precio de venta - metros cuadrados construidos\n",
    "plt.figure(figsize=(12,8))\n",
    "sns.scatterplot(x='buy_price', y='sq_mt_built', data=df)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot: xlabel='n_bathrooms', ylabel='buy_price'>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAi4AAAHECAYAAAAefaGmAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy89olMNAAAACXBIWXMAAA9hAAAPYQGoP6dpAABfIklEQVR4nO3deXxM9/4/8NeZLBMVqVgjVDcNKokk9tIWVb+WJBVdlUZVUNRyUepbquW6VFtLxXpDUEWvIhLdN7durz2EVAkaWpJYExGSiZn5/P5wM82Yk222c07yej4eeTzM+ZyceZlJTt7zOZ/z+UhCCAEiIiIiDdApHYCIiIiosli4EBERkWawcCEiIiLNYOFCREREmsHChYiIiDSDhQsRERFpBgsXIiIi0gwWLkRERKQZLFyIiIhIM2pE4bJixQq88sorVf6+pKQk9OnTByEhIejbty+++uorF6QjIiKiyqr2hcunn36KhQsXVvn7tm/fjrfffhsDBw7EF198gcjISEyYMAGHDh1yfkgiIiKqFE+lA7jKhQsXMGPGDOzduxf33Xdflb5XCIFFixYhNjYWAwcOBACMHDkSBw4cwL59+xAeHu6CxERERFSRatvj8uuvv8LLywvJyclo27atTftPP/2E/v37IzQ0FE8++SQWLlyI4uJiAEBmZibOnz+PqKgoq+9ZtWoVRowY4Zb8REREZKva9rj07NkTPXv2lG37+eefMX78eEydOhWPPPII/vjjD8yaNQuZmZlYtGgRMjMzAQA3b97E0KFDcezYMTRr1gwjR44s85hERETketW2x6U8y5cvxwsvvICXXnoJzZs3R7du3fDee+/h66+/xrlz51BQUAAAmDJlCiIjI7F69Wp07doVo0aNwu7duxVOT0REVHNV2x6X8hw7dgxHjhzB559/btkmhAAAnD59Gl5eXgCAoUOHIiYmBgDQunVrHDt2DImJiejSpYv7QxMREVHNLFzMZjPi4uIsRUlpDRs2xG+//QYACAoKsmpr0aIFdu7c6Y6IREREJKNGXip66KGHkJmZiXvvvdfylZOTg3nz5uHGjRto06YNateujbS0NKvvy8jIQPPmzRVKTURERDWyx2XYsGEYP3484uPj0bdvX+Tk5ODtt99Gs2bN0LBhQwBAXFwclixZgsaNGyM0NBRffPEFfvnlF6xZs0bZ8ERERDVYjSxcnnrqKSxYsAArVqzA8uXLUbduXfTs2ROTJk2y7DNq1CjUqlULCxYswIULF/Dggw9i8eLF6NSpk4LJiYiIajZJlIxKJSIiIlK5GjnGhYiIiLSJhQsRERFpRrUc4yKEgNnMK2BERERaodNJkCSpwv2qZeFiNgtcvXpD6RhERERUSfXq1YaHR8WFCy8VERERkWawcCEiIiLNYOFCREREmsHChYiIiDSDhQsRERFpBgsXIiIi0gwWLkRERKQZLFyIiIhIM1i4EBERkWawcCEiIiLNYOFCRC6RmnoA48ePRGrqAaWjEFE1wsKFiJzOYDAgMXElLl++hMTElTAYDEpHIqJqgoULETldcvJW5OXlAgDy8nKRkrJN4UREVF2wcCEip8rJycaOHUkQQgAAhBBISdmGnJxshZMRUXXAwoWoFK2My9i8eSNiY1/A5s0blY5iRQiBtWsTLEVLRduVppX3Wys5idyBhQvR/2hlXEZ+fj6Sk7fCbDYjOXkr8vPzlY5kkZV1HkePpsFsNlttN5vNOHo0DVlZ5xVKZksr77dWchK5CwsXov/RyriMhQvnWV2GWbToA4UT/SUwsClCQtpCp7M+teh0OoSGhiEwsKlCyWxp5f3WSk4id2HhQgTtjMtITz+CjIzjVttOnPgN6elHFEpkTZIkDB4cB0mSKrVdKVp5v7WSk8idWLhQjaeVcRlmsxnx8fNl2+Lj59tcnlFKQEATREb2sxQpkiQhKioGjRsHKJzsNq2831rJSeRuLFyoxtPKuIzDh1NRUFAg21ZQUIDDh1PdnKhs0dH9UbeuPwDA378eoqJiFE70F62831rJSeRuLFyoxtPKuIywsAj4+vrKtvn61kFYWISbE5VNr9djyJDhaNCgIV59dRj0er3SkSy08n5rJSeRu7FwoRpPK+MydDod3nhjgmzbmDETbP7AKS0ioj0WLlyGiIj2SkexopX3Wys5idxNXWc6IoWofVxGieDgUAQFtbLa1rJla7RpE6JQIm3SyvutlZxE7sTCheh/1Dwuo7RRo8ZZPR45cqxCSbRNK++3VnISuQsLF6L/UfO4jNJ++ul7q8c7d/6gUBJt08r7rZWcRO4iiWp4T53JZMbVqzeUjkHkdDk52ZgyZTxMJpNlm4eHB95/fyECApoomIyIyDH16tWGh0fF/SnscSHSCM7rQUTEwoVIMzivBxERCxcizeC8HkRELFyINIPzehARsXAh0hTO60FENR0LFyKN4bweRFSTsXAh0hjO60FENRnncSEiIiLFcR4XIiIiqnZYuBAREZFmsHAhIiIizWDhQkRERJrBwoWIiIg0g4ULERERaQYLFyIiItIMVRQuRqMRixYtQo8ePRAeHo6BAwfi8OHDSsciUq3U1AMYP34kUlMPKB2FiMitVFG4LFu2DJs3b8asWbOQlJSE+++/H3Fxcbh48aLS0YhUx2AwIDFxJS5fvoTExJUwGAxKRyIichtVFC7ff/89IiMj0a1bN9x777146623cP36dfa6EMlITt6KvLxcAEBeXi5SUrYpnIiIyH1UUbjUr18fP/30E86dOweTyYTPPvsM3t7eaNWqldLRiFQlJycbO3YkoWSlDiEEUlK2IScnW+FkRETuoYq1ik6fPo1x48bh5MmT8PDwgE6nw+LFi9GjRw+7jmcymZGfX+jklETKEkJgzpxZSE8/ArPZbNmu0+kQHByKqVOnQ5IkBRNaO3hwPxITEzBkSBzateugdBxZM2b8H06cOI6WLVvhvff+oXQcohrNz69WpdYq8nRDlgqdOnUKderUwZIlS9C4cWNs3rwZkyZNwvr169G6desqH0+nk+DvX9sFSYmU88cff+DIkcM2281mM44cOYyCgqto3ry5+4PJKCoqwurVK3HlyhUkJv4T3bp1ho+Pj9KxrPz55584ceI4AODEieMoKLiKe+65R+FURFQRxXtcsrOz8eSTT2LNmjVo3769ZfvLL7+MunXrYunSpVU+JntcqDoqr8clJKQt3nprmmp6XD77bAOSkrZACAFJkhAT8xxeeGGA0rGsDBkyEIWFf50natWqhcTETxVMRFSzaabHJS0tDbdu3UJISIjV9rZt2+Lnn3+2+7hGo7ninYg0JjZ2KKZMGW+1TZIkxMYOhckkACh+5Rc5OdlITt5mNQ5n+/ateOSRxxAQ0EThdLft2JFkVbQAQGFhIZKStiIysp8yoYioUhQfnBsQEAAAOHHihNX2jIwM3HfffQokIlKvgIAmiIzsZ+lZkSQJUVExaNw4QOFktwkhsHZtAu7syC1ruxKMRiM2bVov27Zp03oYjUY3JyKiqlC8cAkNDUW7du0wZcoU7NmzB2fOnMHChQuxe/duDB8+XOl4RKoTHd0fdev6AwD8/eshKipG4UR/yco6j6NH06wuZQG3x+EcPZqGrKzzCiX7y7Ztmx1qJyJlKV646HQ6LFu2DJ07d8bUqVPRv39/7NmzB2vWrEHbtm2VjkekOnq9HkOGDEeDBg3x6qvDoNfrlY5kERjYFCEhbaHTWZ9adDodQkPDEBjYVKFkf4mJed6hdiJSluKDc13BZDLj6tUbSscgqpFycrIxZcp4mEwmyzYPDw/Mm7dINZe0duxIkr1cNGBALPr2jVYgERHVq1e7UoNzFe9xIaLqRe3jcAAgMrIffHxqWW2rVasWixYiDWDhQkROp+ZxOCVmzpxr9fi99+aWsScRqQkLFyJyOjWPwykRGNgUQUEtAQBBQS1VMf6GiCrGMS5ERESkOI5xISIiomqHhQsRERFpBgsXIiIi0gwWLkRERKQZLFyIiIhIM1i4EBERkWawcCEiIiLNYOFCREREmsHChYiIiDSDhQsRERFpBgsXIiIi0gwWLkRERKQZLFyIiIhIM1i4EBERkWawcCEiIiLNYOFCREREmsHChYiIiDSDhQsRERFpBgsXIiIi0gwWLkRERKQZLFyIiIhIM1i4EBERkWawcCEiIiLNYOFCREREmsHChYiIiDSDhQsRERFpBgsXolJSUw9g/PiRSE09oHQUIiKSwcKF6H8MBgMSE1fi8uVLSExcCYPBoHQkIiK6AwsXov9JTt6KvLxcAEBeXi5SUrYpnIiIiO7EwoUIQE5ONnbsSIIQAgAghEBKyjbk5GQrnIyIiEpj4UI1nhACa9cmWIqWirZT5WzevBGxsS9g8+aNSkchomqEhQvVeFlZ53H0aBrMZrPVdrPZjKNH05CVdV6hZNqVn5+P5OStMJvNSE7eivz8fKUjEVE1wcKFarzAwKYICWkLnc7610Gn0yE0NAyBgU0VSqZdCxfOs7rstmjRBwonIqLqgoUL1XiSJGHw4DhIklSp7VS+9PQjyMg4brXtxInfkJ5+RKFERFSdsHAhAhAQ0ASRkf0sRYokSYiKikHjxgEKJ9MWs9mM+Pj5sm3x8fNtLscREVUVCxei/4mO7o+6df0BAP7+9RAVFaNwIu05fDgVBQUFsm0FBQU4fDjVzYmIqLph4UL0P3q9HkOGDEeDBg3x6qvDoNfrlY6kOWFhEfD19ZVt8/Wtg7CwCDcnIqLqRhLV8F5Pk8mMq1dvKB2DqEZKTz+CuXNn2myfOnUG2rQJUSAREWlBvXq14eFRcX8Ke1yIyKmCg0MRFNTKalvLlq1ZtBCRU7BwISKnGz9+cqmBzjqMG/emwomIqLpg4UJETufn54fo6P7Q6XSIjo6Bn5+f0pGIqJrgGBciqrFSUw9g3bpViI0dioiI9krHKZNWchI5gmNciIjKYTAYkJi4EpcvX0Ji4koYDAalI8nSSk4id2HhQkQ1UnLyVuTl5QIA8vJykZKyTeFE8rSSk8hdWLgQUY2Tk5ONHTuSrNZTSknZhpycbIWTWdNKTiJ3YuFCRDWKEAJr1ybgzuF9ZW1XilZyErkbCxciqlGyss7j6NE0m3WTzGYzjh5NQ1bWeYWSWdNKTiJ3Y+FCRDVKYGBThIS0hU5nffrT6XQIDQ1DYGBThZJZ00pOIndj4UJENYokSRg8OM4yQV5F25WilZxE7sbChYhqnICAJoiM7Fdqdl8JUVExaNw4QOFk1rSSk8idWLgQUY0UHd0fdev6AwD8/eshKipG4UTytJKTyF1YuBBRjaTX6zFkyHA0aNAQr746DHq9XulIsrSSk8hdOOU/ERERKY5T/hMREVG1w8KFiIiINIOFCxEREWkGCxciIiLSDBYuREREpBksXIiIiEgzWLgQERGRZrBwISIiIs1QReGSlJSEPn36ICQkBH379sVXX32ldCRygdTUAxg/fiRSUw8oHUXztPBabt68EbGxL2Dz5o1KRyGiakTxwmX79u14++23MXDgQHzxxReIjIzEhAkTcOjQIaWjkRMZDAYkJq7E5cuXkJi4EgaDQelImqWF1zI/Px/JyVthNpuRnLwV+fn5SkciompC0cJFCIFFixYhNjYWAwcORPPmzTFy5Eg88sgj2Ldvn5LRyMmSk7ciLy8XAJCXl4uUlG0KJ9IuLbyWCxfOQ8lqIrd/zz9QOBERVReKFi6ZmZk4f/48oqKirLavWrUKI0aMUCgVOVtOTjZ27Eiy+kOWkrINOTnZCifTHi28lunpR5CRcdxq24kTvyE9/YhCiYioOvFU8skzMzMBADdv3sTQoUNx7NgxNGvWDCNHjkTPnj0dOranp+JXwQi3/7CuW7cKd67lWbJ96tTpkCRJoXTaooXX0mw2Iz5+gWxbfPwCrFyZCJ2Ov5tEZD9FC5eCggIAwJQpU/DGG29g0qRJ+OabbzBq1CgkJiaiS5cudh1Xp5Pg71/bmVHJTn/88QeOHDlss91sNuPIkcMoKLiK5s2buz+YBmnhtdy7dy8KCq7LthUUXMepU8fQqVMnN6cioupE0cLFy8sLADB06FDExMQAAFq3bo1jx445VLiYzQL5+TedlpPs5+tbD6GhYUhPPwKz2WzZrtPpEBLSFr6+9ZCbe0PBhNqhhdeyRYuH4etbR7Z4qVOnDlq0eFjxjESkTn5+teDhUXGPrKKFS+PGjQEAQUFBVttbtGiBnTt3OnRso9Fc8U7kFrGxQzFlynirbZIkITZ2KEwmAUDIfh/Z0sJr+cYbf8PcuTNltk+A2Qyroktpixd/hL17d6NTpy4YM2ai0nGIqBIUvdjcpk0b1K5dG2lpaVbbMzIyFO/yJucJCGiCyMh+lvEXkiQhKioGjRsHKJxMe7TwWgYHhyIoqJXVtpYtW6NNmxCFEsm7fPkS9u7dDQDYu3c3Ll++pHAiIqoMRQsXHx8fxMXFYcmSJdixYwf++OMPLFu2DL/88guGDBmiZDRysujo/qhb1x8A4O9fD1FRMQon0i4tvJbjx08uVVzpMG7cmwonsjVz5jSrx7NmTVcoCRFVheLD+0eNGoUxY8ZgwYIF6NOnD77++mssXryYA/iqGb1ejyFDhqNBg4Z49dVh0Ov1SkfSLC28ln5+foiO7g+dTofo6Bj4+fkpHcnKrl07cfXqFattV65cxq5dO90fhoiqRBJ33ltZDZhMZly9ygGARGTLZDLhtddehslksmnz8PDA6tUb4OHhoUAyopqtXr3alRqcq3iPCxGRO/3443eyRQtwu6j58cfv3JyIiKqChQsR1Sg9ez5ZZo+Kh4cHevZ80s2JiKgqWLgQUY3i4eGBuLiRsm3Dho3iZSIilWPhQkQ1zqOPdoePTy2rbbVq1UK3bo8rlIiIKouFCxHVODk52SgqKrTaVlhYqKrFKolIHgsXIqpRhBBYuzbBZrFHnU6HtWsTbBaxJCJ1YeFCRDVKVtZ5HD2aZrP0gNlsxtGjacjKOq9QMiKqDBYuRBq0ePFHGDToOSxe/JHSUcq0efNGxMa+gM2bNyodxUpgYFOEhLSV7XEJDQ1DYGBThZIRUWWwcCHSGC2ssZOfn4/k5K0wm81ITt6K/Px8pSNZSJKEwYPjZNsGD46zLFVAROrEwoVIY7Swxs7ChfMsY0WEEFi06AOFE1kLCGiCFi2sV6V/6KGWqlqskojksXAh0pBdu9S/xk56+hFkZBy32nbixG9ITz+iUCJbOTnZOHXqpNW2U6cyeFcRkQawcCHSCJPJhISEZbJtCQnLypzG3p3MZjPi4+fLtsXHz7cZEKuEkruKACG7nXcVEakbCxcijdDCGjuHD6eioKBAtq2goACHD6e6OZEt3lVEpG0sXIg0Qgtr7ISFRcDX11e2zde3DsLCItycyBbvKiLSNhYuRBqhhTV2dDod3nhjgmzbmDETbIoFJZTcVXTn3UNlbScidVH+LEJElfboo91Rr159q2316zdQ1Ro7wcGhCApqZbWtZcvWaNMmRKFEtgICmiAysp+lSJEkCVFRMbyriEgDWLgQacw77/zd6vH06bMUSlK28eMnlyoKdBg37k2FE9mKju6PunX9AQD+/vUQFRWjcCIiqgwWLkQa06BBQ3Tq1AUA0KlTFzRo0FDhRLb8/PwQHd0fOp0O0dEx8PPzUzqSDb1ejyFDhqNBg4Z49dVh0Ov1SkciokqQRDW8989kMuPq1RtKxyAiIqJKqlevNjw8Ku5PYY8LkQalph7A+PEjkZp6QOkoRERuxcKFSGMMBgMSE1fi8uVLSExcCYPBoHQkIiK3cahwOX36NNatW4cPP/wQFy5cwIEDB8qcfIqInCM5eSvy8nIBAHl5uUhJ2aZwIiIi9/G055vMZjPeeecdbNmyBUIISJKEp59+GkuXLsUff/yB9evXIyCAtxUSOVtOTjZ27EiyWsAwJWUbunV7HAEBTRROR0Tkenb1uCxduhQpKSn4+9//jl9++cVyEn3zzTdhNpuxYMECp4YkorLX0uEaO0RUk9hVuGzZsgVjx47Fs88+i7p161q2t27dGmPHjsUvv/zirHxE9D9cY4eIyM7C5fLly2jdurVsW+PGjZGfn+9QKCKyxTV2iIjsLFzuvfde/Pvf/5Zt27dvH+69916HQhEpZfLkcRg06DlMnjxO6Sg2tLbGDm/ZJiJXsKtwGTx4MNatW4eZM2fiv//9LyRJwtmzZ7F69WqsXr0aL7/8srNzErncmTOZlsstWVnnceZMpsKJbAUENEH37k9Ybeve/QnVrbHDW7aJyFXsnjl3xYoVWLZsGQwGg2VQoJeXF+Li4jBunLKfVjlzLtlj8OCXYDIZLY89PT2xZs0mBRPJi4sbhKKiIstjH59aSEj4RMFEtjZv3ojt27dYHvfr9xyee+4lBRMRkdpVduZcu26HBoARI0Zg4MCBOHToEPLy8uDn54e2bdtaDdYl0oqNGz+xKloAwGg0YuPGTzBgwCsKpbK1Y0eSVdECAEVFhdixIwmRkf2UCXWHnJxspKQkWW3jLdtlE0JU2CNV8uGwvMuBer1edZcLiVzB7h6XgwcPYs+ePRg9ejQA4NixY1ixYgWGDRuG4OBgp4asqprY45KaegDr1q1CbOxQRES0VzqOrMWLP8LevbvRqVMXjBkzUek4Frdu3cKQIQPKbE9M3AgvLy83JpJnNBrx6qtl91qsWbMJnp52fxZxCiEE5s37O9LTj1jdni1JEoKDQzF58jT+cS1FCIGZM6fh5MkTDh8rKKgVpk+fxdeXNMulaxX9+9//xuDBg/Gf//zHsk2SJJw5cwYvv/wyDhzgYDx30sJ4gsuXL2Hv3t0AgL17d+Py5UsKJ/rLJ5+sdqjdXbZt2+xQuzuU3LItN9eMErdsCyFQVFRU5ldhYSEKCwvL3aeoqMilc+Sw0HAeLbzf5Di7elyee+45PPjgg5g7d67NL92UKVPw559/YsOGDU4LWVU1rcdl8+aNSE7eapnF+JlnnlXdeIKxY0fg6tUrlsf16zfAokXLFUz0F/a4OI/ZbMaoUa/JLv3h61sHS5eusrmd21W00ptR0aUig6EIo0fHAQCWLEmAXu8ju19Nv1SklfebyubSHpfTp0+jX79+sm9qv379cPz4cXsOS3Yoawr4nJxshZP9ZdeunVZFCwBcuXIZu3btdH8YGV5eXujb9xnZtsjIGFUULcDtwcIvvTRItm3AgFjFixYAyM7OKnO9soKC68jOznJrHi384ZEkCT4+PmV+lS5U9Pqy99PC/9XV+BrUDHad6erUqYPMzEx06dLFpu3PP//EXXfd5XAwqlhFU8CrYTyByWRCQsIy2baEhGV45JFH4eHh4eZUtgYMeAVff/2FzV1FL700UMFUtiIj+yEpaQuKigot22rVqoW+faMVTPWXkknyjh5Ns2lz9yR5kiRh+vRZZfZmVLYnA2Bvhhbw/a457OpxefLJJ7Fo0SL89NNPVtt37dqFRYsW4cknn3RKOCqfFqaA//HH72AymWTbTCYTfvzxOzcnKtvkyW9bPX7zzbfL2FNZM2fOtXr83ntzy9jT/Uomw5Ob3VeJSfLK682obE8GezO0g+93zWBX4fK3v/0N9957L0aOHImQkBA8+uijCAkJwfDhw3HPPfdg4kT13DFSnWlhCviePZ8ss0fFw8MDPXuqp8jdvHmj1ePPP1ffHC7A7fFBJa+ph4cH6tdvoHAiawEBTRAVFWO1LTq6v+omySMibbKrcPH19cWmTZuwbNkyDB48GN27d8crr7yCJUuWYMOGDfD19XV2TpKhhSngPTw8EBc3UrZt2LBRqrhMBADp6Udw6lSG1baTJ08gPf2IQonKlpy81dLLZjabkZKyTZEc5d3B0bt3H9St6w8A8PevhyeffJp3bxCRU9g9mk+n06FHjx7o0aOHM/NQFQUENEFkZD+ru4qiomJU9en20Ue7Y/PmjTZ3FXXr9riCqf5iNpvx8ccfyrZ9/PGHWL58jdvuhKlIWYOx3T25W1Xu4MjNvYrRo4fKtvHuDSKqqkoXLlOnTsWoUaNwzz33YOrUqeXuK0kS/vGPfzgcjionOro/fv75J+TmXoW/fz2bbno1eOedv2P8+L96XqZPn6VgGmuHDh3EzZs3Zdtu3ryJQ4cOol27Dm5OZUttg7FZbBCREipduOzduxeDBw+2/Ls8PKG5l16vx5Ahwy0z5+r1eqUj2WjQoCE6depimTm3QYOGSkeyaNSosUPt7lIyGPtOpQdjN23azC1ZnHUHB+/eIKKqqnTh8uOPP1r+nZSUBD8/P5cEIvtERLRX7VT/JcaMmYgxY5ROYatZs3tw//0PIjPztE3bAw88iGbN7lEgla2Swdi//nrU6k4ynU6H4OBQtw/GLrmDoyIld3AQETmDXRfu+/btiy+//NLZWYgUIUkSRo8eL9s2evTfVNMjoIXB2ERErmZX4VJcXAx/f39nZyFSTEBAEzz9dKTVtqefjlLVIGfgr8HYJUWKGgdjExG5kl2FS2xsLBYuXIhDhw6hsLCw4m8g0oDnnhtguaTh41NLdes9lYiO7m91q7EaB2MTEbmKXbdDb9++HVlZWXj55Zdl2yVJwrFjxxwKRuRuer0ebduGY+/e3WjbNkyVg5yB2zlr1aqF3Fz8b0ZQdeYkInIFuwqX6Gh1rItC5Ez5+fnYt28PAGDfvj3Iz89X5SD0rKzzluUcSv6thlmSiYjcwa7C5Y033nB2DiLFLVw4z2pit0WLPlDVfDMl3nnnLavHM2a8hX/+8xOF0hARuZfdM+caDAYkJSVh3759uHbtGurXr48uXbogMjISnp52H5ZIEenpR5CRcdxq24kTvyE9/QiCg0MVSmVrx44kq5WhAaCwsBA7diQhMrKfMqGIiNzIrsG5WVlZ6Nu3L2bMmIEjR46goKAABw8exFtvvYXnnnsO165dc3ZOIpcxm82Ij58v2xYfP99m9W2lGI1GbNq0XrZt06b1MBqNbk5EROR+dhUus2fPhtlsxrZt2/Ddd99h06ZN+P777/H555/j2rVrmDdvnrNzErnM4cOpKCgokG0rKCjA4cOpbk4kb9u2zQ61ExFVB3YVLnv37sWkSZPQunVrq+3BwcEYP348fvjhB6eEo+olNfUAxo8fidTUA0pHsRIWFlHmiua+vnUQFhbh5kTyYmKed6idiKg6sKtw0ev18PDwkG3z9fXlUvVkw2AwIDFxJS5fvoTExJVlrnGjBJ1OV+b4kMjIGNWsDO3p6Ym+fZ+RbYuMjOHYMnIpIQSKiorK/CosLERhYWG5+/BvAzmDXWe62NhYzJ8/H8HBwWja9K/bMK9du4bly5cjNjbWaQGpekhO3oq8vFwAQF5eLlJStqlmgjez2YwdO5Jk23bs2Io+fSJVUbwIIfDHH2dk286e/R1CCE77Ty4hhMDMmdNw8uQJh44TFNQK06fP4s8pOcSuwuXMmTPIzc3FU089hXbt2qFx48bIzc3FwYMHUVhYCB8fH8sK0pIkYe3atU4NTdqSk5ONHTuSrG41TknZhm7dHkdAQBOF01VujIsaFrAsa3VoAG5fHZpqHhYbpBZ2FS7nzp1Dy5YtAQAmkwlZWVkAgIcfftiyT+k/UlRzCSGwdm2Czc9ByfbJk6cpfkJs2zYcHh4eMJlMNm0eHh5o2zZcgVS2GjRo6FA7kb0kScL06bPKvMRrMBRh9Og4AMCSJQnQ6+VXA9fr9Yr/vpP22VW4fPJJ5Se72r9/P27evIm77rrLnqcijSurl8BsNqumlyAnJ1u2aAFuF+Y5OdmKZwSAlSvjK2wfM2aim9JQTSNJkmUtr/Lo9T6V2o/IXi69cG8ymRAbG4vMzExXPg2pWGBgU4SEtLUZI6LT6RAaGqaKqepLMspRS0bg9gBcR9qJiKoDl4845KWimk2SJAweHGfTPVzWdiVIkoRHHnlUtq1Ll26qyAgA9913P+rUkV87yc/vbtx33/1uTkRE5H7K3ypB1V5AQBNERvazFACSJCEqKgaNGwconOw2s9mMTz9dI9v26adrVDNzLgA0aRIou10Ng5yJiNyBhQu5RXR0f9St6w8A8Pevh6go9VzW0MrMuVlZ523WUyqRkXHcsmI0EVF1xsKF3EKv12PIkOFo0KAhXn11GPR6vdKRLEJDwxxqd5eKeqjU0oNFRORKLFzIbSIi2mPhwmWqmBOltLS0Qw61u8tPP33vUDsRUXXAwoVqvEaNGjvU7i49ez7pUDsRUXWgqsIlMzMT4eHh2Lp1q9JRLNS6MKAWqfW1bNq0GXx8asm2+fjUUsUcLkDFd+jxDj4iqglUU7jcunULkyZNws2bN5WOYqHmhQG1Rs2vZXZ2FoqKCmXbiooKkZ2d5eZE8rZt2+xQOxFRdWBX4bJixQpcuHChwv08PDwwZ84cNGtW8SfWxYsXw9fX1544LiO3MKBaqbU3o4SaX0utTEAXE/O8Q+1ERNWBXYXLP//5T/Ts2RNxcXH48ssvUVxcXOa+MTExuPvuu8s93v79+/HZZ59h7ty59sRxibIWBszJyVY4mS0192YA6n8tJUlC377PyLb17fuM2yegE0KgqKjI5stoNOLZZ+VX1H7++QEwGo1W+/PSERFVR3atVfSf//wH3377LbZv345JkyahTp066NOnD2JiYhAaGlqlY+Xn52Py5MmYNm0amjRx3iRanp72XwUTQmDdulWyCwOuW7cKU6dOV81sqgCwZcs2q96ML75IwgsvDFA41W1aeC2FEPjyy2RIkmSVU5IkfPHFdoSGhrotoxACM2a8jYyME1X6vs2bN2Lz5o1W21q2bIV3352tyOtb+vfP01Pn0O+jq2ghI6CNnO7IKIRw6EOZyVRs9W+j0bGMXDBSOXYVLj4+PoiOjkZ0dDRycnKwfft2fP3119i0aRNatGiB/v3745lnnkG9evUqPNa7776L8PBwREVF2RNFlk4nwd+/tt3f/8cff+DIkcM2281mM44cOYyCgqto3ry5Awmd5/z580hO3mbVm5GcvA2RkU+jaVPlL3Fo4bUsK6MQwu0ZhRDw8rLr19KGp6cH/P1rK3JyLSrysPzb37+2Khfd00JGQBs53ZGxqKgIAwY865RjjRjxmsPH2L59uyrfi5rA4TNkQEAAhgwZggceeADr1q3D/v37MW/ePCxYsAD9+/fHpEmTyhy7kpSUhAMHDiAlJcXRGFbMZoH8fPsH+fr61kNoaBjS049YTfeu0+kQEtIWvr71kJt7wxlRHSKEwMKFH9v0ZpjNZixc+LEqejO08Fr6+tZDq1atcfz4bzZtrVo97PaM06bNLPeTpcFQZDnxPvjgQ5g+/T3Z/fR6PfLylBnsXlRUZPl3bu4N+PjIr76tJC1kBLSR0x0ZSz+HGqj1vdAyP79a8PCouCfMocJl37592L59O7755hvcvHkTnTt3xvz58/HYY4/h559/xsyZM5GVlYWVK1fKfv+WLVtw5coVdO/e3Wr7jBkz8OWXXyIhIcHubEajY+vLxMYOxZQp4622SZKE2NihMJkEAOXHD5w/f67c3ow//vhTFbfyqv21FELAbJbPYDabYTSa3V4Aenp6l9lW+md76tQZZe6r5GtbOqPRaHb499EVtJAR0EZOd2QsfczhA5bCy7Pqs2+XfMiz9/f5ltGAlRtHWfKo8b2oCewqXBYsWICUlBRkZ2ejSZMmePXVV9G/f38EBv61AFyfPn1w4sQJrFu3rszjfPjhhzZVdO/evTF27FhER0fbE81pShYGTE7eCiGE6hYGBP66G+bXX4/a9GYEB4eq5m6Yktdy+/Ytlm1qei0rswaQGgpAIrrNy1MPLy9epqmp7BqdlJiYiLCwMCQkJOCHH37AG2+8YVW0lAgJCcH48ePLPE7jxo1x7733Wn0BQP369dG4sfKzlap5YUDg9qeGwYPjbD49lLVdSR06dLZ63L59J4WS2NLK7dBERGRn4bJr1y7Mnz8fXbt2LfePY69evTB48GC7wylNzQsDlijpzSh5H9TYMwQAf//7O1aPZ89+p4w93U9tt0MTEVHZ7LpUlJGRUeE+HTp0sOfQOHGiareBulpERHvVLQp4p+jo/vj555+Qm3tVlT1DO3Yk2cxMW1hYiB07khAZ2U+ZUKUIITB/vvwcQh99NAerVn3K4oWISCXsKlxeeeWVCk/kv/1me4cGuUZJz9C6dasQGztUVT1DRqMRmzatl23btGk9nnoqEp6ezrn9116nTp0scxLF4uJinDp1Eg89FOTmVEREJMeuvxhyA25v3ryJAwcOYPv27Vi8eLHDwahq1NozVJn1dZ5/XtnJ8pYuXVBh+4IFy9yUhoiIymNX4dKxY0fZ7d27d8ddd92FZcuWYcWKFQ4Fo+ohJuZ5q7uJ5NqV9ve/f4gRI8oei/X3v3/oxjRERFQep8/L3L59e+zbt8/ZhyWN8vT0xEsvDZJtGzAgVvHLRABQu3ZtNGjQULatQYNGqF3b/lmYiYjIuZxeuPz444880ZOV0nc9lbh9J4+yc/WUMJvNNoOHSxQV3bSaI4eIiJRl18fd2NhYm21msxk5OTk4f/48hg0b5nAwqprU1AOWwblqG+uSnn5EdpHF9PQjCA6u2qKcjpJbqC0tLRUFBQWy+xcUFGD//j1o2zbCajsXWCMiUoZdhcudf4SA27O1BgUFYcSIEXj2WecshEWVYzAYkJi4Erm5V5GYuBJt2oSo5s4is9mM+Pj5sm3x8fOxdOlq6HTuWe1WCIGZM6fh5Mmq3XK/eLFt/qCgVpg+fRaLFyIiN7OrcPnkk09gNpvx888/48CBA8jPz0e9evXQqVMndOnSxdkZqQLJyVuRm3sVAJCbexUpKdvw3HMvKZzqtsOHy+/NOHw41a09RCw0iMhVhBAoKipCcXHZi6QKIcqcfqEqvL29yz2feXvr4ePjUy3PeXYVLrm5uRg+fDjS09Ph4eGBunXrIi8vDytWrEDXrl0RHx/P5b7dJCcnGykp26y2JSdvRbdujyMgoIlCqf5S0aUgd14qkiQJ06fPKnPl5Tlz3sPp0yctjx96qCWmTJkuuy8vFRHRnYqKijBs2CtKx7D45z8/Qa1atZSO4XR29dG///77+PPPPxEfH4+jR4/iP//5D44cOYKPPvoIaWlp+PBD3j7qDkIIrF2bYDN41Gw2Y+3aBNlLeu5W3q3QlWl3NkmS4OPjI/v1xht/s9rvb3+bUua+LFqIiJRhV4/LDz/8gClTpuCJJ56wbNPpdOjTpw/y8vLw8ccfY9q0aU4LSfKyss7j6NE02bajR9NUsaqxFuZxKVGnjp/l3336RMPPz6+cvYmIrJX+QDP78Q/h7WE71lAIgVtmxy8VeenkLxUVmwx4+9+TbPJUJ3YVLpIkoX79+rJt999/v1Ou31HFmjQJhK+vr+wYEl/fOmjSxHbFbncrmcdFbtp/tczjIkdNBRURaU8dbz/oPd1/k4TBWPb4murCrktF0dHRSEhIsBkrYDab8cknnyAyMtIp4ah82dlZ5Qx8vY7s7Cw3J5IXGdnPZsyTj08t1czjQkRE2lHpj7tTp061/NtoNOLw4cN44okn0L17dzRo0ADXrl3DL7/8gkuXLmHAAGXXnqkpAgObIiSkrc08KZIkISSkLQIDmyqYzlrXro/hhx++tXpMRERUVZUuXPbu3Wv1uHHjxgCA//73v1bb/f398c0332Dy5MlOiEflkSQJgwfHYfLk8RDCZNmu0+kweHCcaq5v5uRkY+fOH6y27dz5PZ5+OlIVdz4REZF2VLpw+fHHH12Zg+wUENAEUVH9rAbARkXFoHHjAAVT/aXkzie5mXPXrk3A5MnTVFNgERGR+rlnylJyqejo/vD3rwcAqFevPqKiYhRO9JeSO5/kbtkuufOJiIiosli4VAN6vR5DhgxHgwYN8eqrw1Qz3T/w1zgcOaGhYaoah0NEROqnzntRqcoiItqrbnFF4PY4nDZtQmTnm3n44RBeJiK7yS2YWRUGQ5Hsv+0lN5uyoxkB5+bkjM9UHbBwqSbUujq0yWTC5s0bZds2b96Ap5+OhIeHh5tTUXVgMBgQFzfIKccaPTrO4WMkJKy3ue3fmRkBx3PKZSTSGl4qqgZKVoe+fPkSEhNXOvwJz5l+/PE7mEwm2TaTyYQff/zOzYmIiEjL2ONSDSQnb0VeXi4AIC8vV1WrQ/fo0Qtr1yaU207kKM9BzwN2zMJccreb3ZdPjEYY12+u1K4+g0YDnl52PY1DOY23ULR+iV3PS6RGLFw0LicnGzt2JFlObEIIpKRsU83q0Bcu5FTYrvR6SlQNeHpC8qp6UeDoaI8qLWPq6QXJy9uu53Ekp/JLrRI5Fy8VaVhFc6SoYXXoRo0aO9RORERUGgsXDdPCHClJSZ871E5ERFQaCxcNK5kjRaezfht1Op1q5kipaJVlrsJMRERVwcJFw0rWKrpzwF5Z25Xg6emJl16Svx10wIBYeNoxoJKIiGouFi4aFxDQBJGR/SxFiiRJqlqrCAAiI/vBw8O6QPH09ETfvtEKJSIiIq1i4VINREf3x1131QYA1K7tq6q1ioDbdz6ZTEarbUajETk52QolIiIirWLhUk38dVVI+TuJSiu5w0mOWu58IiIi7eAAg2ogOXkrbty4AQC4ceOGqiagK7nzSU7JnU+cx4VIOVpcT+nWLWVmB6/K8xpM9md0ZMJBR55XK1i4aJzaJ6CrVesuh9qJyLW0sp5S6d7ZlZtGOfQczlBRb/GbP45xU5Kah5eKNEwLE9BNmFD+CaaidiIiotLY46JhZV2GKT0BndKXYYYOfR0rVsSX205E6vDg4PnQeert+l5HLm+YjQacXjuh3H1KH3f4S0vh5WVfTkfcumWw9PbI/T/1ej0SEtY79BwGQ5Gl12rJkgTo9fav5q3Xu/81cgcWLhoWGNgUQUGtkJFx3KatZcvWqpiArnPnruUWLp07d3VjGiIqj85TD50CBUFVeXnp4eVl/x90V5EkSfYyl730eh+nHq+64KWiakoNl4kAYOfOHxxqJyIiKo2Fi4ZlZZ2X7W0BgIyM46pYq6hHj14OtRMREZXGS0UaFhjYFHXq+OH69XybNj+/u1VxqejChZwK25Ueh0PWtHh7LBHVHCxcNKy4uFi2aAGA/PxrKC4uVnxwVoMGDR1qJ/fTyu2xRFQz8VKRhi1aNM+hdndYubLsgbmVaSciIiqNPS4aNm7cZAwdOrDcdqWNGDEGe/fuLred1Mtj8GOAp4dd3+vI7bEwmmBa+7Ndz0tE1RsLFw3T6/UIDQ3DkSOHbdrCwiIUv0wEAN7e3vD3r4fc3Ks2bf7+9eDt7a1AKqo0Tw9IXvYVLo6MSlHHPXFEpEa8VKRhQogyb3s2mUyquCW6qKhItmgBgNzcqygqcmzgJhER1SwsXDSsMgsYKm3u3JkOtRMREZXGwkXDmjQJhK+vr2ybr28dNGkS6OZEtnr37uNQOxERUWksXDQsOzsLBQUFsm0FBdeRnZ3l5kS2Khpno4ZxOEREpB0sXDQsMLApQkLa2ty1IUkSQkPDVDEBXUhIW4faiYiISmPhomGSJGHwYPnJvQYPjlPFbKPbt29xqJ2IiKg0Fi7VwJ13D5V3t5G79ev3nEPtREREpXEeFw0TQmDJkoWybUuWLMTMmXMV73W5ePFChe1cq4gcJW4Za9Tzuor5lmNrVGnteUmbWLho2LlzfyIz87RsW2bmaZw79yfuuae5m1NZq1PHz6F2orKU7lU0fbpZwSS3qaWXs6pK5z69boKCSW7T6utI7sNLRRpWmd4Mpb311niH2omIiEpjj4uGhYe3g4+Pj+zssz4+tRAe3k6BVNbmzl2IUaNeK7edyB6lL4N6DHwekpf7T2filtHS21PRZVlx65Y7IlX5eUvnfjB2PnRe7p+iwHzLYOntUfryNqkfCxcNkyQJAQGBOHPmd5u2Jk0CVXECuH49v8J2Pz9eLiLHSF6ekLy8lI5ho/RlD8OnSxRMcltFl2F0XnpFCpeaRAgBg6HsMT0GQ5Hsv++k1+tVcY5XAgsXDcvKOi9btAC3x7hkZZ1XfOBro0aNHWqvbio6aVVGZU9slVGTT35E7iaEwMyZ03Dy5IlK7T96tPx0FwAQFNQK06fPqpG/vyxcNKxkAjq59YrUMgFdUtLnFbY///wAN6VRnsFgQFzcIKcdr7wTW2UkJKyHj4+Pk9LQnUr/UdEPHK1Ir5C4dcvS21MT/8ipDd8Dx7Fw0TBJktCmTYhs4fLww8Gq+AXp2LFLuZPMdezYxY1piJQjeXlB8vJWOgYpSJIkTJ8+q8Je15JLeuWdw2tybykLlwps3rwRKSnbEBUVo7qeAZPJhH/9a4Ns27/+tQFPPx0FDw8PN6ey1rBhI4faq7PGgwHJzt/AkqEK9py3hBG4sNa+5yUix0iSxF5OB/F26HLk5+cjOXkrzGYzkpO3Ij+//IGm7vbDD9/CbDbLtpnNZvzww7duTmRr9ux3HGqvziRPQOcl2fXl4X37y57vtbdYIiJSAxYu5Vi4cJ6ly04IgUWLPlA4kbVWrR52qN0dBgyIdaidiIioNBYuZUhPP4KMjONW206c+A3p6UcUSmTrnnuaw9tb/tZFb2+94rPmAsC6dQkOtRMREZXGwkWG2WxGfPx82bb4+PllXp5xN4PBgOJi+UFexcUGh2+7rSohBIqKiqy+hg8fW+73DB8+1mp/TvdNRETl4dVuGYcPp6KgoEC2raCgAIcPpyIior2bU9l6//1ZFbbPmDHbLVmqOj9BiXfffcvqcU2em4CIKueW0b4PZZW5W8cVz0vOxcJFRlhYBHx9fWWLF1/fOggLi1Agla3evfuUWyj07t3HjWk4PwERucfKjaOUjkAKYuEiQ6fT4Y03JmDu3Jk2bWPGTIBOp44rbJ06dcGyZYtkL13pdDp06uS+OVLKm59g1Kg4FBfbzvCq1/tgyZKEO7bZPzeBo7PSOnNGWqBmz7PgdkYj7LnI6OgncBiN9n0fEdlN8cIlLy8P8+fPx86dO1FQUICWLVti4sSJaN9e2UsxwcGhCApqZTVAt2XL1mjTJkTBVNZ0Oh2GDRuFFSvibdqGDRvt9gKrrPkJVq9ej0GDnrPZvmrVeqc+vzNnpXV0RlqAs9K6k3H9ZqUjkIvp9XokJNh/zjAYiiy/10uWJECvd+x3U6/nmk5KUbzrYMKECTh06BDmz5+PLVu2oHXr1hg6dCh+/11+DR53GjVqnNXjkSPLH2jqbkII/PTT97JtP/30naoGuvbu/bTV46efjlYoCRFpUckHI3u/Shcqer39xyn5Ym+qchTtcTl79ix++eUXbNiwAe3atQMATJ8+Hbt27UJKSgrGjRtXwRFc6/vvv7F6/MMP3+LFFwcqlMbW+fPnbG7ZLpGRcRznz59Ds2b3uDmVvBdeGIhvv/3K8njgQNfO39LnOcDTjp9uR2akBW5fOfiy/OWZyEn4CZyoZlK0cPH398fKlSsREvLX5RdJkiBJkuKz1ObkZOOLL7ZbbduxIwmPP94TAQFNFEpVPTjyx6ayPD3tK1xIO5w5dXrJJ3CXMt6yaxwO4OBYHOMtO5+VSJ0UPbX7+fnh8ccft9r2zTff4OzZs/i///s/h47t6Wn/VTAhBBISltpcainZPmPG31XRTdisWfmrPzdr1tSh18GZSufw9NS5JJda/q8l5P6fzFh17vh5ccdzFK1f4vTjVxXfb9c+B7mHqj6TpqamYurUqejduze6d+9u93F0Ogn+/rXt/v6zZ8/i+PHfZNuOH/8N169fwb333mv38Z1l69at5bbv2vUD+vfv76Y05Ssq+muxR3//2i75dFv6OdRA7v/JjFXnjp8X/kzeZnZgnhJHeoVKP6+W329yD9UULt9//z0mTZqEiIgIfPjhhw4dy2wWyM+/aff35+cXVtiem3vD7uM7y8GDqRW29+jx/9yUpnxFRX/dXpybewM+PiaXPocayP0/mbHq3PHz4qrnEEJgzRr5Fdwry2AowogRrwEAVqxY7dBYnJs3jSgstD53lX4dTq+dYPexnUXL7zc5xs+vFjw8Ku4JU0Xhsn79esyePRtPPfUU3n//fXh7ezt8TKPR/mn5GzcOtLkVukTLlq3RuHGgQ8d3loKC8ouzgoKbqsgJWL8fRqPZJbnU8n8tIff/1FpGcUuZk3vp53XHz4urngMAPD0dO5+VzuXh4e3Q8UwmAdwx0kYLP5POOq6rn4PcQ/HCZcOGDZg1axZeeeUVvP3226oYOyJJEoYPH41Jk8bYtA0fPloVGQGgV6//h+PHfy23naiqSo/tMq/7WcEkt6nptv7qyNG7swDn3qHFu7OoIooWLpmZmfjHP/6BJ598EiNGjMDly5ctbT4+PqhTp46C6W4XMKVPmnc+VlrHjp2h0+nKmDnXAx07dlYgFRFpiTPvzgLcdIcW1WiKFi7ffPMNbt26he+++w7fffedVVtMTAzmzp2rSC4hBNauTZAtXNauTcDkydNU0esiSRKaN78PZ87YTtZ37733qSIjaU/pnxtd7GOQvNw/wFTcMll6e/hzTESlKVq4vP7663j99deVjCArK+s8jh5Ns9luNptx9GgasrLOo2nTZgoks5aVdV62aAGAzMzTqslJ2iV5eShSuBARlYU3sssIDGyKkJC2Nmv96HQ6hIaGITCw/PlT3KVOHT+H2omIiLSGhYsMSZIweLD8InuDB8epput6ypTyl0SoqJ2IiEhrWLiUISCgCR54oIXVtgcffAiNGwcolMjWwIFDHGonIiLSGhYuZcjJycbp0yettp06lYGcnGyFEtmq6K4h3lWkbuZbgPmWUOBL6f85EZH9FJ/HRY2EEFi5consWkUrVy7B9OmzVHG5KCmp/GWIk5I+xwsvvOymNFQZpX+mLq5TMMj/qOn2fiKiymCPi4zz58/JzpoLABkZx3H+/Dk3J5IXGhruUDsREZHWsMdFw5YvX1Rh+8KFy92UhiqjdE9do1hA5+X+DOZbf/X2qKHnkIioKtjjIqNJk0CH2t3lkUced6idlKXzAnRekgJfSv/PiYjsx8JFxqFDBx1qd5fdu//jUDsREZHWsHCRUVH3uVq6169dy3OonYiISGs4xkVGeHg7eHt7o7i42KbN21uP8PB2CqSy1bp1MNLSyu79ad062I1p1MVorFnPS0RUU7BwKcOd0/3/tV0dvS0AEBkZXW7hEhkZ7cY0yit9a++X5d8p7ha81ZiIyPl4qUjG4cOpKCoqkm0rKirC4cOpbk4kb9Wq8u8YqqidqEJGE8Qt+77MxUaYi412fS+MJqX/50SkUuxxkREWFgFfX18UFBTYtPn61kFYWIQCqWx16fIotm37V7ntNUnpsUd9ngM8FfjpNhr/6u1Ry1goR5jW/qx0BCIiK+xxkaHT6fDGGxNk28aMmVDmZSR327On/LuGKmqvzjw9lfsiIiLX4Wm2DMHBoQgKamU1g27Llq3Rpk2IgqmsvfnmNEyYMKrcdqKq0uv1SEhY79AxDIYijB59e4X1JUsSoNf7OJSHiKgEC5dyjBo1DuPHj7Q8HjlyrIJpbL333v9V2L5kSYKb0lB1IUkSfHzsLzTupNf7OPV4RFSzqeOah0r99NP3Vo937vxBoSTy2rXr6FA7ERGR1rDHpQw5OdnYsSPJaltKyjZ06/Y4AgKaKBPqDidO/OZQe2UJIWAwGBw6hsFQJPtve+j1+mox8JWInKu8c1VVzkHV/RwjhEB+/rVy9zGbBQoKrjv0PL6+dcqdQsTP7267XmcWLjKEEFi7NsFmHo6S7ZMnT1PFD/Vjj/XExo1ry213BoPBgLi4QU45FgDL2Ad7JSSs56UHIrIihMDMmdNw8uSJCvet6BwUFNQK06fPUsV53tmq8jq5mr2vMy8VycjKOo+jR9NgNputtpvNZhw9moasrPMKJbNWXtFSmXZSljAC5lvCri9T8e0ve75XcHZfqqaqY6HhClp/ndjjIiMwsClCQtri6NE0m7bQ0DAEBjZVIJWtsLD2OHz4QLntzvZ6tA5edv7UlPRg2fNLc8sILE82V7yjhlxgXUnkNJIkYfr0WeVe1q7sOag6XyoqeZ14qaiakSQJjzzyqGzh0qXLo6r5gS6vaKlMuz28PAEvT3v//468bpw+n4jK5+w74qorSZJw9911K9zP39/f9WHswMJFhtlsxiefrJZt++STVeja9VFVTEIXHByO9PRD5baTunCOFCIix7BwkXH4cCpu3rwp23bz5k0cPpyKiAjnX4YpS1kj5csrWkraS6+5VJ27P7WCc6QQETmGhYuMhg0bOdTuTI6OAC99N1B1HilPREQ1AwsXGYGBTSFJks3t0MDtT8zuHpzLQoOo+qponqTKzj/CHlWqKVi4yEhLOyRbtAC3TzJpaYfcdqmovJHyRqMRr7/+apnfu3z5GniWWvWvJp3YjHbe8lvyttv7Mtn7vFQzVbVHtbz5R9ijSjUFCxcZoaFhDrU7W3njIl56aRA2bbId7DlgQCx8fX1dHU21vvxc6QRElcNCg6hqWLjIOHLkcIXt7hycW57IyH6yhUvfvtEKpCGiqqjM3CNA5eYfqUk9qlSzsXCRobYel4rMmjUP06dPtjyeN2+RgmmU4+itxs68zbgkD1FFOPcIUdWwcJFx6NDBCts7dOjkpjQVa9Ik0PLvFi2CVDOzr7s58w8AbzMmIlIn5WdRU6Hc3KsOtSvprbfeUToCERGRy7BwkREU1MqhdiIiInINXiqScenSxQrb77vvfjelUZdbRmXWDFLqeYmISF1YuMg4ffpkhe1qGuPiaqXntFmeLKD0godlzbFDRETVHy8Vyahdu/z5TypqJyIiItdgj4uMimaxtHfdIK0qPTfE69ESvDzdP1fELaP4X28PJ+wiIqrJWLjI8PGp5VB7deblqUzhchsvERER1XS8VCTjzJnfHWonIiIi12DhIsNsLv+TfUXtRERE5Bq8VCQjJ+e8Q+3V2S0jYO8lm8qst1L+8xIRUU3HwkVGvXr1cfXqlXLba6rlyWYnHIU9VkREZB8WLjLKK1oq014VQogKV4atiMFQJPtve3CFWapOyvv9qsrvDX8viNSDhYsMT09vGI3F5bY7i8FgQFzcIKcdr2R1Y3slJKy3WVzQ0VWXAeeuvMxVl6kyhBCYOXNapaYvqOj3JiioFaZPn8XihUgFWLjIKK9oqUx7dePMVZcBrrxM7sNCg6j6YeGiIh/1qg29h30nWkcGvhpMAhO/v2HX8xKplSRJmD59VrmXYiv7e8NLRUTqwcJFRfQeEvR2T+7GkyrRnZzdW0hEyqvxhYu9g2OLiqwH8/ETGRFVZ84Y6MzzJDlDjS5cqjJ47053Dqjl4D0iqq6cNdCZ50lyhhpduADqGrxnMCozv4lSz0tE2qGmcyXVbDW6cClr8N7Vq1cxefLYMr9v3ryPUa9ePatt9naBlgwOBICJPyg/QLZ0HiIiwHkDnXmpiJyhRhcugPzgvcDAQPj4+NiMYwEAHx8fBAYGuiseEZEqcKAzqUWNL1zKkpCwHoMGPSe73ZlKf/r46InaDtxVZD+DUVh6e/hpiCqLgzWJSAksXMrRu/fT+PbbryyPn3462qXPp/d05HZoIvfhYE0iUgoLl3K88MJAq8Jl4MBYBdNQTVHRLfpq6c1gsUFESqjWhYujCxiW/qOwZEmC7JiXqqjoD4nBZP/AWEdnziV1qOot+kr1ZnCwJhEppVoXLs5cwNDRxQsB+QUMS+O0+wRopyeDgzWJSAnVunAh0prK9GQA7M0gopqrxhQuS55+DnqPqv93HbkEAwAGkxGjv/q8zHa9Xu/wnUoGQ5GlR2jJkgTo9fZ/Ctbr9Q5lIcexJ4OIqGw1pnCBncM4HP7EWsHzOvuPlF7vwz965XDGLbwAezOIiJRSrQuX0rPAjv667F4Pd+GstMpy1i28AG/jJSJSik7pAETuxEKDiEjbqnWPi9r+SKktT03jrFt4AV4qIiJSSrUuXMi9tDAFPAe+EhFpmyoKF7PZjPj4eGzevBnXr19Hhw4d8M477+Cee+5x6LiO3rHjzLt1SvLYQwszqXIKeCIicgdVFC5Lly7Fhg0bMHfuXAQEBOCDDz5AXFwcUlJS4O3tbfdxK/Pp2tHZdUu4qijQykyqAC+FERGR6yleuBQXF2P16tWYNGkSunfvDgBYsGABHn30UXz77beIjIx02XNr5S4TLRQEnAKeiIjcQfHC5fjx47hx4wa6dOli2ebn54eHH34Y+/fvd2nhAqi/KNDSTKocP0JERK6meOGSk5MDAGjSpInV9kaNGlna7OHpWbk7vd97b7Ym7jLx8rrLZccmIiLSCsULl8LCQgCwGcui1+tx7do1u46p00nw969dhe/wtet5iIiIyL0UL1xKLi0UFxdbXWYwGAyoVauWXcc0mwXy8286JR8RERG5np9fLXh4VHy1RPHCpeQS0cWLF9G8eXPL9osXL6Jly5Z2H9doNDucjYiIiNRF8Sn/W7VqBV9fX+zdu9eyLT8/H8eOHUOHDh0UTEZERERqo3iPi7e3NwYNGoQPP/wQ9erVQ9OmTfHBBx8gICAAvXv3VjoeERERqYjihQsAjB07FkajEdOmTUNRURE6dOiAVatWwcvLS+loREREpCKSKLnftxoxmcy4evWG0jGIiIiokurVq12pwbmKj3EhIiIiqiwWLkRERKQZLFyIiIhIM1i4EBERkWawcCEiIiLNYOFCREREmsHChYiIiDSjWs7jIoSA2Vzt/ltERETVlk4nQZKkCverloULERERVU+8VERERESawcKFiIiINIOFCxEREWkGCxciIiLSDBYuREREpBksXIiIiEgzWLgQERGRZrBwISIiIs1g4UJERESawcKFiIiINIOFCxEREWkGCxciIiLSDBYuREREpBksXEpZsWIFXnnllXL3yc3NxcSJE9GhQwd07NgR7733HgoLC12aKy8vD++88w4ee+wxREREYMCAAThw4ECZ+587dw4jRoxAREQEunXrhoULF8JkMrk0IwBcuXIFb775Jjp37ozw8HAMHz4cp0+fLnN/JV7L0jIzMxEeHo6tW7eqLuOFCxfQsmVLm6+ysiqVMykpCX369EFISAj69u2Lr776qsx9DQYD3nvvPXTp0gXh4eGYOHEirl696rJse/fulX0NW7ZsiSeeeEIVGUsYjUYsWrQIPXr0QHh4OAYOHIjDhw+Xub8S73dBQQFmzJiBbt26oWPHjpg0aRKuXLlS5v7uPg/Jnb9/++03DBo0CGFhYejZsyfWrVtX4XG++uor9OnTB6GhoejXrx92797t8pwAcPbsWYSFheHcuXMVHuPTTz/FE088gdDQULz88ss4duyYWzKWmDZtGnr27FnhcVz2WgoSQgixfv160apVKzFo0KBy9xs0aJB49tlnRXp6uvjvf/8revToISZPnuzSbEOGDBGRkZFi//794vfffxfvvfeeCA0NFadPn7bZt7i4WPTu3VsMHz5cnDhxQnz33XeiY8eOYtGiRS7NKIQQL774onj++edFWlqaOHXqlBgzZozo1q2buHnzpuz+SryWJYqLi0X//v1FUFCQ2LJlS5n7KZVx586dIiQkRFy4cEFcvHjR8lVYWKianElJSeLhhx8W69evF2fPnhVLly4VrVq1EqmpqbL7v/XWW6JXr15i//79Ii0tTfTr108MHDjQZfkMBoPVa3fx4kXx7bffipYtW4rPP/9cFRlLfPzxx6Jr165i165d4syZM+Ltt98W7dq1ExcuXJDdX4n3+7XXXhOPP/642Llzp8jIyBCjRo0Sffr0EQaDwWZfd5+H5M7fV69eFZ06dRJTp04Vp06dEp9//rkICQkp870XQojdu3eLNm3aiLVr14pTp06JuXPniuDgYHHq1CmX5RRCiFOnTomePXuKoKAg8eeff5Z7jK1bt4rQ0FCxfft2cfLkSfHmm2+Kjh07iitXrrg0Y4nvvvtOBAUFiR49epR7HFe+ljW+cMnJyREjRowQYWFh4qmnniq3cElNTRVBQUFWL/yuXbtEy5YtRU5OjkvynTlzRgQFBYkDBw5YtpnNZtGrVy+xcOFCm/1TUlJEcHCwyMvLs2zbtGmTiIiIkD3BOEteXp6YMGGCOHHihGXbb7/9JoKCgkRaWprN/kq8lqV99NFHIjY2ttzCRcmMK1euFFFRUZXaV4mcZrNZ9OjRQ8ydO9dq+2uvvSaWL19us39OTo5o1aqV2Llzp2Xb77//LoKCgsosdJztxo0bokePHuKtt96SbVcyY3R0tJgzZ47l8fXr10VQUJD45ptvbPZV4v0+duyYCAoKEv/+978t2woKCkT79u3F1q1bbfZ313movPP38uXLRbdu3cStW7cs2z766CPRu3fvMo/32muviXHjxllte/HFF8X06dNdmjMsLEzExMRUqnDp3bu3mDdvnuXxrVu3xOOPPy77e+esjCUuXLggOnfuLAYNGlRh4eKq11IIIWr8paJff/0VXl5eSE5ORtu2bcvd98CBA2jYsCEefPBBy7aOHTtCkiQcPHjQJfn8/f2xcuVKhISEWLZJkgRJkpCfny+bsU2bNrj77rst2zp37oyCggL89ttvLskIAHfffTc++ugjBAUFAQCuXr2KNWvWICAgAC1atJDN6e7XssT+/fvx2WefYe7cueXup2TGEydOWD1veZTImZmZifPnzyMqKspq+6pVqzBixAib/UtydO7c2bLt/vvvR+PGjbF//36XZLzT8uXLUVhYiClTpsi2K5mxfv36+Omnn3Du3DmYTCZ89tln8Pb2RqtWrWz2VeL9PnPmDACgffv2lm21a9fGvffei3379slmdMd5qLzz94EDB9CxY0d4enpaZThz5gwuX75scyyz2YzU1FR06dLFanunTp0cfv/Ly/n9999jzpw5Zf5clnblyhWcOXPGKqOnpyfat2/v0owAIITAW2+9hWeeeQYdO3Ys91iufC0BwLPiXaq3nj17VupaHXB73EGTJk2stnl7e6Nu3brIzs52RTz4+fnh8ccft9r2zTff4OzZs/i///s/m/1zcnIQEBBgta1Ro0YAgOzs7AqLM2eYPn06/vWvf8Hb2xvLli3DXXfdZbOPEq8lAOTn52Py5MmYNm2azfOrJSMAZGRkwN/fHwMHDkRmZibuvfdejBw5Eo899pgqcmZmZgIAbt68iaFDh+LYsWNo1qwZRo4cKfv7dOHCBfj7+0Ov11ttb9SoEXJyclySsbSSQnrixImoW7eu7D5KZnz77bcxbtw4PPHEE/Dw8IBOp8PixYvRvHlz2Zzufr9Ln0NKCiaTyYScnBzUr1/fZn93nYfKO3/n5ORYPkjJZWjQoIFVW35+Pm7evCmb29H3v7ycmzdvBnB7TFZFSnLc+f43atQIx48fd1lGAFizZg0uXbqE5cuXY8WKFeUey5WvJcDBuVVSWFgIb29vm+16vR4Gg8EtGVJTUzF16lT07t0b3bt3t2kvKiqyyVhyInZXxsGDB2PLli2IjIzE6NGj8euvv9rso9Rr+e677yI8PNymp0COUhmNRiN+//13XLt2DWPGjMHKlSsRFhaG4cOHyw5uUyJnQUEBAGDKlCmIjIzE6tWr0bVrV4waNUo1GUvbsGED6tSpgxdffLHMfZTMeOrUKdSpUwdLlizBZ599hv79+2PSpEmyvRNK5AwJCcEDDzyAGTNm4MKFCygqKsJHH32E3Nxc3Lp1y2Z/NZyHqpqhqKgIAGS/x12ZK1IyANvdGY8fP474+Hh88MEHsj97d3L1a1nje1yqwsfHB8XFxTbbDQaDbK+Cs33//feYNGkSIiIi8OGHH8ruI5ex5AfFHRkBWC4NzZ49G2lpaVi/fj3mzJlTYU7Ata9lUlISDhw4gJSUlErtr9T77enpib1798LDwwM+Pj4AgODgYJw8eRKrVq2y6X5VIqeXlxcAYOjQoYiJiQEAtG7dGseOHUNiYmKVMtaqVcslGUtLSkpCv379LK+nHKUyZmdnY+LEiVizZo3lUkxISAhOnTqFxYsXY+nSpZXO6ar329vbG/Hx8Zg8eTIee+wxeHl5ISoqCj169IBOZ/v5Vw3noapmKClq5L7HHT+jlVHy8+vOjAaDAZMmTcLIkSNlL13KcfVryR6XKggICMDFixetthUXFyMvL8/SBekq69evx5gxY9CjRw8sX77cpju7vIwljxs3buyyfFevXsUXX3wBo9Fo2abT6dCiRQubPGXldPVruWXLFly5cgXdu3dHeHg4wsPDAQAzZsxAXFycKjKWqF27ts0f2YceeggXLlxQRc6Sn6U7u+JbtGgheztnQEAA8vLybE5kFy9edOnPJXD70+Kff/5ZYS+bUhnT0tJw69Ytq3FsANC2bVucPXtWNqcSP5cPPvggtmzZgr1792LPnj2YM2cOcnJyZC9nKXUeciRD3bp1cdddd8l+j7syV6TkEpE7M6alpeHkyZOIj4+3nDdXrFiBrKwshIeHy07N4erXkoVLFXTo0AE5OTlWJ5OSgWnt2rVz2fNu2LABs2bNwsCBAzF//vxyu+o6dOiAY8eOWbryAWDPnj2oXbt2patle1y+fBkTJkywukxw69YtHDt2THaQqRKv5Ycffogvv/wSSUlJli8AGDt2LGbPnq2KjABw8uRJRERE2FzzTk9Plx3orETONm3aoHbt2khLS7PanpGRIfuHrF27djCbzVaDRzMzM3HhwgV06NDBJRlLHDhwAPXr16/w51+pjCXjAE6cOGG1PSMjA/fdd5/N/kq83wUFBRg0aBCOHz+OunXrwtfXF+fOncOxY8fQtWtX2YxKnIfuzHDw4EGruWP27NmD+++/X3ZcjiRJiIiIsBlsvHfvXqtByUqqX78+7r//fqtzg9FoxIEDB1z2MxoaGopvv/0W27dvt5w3X3rpJTRq1AhJSUkIDg62+R6Xv5YO35dUjUyZMsXqFjCj0Wg1d4bZbBYvvfSSiImJEWlpaWL37t3l3l7pDL///rto06aNGD16tM2cFPn5+Za5KkpuMSwqKhK9evUSQ4cOFb/99ptl/oTFixe7LGOJuLg40bt3b7Fv3z5x4sQJMWHCBNGhQwdx/vx5VbyWckrfDq2WjCaTSTz77LOiT58+Yv/+/eLUqVPiH//4hwgODhYnTpxQTc4lS5aI8PBwkZKSYjWPy549e4QQQly8eFEUFBRY9p8wYYLo2bOn2LNnj2WOlIrmTXKGqVOnildffVW2TQ0ZTSaTGDBggHjqqafE7t27RWZmpliwYIFo3bq1OHz4sGre75dfflkMGjRIZGRkiCNHjojIyEgxZMgQIYRQxXnozvP35cuXRYcOHcSUKVPEyZMnxZYtW0RISIjV7dv5+flW85/s2rVLtG7dWqxevVqcOnVKvP/++yI0NNRp87jI5SyxZ88e2duhc3NzRW5uruXxZ599JkJDQ8XWrVst87h06tTJafO4lJexxMcff2xzO7Q7X0sWLqXc+Wb9+eefNvN8XL58WYwZM0aEhYWJTp06iRkzZoiioiKXZVq2bJkICgqS/ZoyZYrlh73kj4UQt+d+GTJkiAgJCRHdunUTCxcuFCaTyWUZS+Tn54sZM2aIrl27itDQUPHaa6+JjIwMIYQ6Xks5pTOpKeOlS5fEW2+9Jbp27SpCQkLEiy++KPbv36+6nKtXrxY9e/YUbdq0EdHR0eK7776ztAUFBYmPP/7Y8vjGjRvi7bffFu3btxft27cXEyZMEFevXnV5xri4ODF+/HjZNrVkzMvLE++++67o3r27CA8PFy+++KLYu3evEEI973dOTo4YPXq0aNeunejSpYuYMWOGpehTw3lI7o9tWlqaeOGFF0RwcLDo0aOH+OSTT2y+584/wNu2bRNPPvmkCAkJETExMeK///2vy3MKUXbhMmjQIJv9ExISxGOPPSZCQ0PFyy+/LI4dO+aWjCXkChd3vpaSEEI43m9DRERE5Hoc40JERESawcKFiIiINIOFCxEREWkGCxciIiLSDBYuREREpBksXIiIiEgzWLgQERGRZrBwISLFcTopIqosFi5EVGVbt25Fy5YtZRdUrKoffvgBU6ZMsTzeu3cvWrZsabNWExERAHgqHYCIarY1a9YoHYGINIQ9LkRERKQZLFyIapiePXvi448/xvvvv49HHnkEoaGhGDp0KM6cOVPlY6WmpqJfv34IDg5GZGQkvvzyS6v2c+fOYfLkyejWrRvatGmDLl26YPLkycjNzQUAvPLKK9i3bx/27dtnc3no999/x9ChQ9G2bVt07doVH374IYxGo6W9ZcuWiI+PR//+/REaGor4+HgAwJkzZzB27Fh07doVYWFheOWVV3Dw4EGrXNevX8ecOXPQq1cvhISEIDIyEp9//rnN6xQfH49//OMf6NSpE8LDwzFx4kTcuHEDK1euxGOPPYZ27dphzJgxlv8PAKSnp2Pw4MFo164dwsPD8eqrr+Lw4cNVfm2JSB4XWSSqYXr27In8/Hy0a9cOL7/8Mq5du4bZs2fjvvvuw2effVapY2zduhVTp05FrVq1MHLkSLRu3Rrbtm3Dl19+iSVLlqBXr14oLCxE37594e/vj9dffx116tTBoUOHEB8fj2effRYzZ87EqVOn8OabbwIAZsyYgRYtWuDXX39FbGws9Ho9Xn/9dYSHh+OHH37AJ598gunTp2PQoEEAbhcuXl5emDhxIu6//340bdoUkiThhRdewH333Ydhw4bBy8sL69atQ2pqKlavXo2OHTuiqKgIzz//PK5cuYKxY8eiadOm+P7777Fp0yb87W9/w+uvv255nfLy8tC1a1cMGDAA6enp+Oijj3D//fejUaNGGDp0KM6dO4fZs2fjxRdfxIwZM1BQUIBevXqhc+fOeOGFF1BcXIxly5bh1KlT2LlzJ+rUqeOaN5WoBuEYF6IayM/PD0uXLoWHhwcA4I8//sDixYuRm5sLf3//Sh9nzJgxGDp0KADgsccew5kzZ7B06VL06tULZ86cQUBAAN5//33cc889AIDOnTsjLS0N+/btAwC0aNECvr6+AICwsDCrY8fGxmLUqFGW7/v++++xZ88eS+ECAO3bt8eQIUMsj8ePHw9vb2+sW7fOctzu3bsjMjIS8+bNw+eff46tW7ciIyMDmzZtQnh4OADg0UcfhdFoxNKlS/HSSy+hbt26AABfX18sWLAAnp6eeOSRR7Bt2zZcuHABmzdvthQhu3btQmpqKgDg1KlTyM3NRWxsLCIiIgAADzzwAD777DPcuHGDhQuRE/BSEVENFBISYilaACAgIAAAUFhYWKXj9OnTx+pxr169cOzYMdy4cQOtW7fGhg0b0LRpU5w5cwb//ve/sWrVKvz+++8oLi6u8Njt27e3/FuSJDRt2hT5+flW+7Ru3drq8b59+9CjRw9L0QIAnp6e6Nu3L9LT03Hjxg3s27cPTZs2tRQtJaKjo2EwGJCWlmbZFhoaCk/Pvz7fNWjQAPfff79VAVK3bl1cv34dAPDQQw+hXr16eP311/HOO+/gu+++Q4MGDfDmm29aXmMicgx7XIhqoFq1alk91uluf4Yxm81VOk6DBg2sHtevXx9CCBQUFKB27dpITEzE8uXLkZeXhwYNGiA4OBi1atWy/KGvasY7r2zfddddVo+vXbtmk6kkZ0mua9euoWHDhmX+X0oXR6ULoLKes7TatWvj008/xbJly/DVV1/hs88+g4+PD5555hlMmzYN3t7eZX4vEVUOCxcistudhcLly5fh4eGBu+++GykpKZg7dy7efPNN9O/fH/Xq1QMAjBs3DkePHnVJnrvvvhuXL1+22X7p0iUAgL+/P+6++26cPXu23H0c8cADD+CDDz6AyWTCkSNHsH37dmzcuBHNmzdHXFycQ8cmIl4qIiIH7Ny50/Jvs9mMr7/+Gm3btoWPjw8OHjwIPz8/xMXFWYqWGzdu4ODBg1Y9OyW9Pc7QoUMH/PTTTygoKLBsM5lM+OKLLxASEgJvb2906NAB58+fx6FDh6y+Nzk5GV5eXggNDbX7+b/++mt07twZly5dgoeHB8LDw/Huu+/Cz88PWVlZdh+XiP7CHhcistvChQthMpnQpEkTbNy4EZmZmUhMTARwe3zIxo0bMXfuXPTo0QMXL17EqlWrcPnyZdx9992WY/j5+eHQoUPYvXs3Hn74YYfyvPHGG/j5558RGxuL4cOHw8vLC+vXr8eff/6JhIQEAED//v2xYcMGjB49GmPHjkWzZs3w448/YsuWLXjjjTfg5+dn9/NHRETAbDZj9OjRGD58OGrXro2vvvoK169fR+/evR36vxHRbSxciMhuc+bMwdy5c3H27FkEBQXhn//8Jzp27AgAiImJwblz57BlyxZs2LABjRs3xuOPP46XX34Z06dPx+nTp/Hggw9i4MCBSE9Px7BhwzBnzhw0atTI7jwPPfQQNmzYgPnz52Pq1KmQJAmhoaFYt26dZbBvrVq18Mknn+Cjjz7CokWLUFBQgAceeACzZ8/Gc88959Dr0ahRIyQkJGDRokV4++23UVhYiIceegiLFy9G586dHTo2Ed3GeVyIiIhIM9jjQkQWpWemLYtOp3PquBQioqpgjwsRWbRs2bLCfWJiYjB37lw3pCEissXChYgsKnObsr+/P5o1a+aGNEREtli4EBERkWbwQjURERFpBgsXIiIi0gwWLkRERKQZLFyIiIhIM1i4EBERkWawcCEiIiLNYOFCREREmvH/AURzI+9yf4F5AAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#   Relación precio de venta - cantidad de baños\n",
    "sns.boxplot(x='n_bathrooms', y='buy_price', data=df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nSi el dataframe tuviese buena información geográfica se podrían visualizar los plot geográficos\\n'"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "Si el dataframe tuviese buena información geográfica se podrían visualizar los plot geográficos\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Limpieza **final** del dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>sq_mt_built</th>\n",
       "      <th>n_rooms</th>\n",
       "      <th>n_bathrooms</th>\n",
       "      <th>is_floor_under</th>\n",
       "      <th>rent_price</th>\n",
       "      <th>buy_price</th>\n",
       "      <th>has_parking</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>64.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>471.0</td>\n",
       "      <td>85000</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>70.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>666.0</td>\n",
       "      <td>129900</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>94.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>722.0</td>\n",
       "      <td>144247</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>64.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>583.0</td>\n",
       "      <td>109900</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>108.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1094.0</td>\n",
       "      <td>260000</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sq_mt_built  n_rooms  n_bathrooms  is_floor_under  rent_price  buy_price  \\\n",
       "0         64.0        2          1.0             0.0       471.0      85000   \n",
       "1         70.0        3          1.0             0.0       666.0     129900   \n",
       "2         94.0        2          2.0             0.0       722.0     144247   \n",
       "3         64.0        2          1.0             1.0       583.0     109900   \n",
       "4        108.0        2          2.0             0.0      1094.0     260000   \n",
       "\n",
       "   has_parking  \n",
       "0        False  \n",
       "1        False  \n",
       "2        False  \n",
       "3        False  \n",
       "4         True  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "Si una columna no tiene una correlación alta con la variable objetivo, no conviene utilizarla como variable.\n",
    "\"\"\"\n",
    "\n",
    "# Eliminamos las columnas de identificación y las columnas no correlacionadas con el método .drop()\n",
    "df = df.drop(\n",
    "    [\n",
    "    'id',\n",
    "    'title',\n",
    "    'subtitle',\n",
    "    'is_exact_address_hidden',\n",
    "    'is_renewal_needed',\n",
    "    'is_new_development',\n",
    "    'neighborhood_id',\n",
    "    'operation',\n",
    "    'is_rent_price_known',\n",
    "    'is_buy_price_known',\n",
    "    'buy_price_by_area',\n",
    "    'is_buy_price_known',\n",
    "    'house_type_id',\n",
    "    'energy_certificate',\n",
    "],\n",
    "axis=1\n",
    ")\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Limpieza temporal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nSi el dataset tuviese datos temporales es importante visualizar la evolución temporal y, si ha habido un importante cambio, quedarse sólo\\ncon la porción de datos que corresponde a la situación actual para mejorar la futura predicción.\\n'"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "Si el dataset tuviese datos temporales es importante visualizar la evolución temporal y, si ha habido un importante cambio, quedarse sólo\n",
    "con la porción de datos que corresponde a la situación actual para mejorar la futura predicción.\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Variables categóricas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nSi el dataset tuviese variables categóricas se debería utilizar OneHotEncoder para reemplazar cada categoría por un número\\n'"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "Si el dataset tuviese variables categóricas se debería utilizar OneHotEncoder para reemplazar cada categoría por un número\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train/Test split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nDebemos dividir todo el conjunto de datos en grupos de entrenamiento y prueba. Por lo general se dedica el 80% para entrenamiento\\ny el 20% restante para prueba.\\n'"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "Debemos dividir todo el conjunto de datos en grupos de entrenamiento y prueba. Por lo general se dedica el 80% para entrenamiento\n",
    "y el 20% restante para prueba.\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Separamos el dataframe en X e y, donde X representa a todas las variables excepto la variable de salida, la cual es representada por y\n",
    "X = df.drop('buy_price', axis=1)\n",
    "y = df['buy_price']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importamos train_test_split desde sklearn.model_selection\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Utilizamos el método train_test_split() para dividir el dataframe y generar las variables de entrada y salida de los conjuntos de train y test\n",
    "#   train_test_split(variables de entrada, variables de salida, test_size=tamaño del conjunto de test (en porcentaje decimal), random_state=???)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.2, random_state=101)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15448, 6)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Vemos el tamaño de los conjuntos con .shape (retorna una tupla (celdas, columnas))\n",
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3862, 6)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Escalado"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "Para lograr un entrenamiento correcto, sin dar mayor importancia a variables sólo por su orden de magnitud (mayor importancia a una variable\n",
    "que va de 1 a 100 sobre otra variable que va de 0 a 1) se debe realizar un escalado de todas las variables entre 0 y 1\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "#  Importamos MinMaxScaler desde sklearn.preprocessing\n",
    "from sklearn.preprocessing import MinMaxScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creamos una instancia de MinMaxScaler\n",
    "scaler = MinMaxScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>sq_mt_built</th>\n",
       "      <th>n_rooms</th>\n",
       "      <th>n_bathrooms</th>\n",
       "      <th>is_floor_under</th>\n",
       "      <th>rent_price</th>\n",
       "      <th>has_parking</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2160</th>\n",
       "      <td>87.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>913.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13201</th>\n",
       "      <td>150.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1839.0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14224</th>\n",
       "      <td>682.0</td>\n",
       "      <td>24</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6864.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17339</th>\n",
       "      <td>154.0</td>\n",
       "      <td>5</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2510.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5298</th>\n",
       "      <td>70.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>621.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6473</th>\n",
       "      <td>67.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>762.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9240</th>\n",
       "      <td>146.0</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2070.0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19993</th>\n",
       "      <td>104.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1890.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20197</th>\n",
       "      <td>85.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1197.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14945</th>\n",
       "      <td>116.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2285.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>15448 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       sq_mt_built  n_rooms  n_bathrooms  is_floor_under  rent_price  \\\n",
       "2160          87.0        2          1.0             0.0       913.0   \n",
       "13201        150.0        2          2.0             1.0      1839.0   \n",
       "14224        682.0       24          4.0             0.0      6864.0   \n",
       "17339        154.0        5          5.0             0.0      2510.0   \n",
       "5298          70.0        3          1.0             0.0       621.0   \n",
       "...            ...      ...          ...             ...         ...   \n",
       "6473          67.0        2          1.0             0.0       762.0   \n",
       "9240         146.0        3          2.0             0.0      2070.0   \n",
       "19993        104.0        2          1.0             0.0      1890.0   \n",
       "20197         85.0        2          2.0             0.0      1197.0   \n",
       "14945        116.0        2          2.0             0.0      2285.0   \n",
       "\n",
       "       has_parking  \n",
       "2160         False  \n",
       "13201         True  \n",
       "14224        False  \n",
       "17339        False  \n",
       "5298         False  \n",
       "...            ...  \n",
       "6473         False  \n",
       "9240          True  \n",
       "19993        False  \n",
       "20197        False  \n",
       "14945        False  \n",
       "\n",
       "[15448 rows x 6 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Aplicamos el método .fit_transform() sobre X_train (ajusta los datos antes de escalarlos)\n",
    "X_train = scaler.fit_transform(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.07230143, 0.08333333, 0.        , 0.        , 0.07122394,\n",
       "        0.        ],\n",
       "       [0.13645621, 0.08333333, 0.07692308, 1.        , 0.14491485,\n",
       "        1.        ],\n",
       "       [0.67820774, 1.        , 0.23076923, 0.        , 0.54480344,\n",
       "        0.        ],\n",
       "       ...,\n",
       "       [0.08961303, 0.08333333, 0.        , 0.        , 0.14897342,\n",
       "        0.        ],\n",
       "       [0.07026477, 0.08333333, 0.07692308, 0.        , 0.09382461,\n",
       "        0.        ],\n",
       "       [0.10183299, 0.08333333, 0.07692308, 0.        , 0.18040745,\n",
       "        0.        ]])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>sq_mt_built</th>\n",
       "      <th>n_rooms</th>\n",
       "      <th>n_bathrooms</th>\n",
       "      <th>is_floor_under</th>\n",
       "      <th>rent_price</th>\n",
       "      <th>has_parking</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15884</th>\n",
       "      <td>78.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>724.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12837</th>\n",
       "      <td>149.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1966.0</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16432</th>\n",
       "      <td>60.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>625.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18767</th>\n",
       "      <td>56.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1197.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20433</th>\n",
       "      <td>59.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>884.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20290</th>\n",
       "      <td>82.0</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1250.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11924</th>\n",
       "      <td>150.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2515.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17715</th>\n",
       "      <td>60.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1225.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6430</th>\n",
       "      <td>130.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1405.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2813</th>\n",
       "      <td>52.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1041.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3862 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       sq_mt_built  n_rooms  n_bathrooms  is_floor_under  rent_price  \\\n",
       "15884         78.0        3          1.0             0.0       724.0   \n",
       "12837        149.0        4          2.0             0.0      1966.0   \n",
       "16432         60.0        2          1.0             0.0       625.0   \n",
       "18767         56.0        1          1.0             0.0      1197.0   \n",
       "20433         59.0        2          2.0             1.0       884.0   \n",
       "...            ...      ...          ...             ...         ...   \n",
       "20290         82.0        3          2.0             0.0      1250.0   \n",
       "11924        150.0        2          2.0             0.0      2515.0   \n",
       "17715         60.0        1          1.0             0.0      1225.0   \n",
       "6430         130.0        4          2.0             0.0      1405.0   \n",
       "2813          52.0        1          1.0             1.0      1041.0   \n",
       "\n",
       "       has_parking  \n",
       "15884        False  \n",
       "12837         True  \n",
       "16432        False  \n",
       "18767        False  \n",
       "20433        False  \n",
       "...            ...  \n",
       "20290        False  \n",
       "11924        False  \n",
       "17715        False  \n",
       "6430         False  \n",
       "2813         False  \n",
       "\n",
       "[3862 rows x 6 columns]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Aplicamos el método .transform() sobre X_test (como los datos ya fueron ajustados, solo se necesita escalarlos)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.06313646, 0.125     , 0.        , 0.        , 0.05618335,\n",
       "        0.        ],\n",
       "       [0.13543788, 0.16666667, 0.07692308, 0.        , 0.15502149,\n",
       "        1.        ],\n",
       "       [0.04480652, 0.08333333, 0.        , 0.        , 0.04830495,\n",
       "        0.        ],\n",
       "       ...,\n",
       "       [0.04480652, 0.04166667, 0.        , 0.        , 0.09605284,\n",
       "        0.        ],\n",
       "       [0.11608961, 0.16666667, 0.07692308, 0.        , 0.11037721,\n",
       "        0.        ],\n",
       "       [0.03665988, 0.04166667, 0.        , 1.        , 0.08141015,\n",
       "        0.        ]])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Creación del modelo"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Para un problema de regresión utilizamos como optimizador 'adam' y en el argumento loss indicamos 'mse'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importamos las clases necesarias para la creación del modelo\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Activation\n",
    "from keras.optimizers import Adam"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Creamos un modelo secuencial\n",
    "model = Sequential()\n",
    "\n",
    "# Indicamos un número de neuronas correspondiente al número de variables en X_train (recordar que .shape retorna una tupla (celdas, columnas))\n",
    "num_neuronas = X_train.shape[1]\n",
    "\n",
    "# Añadimos 4 capas de neuronas con función de activación relu\n",
    "model.add(Dense(num_neuronas, activation='relu'))\n",
    "model.add(Dense(num_neuronas, activation='relu'))\n",
    "model.add(Dense(num_neuronas, activation='relu'))\n",
    "model.add(Dense(num_neuronas, activation='relu'))\n",
    "\n",
    "# Añadimos una capa de salida con una neurona por cada variable de salida\n",
    "model.add(Dense(1))\n",
    "\n",
    "# Compilamos el modelo indicando el optimizador a utilizar (optimizer) y el cálculo para valorar las pérdidas (loss)\n",
    "model.compile(optimizer='adam', loss='mse')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Entrenamiento del modelo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/300\n",
      "121/121 [==============================] - 2s 7ms/step - loss: 935847657472.0000 - val_loss: 850748375040.0000\n",
      "Epoch 2/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 935832846336.0000 - val_loss: 850714558464.0000\n",
      "Epoch 3/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 935735132160.0000 - val_loss: 850523914240.0000\n",
      "Epoch 4/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 935325663232.0000 - val_loss: 849850728448.0000\n",
      "Epoch 5/300\n",
      "121/121 [==============================] - 1s 6ms/step - loss: 934125961216.0000 - val_loss: 848086171648.0000\n",
      "Epoch 6/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 931249324032.0000 - val_loss: 844098043904.0000\n",
      "Epoch 7/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 925294460928.0000 - val_loss: 836428169216.0000\n",
      "Epoch 8/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 914575327232.0000 - val_loss: 823365009408.0000\n",
      "Epoch 9/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 897245184000.0000 - val_loss: 802991702016.0000\n",
      "Epoch 10/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 871478067200.0000 - val_loss: 773714870272.0000\n",
      "Epoch 11/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 835650650112.0000 - val_loss: 734722719744.0000\n",
      "Epoch 12/300\n",
      "121/121 [==============================] - 1s 8ms/step - loss: 789286092800.0000 - val_loss: 685792100352.0000\n",
      "Epoch 13/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 732967141376.0000 - val_loss: 628415004672.0000\n",
      "Epoch 14/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 669320937472.0000 - val_loss: 566137454592.0000\n",
      "Epoch 15/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 603050999808.0000 - val_loss: 504621105152.0000\n",
      "Epoch 16/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 539556806656.0000 - val_loss: 448086016000.0000\n",
      "Epoch 17/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 483782787072.0000 - val_loss: 402389762048.0000\n",
      "Epoch 18/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 439482646528.0000 - val_loss: 368278765568.0000\n",
      "Epoch 19/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 407998267392.0000 - val_loss: 346607747072.0000\n",
      "Epoch 20/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 387914530816.0000 - val_loss: 333608026112.0000\n",
      "Epoch 21/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 375746985984.0000 - val_loss: 326219202560.0000\n",
      "Epoch 22/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 368063414272.0000 - val_loss: 321377370112.0000\n",
      "Epoch 23/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 362446159872.0000 - val_loss: 317301784576.0000\n",
      "Epoch 24/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 357624643584.0000 - val_loss: 313522094080.0000\n",
      "Epoch 25/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 353092206592.0000 - val_loss: 309736046592.0000\n",
      "Epoch 26/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 348633956352.0000 - val_loss: 305882595328.0000\n",
      "Epoch 27/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 344162795520.0000 - val_loss: 301939556352.0000\n",
      "Epoch 28/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 339646480384.0000 - val_loss: 297943826432.0000\n",
      "Epoch 29/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 335036940288.0000 - val_loss: 293782650880.0000\n",
      "Epoch 30/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 330371858432.0000 - val_loss: 289642348544.0000\n",
      "Epoch 31/300\n",
      "121/121 [==============================] - 1s 6ms/step - loss: 325611028480.0000 - val_loss: 285355278336.0000\n",
      "Epoch 32/300\n",
      "121/121 [==============================] - 1s 7ms/step - loss: 320733315072.0000 - val_loss: 281015189504.0000\n",
      "Epoch 33/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 315743797248.0000 - val_loss: 276561592320.0000\n",
      "Epoch 34/300\n",
      "121/121 [==============================] - 1s 6ms/step - loss: 310644146176.0000 - val_loss: 272006610944.0000\n",
      "Epoch 35/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 305456676864.0000 - val_loss: 267348787200.0000\n",
      "Epoch 36/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 300125618176.0000 - val_loss: 262620594176.0000\n",
      "Epoch 37/300\n",
      "121/121 [==============================] - 1s 7ms/step - loss: 294726729728.0000 - val_loss: 257701986304.0000\n",
      "Epoch 38/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 289233207296.0000 - val_loss: 252687237120.0000\n",
      "Epoch 39/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 283519385600.0000 - val_loss: 247586684928.0000\n",
      "Epoch 40/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 277741043712.0000 - val_loss: 242380750848.0000\n",
      "Epoch 41/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 271780413440.0000 - val_loss: 237047349248.0000\n",
      "Epoch 42/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 265744777216.0000 - val_loss: 231688486912.0000\n",
      "Epoch 43/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 259619864576.0000 - val_loss: 226263842816.0000\n",
      "Epoch 44/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 253447962624.0000 - val_loss: 220654026752.0000\n",
      "Epoch 45/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 247170990080.0000 - val_loss: 215022829568.0000\n",
      "Epoch 46/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 240850190336.0000 - val_loss: 209441783808.0000\n",
      "Epoch 47/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 234395844608.0000 - val_loss: 203523850240.0000\n",
      "Epoch 48/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 227938238464.0000 - val_loss: 197707055104.0000\n",
      "Epoch 49/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 221399498752.0000 - val_loss: 191895011328.0000\n",
      "Epoch 50/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 214852632576.0000 - val_loss: 185978667008.0000\n",
      "Epoch 51/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 208220979200.0000 - val_loss: 179985580032.0000\n",
      "Epoch 52/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 201592963072.0000 - val_loss: 174079770624.0000\n",
      "Epoch 53/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 195006103552.0000 - val_loss: 168141537280.0000\n",
      "Epoch 54/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 188326412288.0000 - val_loss: 162177204224.0000\n",
      "Epoch 55/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 181802532864.0000 - val_loss: 156284747776.0000\n",
      "Epoch 56/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 175265366016.0000 - val_loss: 150593748992.0000\n",
      "Epoch 57/300\n",
      "121/121 [==============================] - 1s 9ms/step - loss: 168895922176.0000 - val_loss: 144716726272.0000\n",
      "Epoch 58/300\n",
      "121/121 [==============================] - 1s 10ms/step - loss: 162671362048.0000 - val_loss: 139282235392.0000\n",
      "Epoch 59/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 156621455360.0000 - val_loss: 133998682112.0000\n",
      "Epoch 60/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 150800744448.0000 - val_loss: 128895139840.0000\n",
      "Epoch 61/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 145293099008.0000 - val_loss: 124080758784.0000\n",
      "Epoch 62/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 140103761920.0000 - val_loss: 119667810304.0000\n",
      "Epoch 63/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 135314841600.0000 - val_loss: 115544350720.0000\n",
      "Epoch 64/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 130928041984.0000 - val_loss: 111904661504.0000\n",
      "Epoch 65/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 127000952832.0000 - val_loss: 108672270336.0000\n",
      "Epoch 66/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 123463720960.0000 - val_loss: 105833357312.0000\n",
      "Epoch 67/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 120295170048.0000 - val_loss: 103342317568.0000\n",
      "Epoch 68/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 117548867584.0000 - val_loss: 101089460224.0000\n",
      "Epoch 69/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 115119489024.0000 - val_loss: 99201122304.0000\n",
      "Epoch 70/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 113040400384.0000 - val_loss: 97574281216.0000\n",
      "Epoch 71/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 111265529856.0000 - val_loss: 96247701504.0000\n",
      "Epoch 72/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 109703348224.0000 - val_loss: 94984167424.0000\n",
      "Epoch 73/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 108340658176.0000 - val_loss: 93955563520.0000\n",
      "Epoch 74/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 107113938944.0000 - val_loss: 93018103808.0000\n",
      "Epoch 75/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 106113007616.0000 - val_loss: 92231180288.0000\n",
      "Epoch 76/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 105142550528.0000 - val_loss: 91439742976.0000\n",
      "Epoch 77/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 104285208576.0000 - val_loss: 90780884992.0000\n",
      "Epoch 78/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 103428669440.0000 - val_loss: 90092093440.0000\n",
      "Epoch 79/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 102653771776.0000 - val_loss: 89429745664.0000\n",
      "Epoch 80/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 101884067840.0000 - val_loss: 88812150784.0000\n",
      "Epoch 81/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 101212119040.0000 - val_loss: 88145829888.0000\n",
      "Epoch 82/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 100454989824.0000 - val_loss: 87522738176.0000\n",
      "Epoch 83/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 99745538048.0000 - val_loss: 86896320512.0000\n",
      "Epoch 84/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 99025207296.0000 - val_loss: 86276571136.0000\n",
      "Epoch 85/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 98295021568.0000 - val_loss: 85645983744.0000\n",
      "Epoch 86/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 97604902912.0000 - val_loss: 85032345600.0000\n",
      "Epoch 87/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 96914202624.0000 - val_loss: 84393730048.0000\n",
      "Epoch 88/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 96201711616.0000 - val_loss: 83772850176.0000\n",
      "Epoch 89/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 95516606464.0000 - val_loss: 83140739072.0000\n",
      "Epoch 90/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 94848319488.0000 - val_loss: 82543812608.0000\n",
      "Epoch 91/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 94140375040.0000 - val_loss: 81906810880.0000\n",
      "Epoch 92/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 93464895488.0000 - val_loss: 81290797056.0000\n",
      "Epoch 93/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 92771000320.0000 - val_loss: 80717217792.0000\n",
      "Epoch 94/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 92089466880.0000 - val_loss: 80111681536.0000\n",
      "Epoch 95/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 91407024128.0000 - val_loss: 79509217280.0000\n",
      "Epoch 96/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 90775011328.0000 - val_loss: 78932688896.0000\n",
      "Epoch 97/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 90069450752.0000 - val_loss: 78286176256.0000\n",
      "Epoch 98/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 89393258496.0000 - val_loss: 77687947264.0000\n",
      "Epoch 99/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 88784109568.0000 - val_loss: 77083795456.0000\n",
      "Epoch 100/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 88100093952.0000 - val_loss: 76495020032.0000\n",
      "Epoch 101/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 87464796160.0000 - val_loss: 75915182080.0000\n",
      "Epoch 102/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 86794166272.0000 - val_loss: 75344093184.0000\n",
      "Epoch 103/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 86154878976.0000 - val_loss: 74761912320.0000\n",
      "Epoch 104/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 85516001280.0000 - val_loss: 74192912384.0000\n",
      "Epoch 105/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 84849901568.0000 - val_loss: 73600319488.0000\n",
      "Epoch 106/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 84213104640.0000 - val_loss: 73028870144.0000\n",
      "Epoch 107/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 83580067840.0000 - val_loss: 72463474688.0000\n",
      "Epoch 108/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 82932498432.0000 - val_loss: 71887618048.0000\n",
      "Epoch 109/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 82315927552.0000 - val_loss: 71315726336.0000\n",
      "Epoch 110/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 81648754688.0000 - val_loss: 70764584960.0000\n",
      "Epoch 111/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 81033945088.0000 - val_loss: 70219382784.0000\n",
      "Epoch 112/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 80393936896.0000 - val_loss: 69646966784.0000\n",
      "Epoch 113/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 79766487040.0000 - val_loss: 69096128512.0000\n",
      "Epoch 114/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 79136301056.0000 - val_loss: 68537815040.0000\n",
      "Epoch 115/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 78508802048.0000 - val_loss: 68028522496.0000\n",
      "Epoch 116/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 77889011712.0000 - val_loss: 67464540160.0000\n",
      "Epoch 117/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 77277061120.0000 - val_loss: 66935013376.0000\n",
      "Epoch 118/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 76675948544.0000 - val_loss: 66382934016.0000\n",
      "Epoch 119/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 76073205760.0000 - val_loss: 65857249280.0000\n",
      "Epoch 120/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 75470684160.0000 - val_loss: 65368420352.0000\n",
      "Epoch 121/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 74832871424.0000 - val_loss: 64796401664.0000\n",
      "Epoch 122/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 74243481600.0000 - val_loss: 64280252416.0000\n",
      "Epoch 123/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 73645785088.0000 - val_loss: 63760334848.0000\n",
      "Epoch 124/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 73052135424.0000 - val_loss: 63219142656.0000\n",
      "Epoch 125/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 72452603904.0000 - val_loss: 62693888000.0000\n",
      "Epoch 126/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 71855816704.0000 - val_loss: 62198362112.0000\n",
      "Epoch 127/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 71269015552.0000 - val_loss: 61671120896.0000\n",
      "Epoch 128/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 70694461440.0000 - val_loss: 61159157760.0000\n",
      "Epoch 129/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 70119079936.0000 - val_loss: 60671004672.0000\n",
      "Epoch 130/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 69565661184.0000 - val_loss: 60150468608.0000\n",
      "Epoch 131/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 69027201024.0000 - val_loss: 59661312000.0000\n",
      "Epoch 132/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 68423016448.0000 - val_loss: 59164442624.0000\n",
      "Epoch 133/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 67844714496.0000 - val_loss: 58676736000.0000\n",
      "Epoch 134/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 67328040960.0000 - val_loss: 58171830272.0000\n",
      "Epoch 135/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 66773344256.0000 - val_loss: 57710481408.0000\n",
      "Epoch 136/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 66178330624.0000 - val_loss: 57208147968.0000\n",
      "Epoch 137/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 65656512512.0000 - val_loss: 56739672064.0000\n",
      "Epoch 138/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 65106759680.0000 - val_loss: 56285724672.0000\n",
      "Epoch 139/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 64596279296.0000 - val_loss: 55777255424.0000\n",
      "Epoch 140/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 64030355456.0000 - val_loss: 55309180928.0000\n",
      "Epoch 141/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 63494422528.0000 - val_loss: 54832418816.0000\n",
      "Epoch 142/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 62964129792.0000 - val_loss: 54372782080.0000\n",
      "Epoch 143/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 62464155648.0000 - val_loss: 53911842816.0000\n",
      "Epoch 144/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 61920407552.0000 - val_loss: 53463805952.0000\n",
      "Epoch 145/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 61413629952.0000 - val_loss: 53043314688.0000\n",
      "Epoch 146/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 60920520704.0000 - val_loss: 52571545600.0000\n",
      "Epoch 147/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 60396089344.0000 - val_loss: 52124676096.0000\n",
      "Epoch 148/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 59919753216.0000 - val_loss: 51705401344.0000\n",
      "Epoch 149/300\n",
      "121/121 [==============================] - 1s 6ms/step - loss: 59408527360.0000 - val_loss: 51263344640.0000\n",
      "Epoch 150/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 58936623104.0000 - val_loss: 50838880256.0000\n",
      "Epoch 151/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 58464284672.0000 - val_loss: 50426945536.0000\n",
      "Epoch 152/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 57980215296.0000 - val_loss: 50019233792.0000\n",
      "Epoch 153/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 57520795648.0000 - val_loss: 49612972032.0000\n",
      "Epoch 154/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 57045753856.0000 - val_loss: 49236262912.0000\n",
      "Epoch 155/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 56617091072.0000 - val_loss: 48818003968.0000\n",
      "Epoch 156/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 56183758848.0000 - val_loss: 48454098944.0000\n",
      "Epoch 157/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 55754715136.0000 - val_loss: 48071532544.0000\n",
      "Epoch 158/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 55321382912.0000 - val_loss: 47708155904.0000\n",
      "Epoch 159/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 54901530624.0000 - val_loss: 47344545792.0000\n",
      "Epoch 160/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 54489436160.0000 - val_loss: 46995189760.0000\n",
      "Epoch 161/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 54083264512.0000 - val_loss: 46661296128.0000\n",
      "Epoch 162/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 53701439488.0000 - val_loss: 46328946688.0000\n",
      "Epoch 163/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 53304700928.0000 - val_loss: 46028869632.0000\n",
      "Epoch 164/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 52937007104.0000 - val_loss: 45689864192.0000\n",
      "Epoch 165/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 52558905344.0000 - val_loss: 45375328256.0000\n",
      "Epoch 166/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 52212371456.0000 - val_loss: 45080678400.0000\n",
      "Epoch 167/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 51869192192.0000 - val_loss: 44781486080.0000\n",
      "Epoch 168/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 51532640256.0000 - val_loss: 44503547904.0000\n",
      "Epoch 169/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 51198423040.0000 - val_loss: 44235780096.0000\n",
      "Epoch 170/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 50875822080.0000 - val_loss: 43961323520.0000\n",
      "Epoch 171/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 50554159104.0000 - val_loss: 43736141824.0000\n",
      "Epoch 172/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 50300940288.0000 - val_loss: 43457925120.0000\n",
      "Epoch 173/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 49973391360.0000 - val_loss: 43258208256.0000\n",
      "Epoch 174/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 49713135616.0000 - val_loss: 43003461632.0000\n",
      "Epoch 175/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 49427476480.0000 - val_loss: 42760011776.0000\n",
      "Epoch 176/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 49178521600.0000 - val_loss: 42553626624.0000\n",
      "Epoch 177/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 48926003200.0000 - val_loss: 42336239616.0000\n",
      "Epoch 178/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 48678117376.0000 - val_loss: 42146344960.0000\n",
      "Epoch 179/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 48415805440.0000 - val_loss: 41936855040.0000\n",
      "Epoch 180/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 48218750976.0000 - val_loss: 41740382208.0000\n",
      "Epoch 181/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 47971508224.0000 - val_loss: 41557377024.0000\n",
      "Epoch 182/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 47758721024.0000 - val_loss: 41373392896.0000\n",
      "Epoch 183/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 47533068288.0000 - val_loss: 41217028096.0000\n",
      "Epoch 184/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 47331274752.0000 - val_loss: 41065680896.0000\n",
      "Epoch 185/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 47114387456.0000 - val_loss: 40885460992.0000\n",
      "Epoch 186/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 46927785984.0000 - val_loss: 40738275328.0000\n",
      "Epoch 187/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 46742155264.0000 - val_loss: 40561897472.0000\n",
      "Epoch 188/300\n",
      "121/121 [==============================] - 1s 7ms/step - loss: 46527868928.0000 - val_loss: 40539230208.0000\n",
      "Epoch 189/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 46377222144.0000 - val_loss: 40323174400.0000\n",
      "Epoch 190/300\n",
      "121/121 [==============================] - 1s 6ms/step - loss: 46166323200.0000 - val_loss: 40121077760.0000\n",
      "Epoch 191/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 46011740160.0000 - val_loss: 39999406080.0000\n",
      "Epoch 192/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 45838991360.0000 - val_loss: 39856168960.0000\n",
      "Epoch 193/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 45683200000.0000 - val_loss: 39722516480.0000\n",
      "Epoch 194/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 45538185216.0000 - val_loss: 39600979968.0000\n",
      "Epoch 195/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 45380767744.0000 - val_loss: 39560433664.0000\n",
      "Epoch 196/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 45224824832.0000 - val_loss: 39402536960.0000\n",
      "Epoch 197/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 45075365888.0000 - val_loss: 39253938176.0000\n",
      "Epoch 198/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 44926369792.0000 - val_loss: 39217516544.0000\n",
      "Epoch 199/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 44803411968.0000 - val_loss: 39036903424.0000\n",
      "Epoch 200/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 44678844416.0000 - val_loss: 38967787520.0000\n",
      "Epoch 201/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 44532506624.0000 - val_loss: 38856826880.0000\n",
      "Epoch 202/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 44419002368.0000 - val_loss: 38779604992.0000\n",
      "Epoch 203/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 44291796992.0000 - val_loss: 38708170752.0000\n",
      "Epoch 204/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 44163940352.0000 - val_loss: 38570532864.0000\n",
      "Epoch 205/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 44050558976.0000 - val_loss: 38476599296.0000\n",
      "Epoch 206/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 43931656192.0000 - val_loss: 38531117056.0000\n",
      "Epoch 207/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 43854049280.0000 - val_loss: 38318567424.0000\n",
      "Epoch 208/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 43710689280.0000 - val_loss: 38252077056.0000\n",
      "Epoch 209/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 43610841088.0000 - val_loss: 38124023808.0000\n",
      "Epoch 210/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 43506458624.0000 - val_loss: 38054637568.0000\n",
      "Epoch 211/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 43393064960.0000 - val_loss: 37979291648.0000\n",
      "Epoch 212/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 43296854016.0000 - val_loss: 37902581760.0000\n",
      "Epoch 213/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 43174756352.0000 - val_loss: 37866893312.0000\n",
      "Epoch 214/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 43077967872.0000 - val_loss: 37728489472.0000\n",
      "Epoch 215/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 43000418304.0000 - val_loss: 37672153088.0000\n",
      "Epoch 216/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 42894446592.0000 - val_loss: 37599764480.0000\n",
      "Epoch 217/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 42788003840.0000 - val_loss: 37580771328.0000\n",
      "Epoch 218/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 42690007040.0000 - val_loss: 37450145792.0000\n",
      "Epoch 219/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 42600062976.0000 - val_loss: 37377622016.0000\n",
      "Epoch 220/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 42514526208.0000 - val_loss: 37314105344.0000\n",
      "Epoch 221/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 42424893440.0000 - val_loss: 37263835136.0000\n",
      "Epoch 222/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 42335125504.0000 - val_loss: 37200302080.0000\n",
      "Epoch 223/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 42264764416.0000 - val_loss: 37129510912.0000\n",
      "Epoch 224/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 42202894336.0000 - val_loss: 37132349440.0000\n",
      "Epoch 225/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 42084806656.0000 - val_loss: 36995756032.0000\n",
      "Epoch 226/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 42012209152.0000 - val_loss: 36932276224.0000\n",
      "Epoch 227/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 41920917504.0000 - val_loss: 36868091904.0000\n",
      "Epoch 228/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 41840766976.0000 - val_loss: 36836872192.0000\n",
      "Epoch 229/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 41765969920.0000 - val_loss: 36743835648.0000\n",
      "Epoch 230/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 41681469440.0000 - val_loss: 36720193536.0000\n",
      "Epoch 231/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 41599508480.0000 - val_loss: 36735115264.0000\n",
      "Epoch 232/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 41538654208.0000 - val_loss: 36599660544.0000\n",
      "Epoch 233/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 41445650432.0000 - val_loss: 36581376000.0000\n",
      "Epoch 234/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 41380634624.0000 - val_loss: 36453433344.0000\n",
      "Epoch 235/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 41299185664.0000 - val_loss: 36380790784.0000\n",
      "Epoch 236/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 41230303232.0000 - val_loss: 36358037504.0000\n",
      "Epoch 237/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 41138094080.0000 - val_loss: 36277088256.0000\n",
      "Epoch 238/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 41089245184.0000 - val_loss: 36266602496.0000\n",
      "Epoch 239/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 41010528256.0000 - val_loss: 36222840832.0000\n",
      "Epoch 240/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 40934387712.0000 - val_loss: 36226174976.0000\n",
      "Epoch 241/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 40884723712.0000 - val_loss: 36147884032.0000\n",
      "Epoch 242/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 40815013888.0000 - val_loss: 36073037824.0000\n",
      "Epoch 243/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 40762322944.0000 - val_loss: 36002414592.0000\n",
      "Epoch 244/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 40681316352.0000 - val_loss: 35987468288.0000\n",
      "Epoch 245/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 40608350208.0000 - val_loss: 35891392512.0000\n",
      "Epoch 246/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 40558653440.0000 - val_loss: 35860283392.0000\n",
      "Epoch 247/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 40490180608.0000 - val_loss: 35811549184.0000\n",
      "Epoch 248/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 40421470208.0000 - val_loss: 35793010688.0000\n",
      "Epoch 249/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 40378007552.0000 - val_loss: 35726663680.0000\n",
      "Epoch 250/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 40310255616.0000 - val_loss: 35691671552.0000\n",
      "Epoch 251/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 40243515392.0000 - val_loss: 35637493760.0000\n",
      "Epoch 252/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 40185081856.0000 - val_loss: 35686674432.0000\n",
      "Epoch 253/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 40123240448.0000 - val_loss: 35562512384.0000\n",
      "Epoch 254/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 40070225920.0000 - val_loss: 35525304320.0000\n",
      "Epoch 255/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39994560512.0000 - val_loss: 35533398016.0000\n",
      "Epoch 256/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39951097856.0000 - val_loss: 35416219648.0000\n",
      "Epoch 257/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39917481984.0000 - val_loss: 35341037568.0000\n",
      "Epoch 258/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 39830474752.0000 - val_loss: 35336060928.0000\n",
      "Epoch 259/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39784845312.0000 - val_loss: 35257307136.0000\n",
      "Epoch 260/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 39723102208.0000 - val_loss: 35237126144.0000\n",
      "Epoch 261/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 39667490816.0000 - val_loss: 35207401472.0000\n",
      "Epoch 262/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 39620423680.0000 - val_loss: 35237236736.0000\n",
      "Epoch 263/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39576776704.0000 - val_loss: 35114618880.0000\n",
      "Epoch 264/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 39524794368.0000 - val_loss: 35174486016.0000\n",
      "Epoch 265/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39508897792.0000 - val_loss: 35082604544.0000\n",
      "Epoch 266/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39423090688.0000 - val_loss: 35020009472.0000\n",
      "Epoch 267/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39375151104.0000 - val_loss: 35007008768.0000\n",
      "Epoch 268/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39340912640.0000 - val_loss: 34986270720.0000\n",
      "Epoch 269/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39283650560.0000 - val_loss: 34903658496.0000\n",
      "Epoch 270/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 39243411456.0000 - val_loss: 34897870848.0000\n",
      "Epoch 271/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 39188836352.0000 - val_loss: 34829975552.0000\n",
      "Epoch 272/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39163977728.0000 - val_loss: 35001851904.0000\n",
      "Epoch 273/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39136481280.0000 - val_loss: 34808401920.0000\n",
      "Epoch 274/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39086809088.0000 - val_loss: 34799079424.0000\n",
      "Epoch 275/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 39053316096.0000 - val_loss: 34757996544.0000\n",
      "Epoch 276/300\n",
      "121/121 [==============================] - 1s 5ms/step - loss: 39007612928.0000 - val_loss: 34731544576.0000\n",
      "Epoch 277/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38976532480.0000 - val_loss: 34691313664.0000\n",
      "Epoch 278/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38932193280.0000 - val_loss: 34681348096.0000\n",
      "Epoch 279/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38898397184.0000 - val_loss: 34687836160.0000\n",
      "Epoch 280/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38866997248.0000 - val_loss: 34669424640.0000\n",
      "Epoch 281/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38823821312.0000 - val_loss: 34585108480.0000\n",
      "Epoch 282/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38793146368.0000 - val_loss: 34618871808.0000\n",
      "Epoch 283/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38759211008.0000 - val_loss: 34597257216.0000\n",
      "Epoch 284/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38721077248.0000 - val_loss: 34530816000.0000\n",
      "Epoch 285/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38690811904.0000 - val_loss: 34500263936.0000\n",
      "Epoch 286/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38666280960.0000 - val_loss: 34576859136.0000\n",
      "Epoch 287/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 38637932544.0000 - val_loss: 34501349376.0000\n",
      "Epoch 288/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38608535552.0000 - val_loss: 34519732224.0000\n",
      "Epoch 289/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38587604992.0000 - val_loss: 34449141760.0000\n",
      "Epoch 290/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38540320768.0000 - val_loss: 34461659136.0000\n",
      "Epoch 291/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38523654144.0000 - val_loss: 34398896128.0000\n",
      "Epoch 292/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38501605376.0000 - val_loss: 34395062272.0000\n",
      "Epoch 293/300\n",
      "121/121 [==============================] - 0s 4ms/step - loss: 38477549568.0000 - val_loss: 34412072960.0000\n",
      "Epoch 294/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38441717760.0000 - val_loss: 34347227136.0000\n",
      "Epoch 295/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38415355904.0000 - val_loss: 34384113664.0000\n",
      "Epoch 296/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38387347456.0000 - val_loss: 34339186688.0000\n",
      "Epoch 297/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38344839168.0000 - val_loss: 34307553280.0000\n",
      "Epoch 298/300\n",
      "121/121 [==============================] - 1s 4ms/step - loss: 38319263744.0000 - val_loss: 34270537728.0000\n",
      "Epoch 299/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38300692480.0000 - val_loss: 34299430912.0000\n",
      "Epoch 300/300\n",
      "121/121 [==============================] - 0s 3ms/step - loss: 38281662464.0000 - val_loss: 34265344000.0000\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x14238ba1d80>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "Para realizar el entrenamiento del modelo utilizamos el método .fit(), donde indicamos los siguientes parámetros:\n",
    "\n",
    "    > x= Conjunto X de entrenamiento\n",
    "\n",
    "    > y= Valores del conjunto Y de entrenamiento\n",
    "\n",
    "    > validation_data= Tupla con el conjunto X de prueba y los valores del conjunto Y de prueba. Si se indica, calcula sobre el conjunto de pruebas el resultado de \n",
    "                       pérdidas. Es recomendable realizarlo ya que, de esta forma, se puede realizar luego un gráfico para comparar las pérdidas durante el \n",
    "                       entrenamiento y durante la prueba\n",
    "\n",
    "    > batch_size= Tamaño de datos que toma en cada cálculo (se recomienda indicar valores que sean potencia de 2). Mientras mas bajo, mayor velocidad de entrenamiento \n",
    "                  pero corre riesgo de no aprender las características y detalles que pueden ser significativos en la predicción. Mientras más alto, menor velocidad\n",
    "                  entrenamiento pero mayor probabilidad de aprendizaje\n",
    "\n",
    "    > epochs= Cantidad de iteraciones de datos que se harán durante el aprendizaje. Un número muy bajo puede resultar en Underfitting y un número muy alto puede \n",
    "              resultar en Overfitting.\n",
    "\"\"\"\n",
    "model.fit(\n",
    "    x=X_train,\n",
    "    y=y_train.values,\n",
    "    validation_data=(X_test, y_test.values),\n",
    "    batch_size=128,\n",
    "    epochs=300\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cargamos los datos de las perdidas en un dataframe a partir de método .history.history\n",
    "losses = pd.DataFrame(model.history.history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot: >"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualizamos un plot de este dataframe para comparar las pérdidas en el entrenamiento y la prueba\n",
    "losses.plot(figsize=(10, 5))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluación del modelo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importamos los métodos necesarios desde la librería sklearn (https://scikit-learn.org/stable/modules/model_evaluation.html#regression-metrics)\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error, explained_variance_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "121/121 [==============================] - 3s 5ms/step\n"
     ]
    }
   ],
   "source": [
    "# Realizamos la predicción de nuestro modelo con el conjunto de prueba (x) e instanciamos el resultado de cada entrada en una variable\n",
    "predictions = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "109868.39278142802"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Comparamos los valores reales del conjunto de prueba (y) y los comparamos con las predicciones utilizando mean_absolute_error()\n",
    "mean_absolute_error(y_test, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "185108.9987387719"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Realizamos la raíz cuadrática (con el método sqrt() de Numpy) del error cuadrático medio (mean_squared_error() de sklearn)\n",
    "np.sqrt(mean_squared_error(y_test, predictions))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9266042481106153"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calculamos la variancia (explained_variance_score) entre los valores reales del cojunto de prueba (y) y las predicciones, cuanto más cercano a 1 mejor ya que\n",
    "# significa que conforme más aumenta o disminuye el valor real, más aumenta o disminuye el valor predecido\n",
    "explained_variance_score(y_test, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.1736796115719512"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Comparamos el valor absoluto (mean_absolute_error) con el promedio de la variable objetivo. Esto nos muestra en qué porcentaje nos estamos\n",
    "# equivaocando respecto al promedio (más recomendable si hay valores outliers)\n",
    "mean_absolute_error(y_test, predictions)/df['buy_price'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.28912734942481055"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Comparamos el valor absoluto (mean_absolute_error) con la mediana de la variable objetivo. Esto nos muestra en qué porcentaje nos estamos\n",
    "# equivocando respecto a la mediana\n",
    "mean_absolute_error(y_test, predictions)/df['buy_price'].median()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x14238c53ac0>]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualizamos un plot con la comparación entre la predicción realizada y los datos reales\n",
    "# Nuestra predicción\n",
    "plt.scatter(y_test, predictions)\n",
    "plt.xlabel('Valor Real')\n",
    "plt.ylabel('Predicción')\n",
    "\n",
    "# Predición \"perfecta\"\n",
    "plt.plot(y_test, y_test, 'r')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x142455468c0>"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 563.5x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Evaluamos los errores rotando los valores reales y restándoles las predicciones\n",
    "errors = y_test.values.reshape(y_test.shape[0], 1) - predictions\n",
    "\n",
    "# Visualizamos un displot de los errores. Lo ideal es que estos se encuentren centrados en el 0\n",
    "sns.displot(errors, kind='kde', fill=True, palette=sns.color_palette('bright')[:1])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicción con nuevos datos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "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>sq_mt_built</th>\n",
       "      <th>n_rooms</th>\n",
       "      <th>n_bathrooms</th>\n",
       "      <th>is_floor_under</th>\n",
       "      <th>rent_price</th>\n",
       "      <th>buy_price</th>\n",
       "      <th>has_parking</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>65</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>420</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>72</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>650</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>97</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>730</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>63</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>True</td>\n",
       "      <td>590</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>107</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>125</td>\n",
       "      <td>4</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>890</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>130</td>\n",
       "      <td>5</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>885</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>135</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>912</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>76</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>970</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>87</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>701</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sq_mt_built  n_rooms  n_bathrooms  is_floor_under  rent_price  buy_price  \\\n",
       "0           65        2          1.0           False         420        NaN   \n",
       "1           72        3          1.0           False         650        NaN   \n",
       "2           97        2          2.0           False         730        NaN   \n",
       "3           63        2          1.0            True         590        NaN   \n",
       "4          107        2          2.0           False        1000        NaN   \n",
       "5          125        4          2.0           False         890        NaN   \n",
       "6          130        5          2.0           False         885        NaN   \n",
       "7          135        3          2.0           False         912        NaN   \n",
       "8           76        3          2.0           False         970        NaN   \n",
       "9           87        2          1.0           False         701        NaN   \n",
       "\n",
       "   has_parking  \n",
       "0        False  \n",
       "1        False  \n",
       "2        False  \n",
       "3        False  \n",
       "4         True  \n",
       "5         True  \n",
       "6         True  \n",
       "7        False  \n",
       "8         True  \n",
       "9        False  "
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Insertamos los nuevos datos de entrada en un dataframe\n",
    "df_pred = pd.read_excel('./datasets/viviendas_Madrid.xlsx', sheet_name='Datos_Predecir')\n",
    "\n",
    "# Filtramos el nuevo dataframe para quedarnos con las mismas columnas de interés que en el dataframe de datos históricos\n",
    "df_pred = df_pred[df.columns]\n",
    "\n",
    "df_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "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>sq_mt_built</th>\n",
       "      <th>n_rooms</th>\n",
       "      <th>n_bathrooms</th>\n",
       "      <th>is_floor_under</th>\n",
       "      <th>rent_price</th>\n",
       "      <th>has_parking</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>65</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>420</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>72</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>650</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>97</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>730</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>63</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>True</td>\n",
       "      <td>590</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>107</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1000</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>125</td>\n",
       "      <td>4</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>890</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>130</td>\n",
       "      <td>5</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>885</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>135</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>912</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>76</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>970</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>87</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>701</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sq_mt_built  n_rooms  n_bathrooms  is_floor_under  rent_price  has_parking\n",
       "0           65        2          1.0           False         420        False\n",
       "1           72        3          1.0           False         650        False\n",
       "2           97        2          2.0           False         730        False\n",
       "3           63        2          1.0            True         590        False\n",
       "4          107        2          2.0           False        1000         True\n",
       "5          125        4          2.0           False         890         True\n",
       "6          130        5          2.0           False         885         True\n",
       "7          135        3          2.0           False         912        False\n",
       "8           76        3          2.0           False         970         True\n",
       "9           87        2          1.0           False         701        False"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Eliminamos la columna de la variable objetivo\n",
    "X_pred = df_pred.drop('buy_price', axis=1)\n",
    "\n",
    "X_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.04989817, 0.08333333, 0.        , 0.        , 0.03199109,\n",
       "        0.        ],\n",
       "       [0.05702648, 0.125     , 0.        , 0.        , 0.05029445,\n",
       "        0.        ],\n",
       "       [0.08248473, 0.08333333, 0.07692308, 0.        , 0.05666083,\n",
       "        0.        ],\n",
       "       [0.04786151, 0.08333333, 0.        , 1.        , 0.04551966,\n",
       "        0.        ],\n",
       "       [0.09266802, 0.08333333, 0.07692308, 0.        , 0.07814738,\n",
       "        1.        ],\n",
       "       [0.11099796, 0.16666667, 0.07692308, 0.        , 0.0693936 ,\n",
       "        1.        ],\n",
       "       [0.11608961, 0.20833333, 0.07692308, 0.        , 0.0689957 ,\n",
       "        1.        ],\n",
       "       [0.12118126, 0.125     , 0.07692308, 0.        , 0.07114436,\n",
       "        0.        ],\n",
       "       [0.0610998 , 0.125     , 0.07692308, 0.        , 0.07575999,\n",
       "        1.        ],\n",
       "       [0.07230143, 0.08333333, 0.        , 0.        , 0.05435302,\n",
       "        0.        ]])"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Escalamos las variables de entrada\n",
    "X_pred  = scaler.transform(X_pred)\n",
    "\n",
    "X_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 337ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[ 80540.46],\n",
       "       [137216.75],\n",
       "       [223319.77],\n",
       "       [121027.35],\n",
       "       [315907.53],\n",
       "       [298760.38],\n",
       "       [300748.25],\n",
       "       [317857.1 ],\n",
       "       [275794.5 ],\n",
       "       [162815.89]], dtype=float32)"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Realizamos la predicción\n",
    "resultado = model.predict(X_pred)\n",
    "\n",
    "resultado"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "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>sq_mt_built</th>\n",
       "      <th>n_rooms</th>\n",
       "      <th>n_bathrooms</th>\n",
       "      <th>is_floor_under</th>\n",
       "      <th>rent_price</th>\n",
       "      <th>buy_price</th>\n",
       "      <th>has_parking</th>\n",
       "      <th>PREDICCIÓN</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>65</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>420</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>80540.460938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>72</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>650</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>137216.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>97</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>730</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>223319.765625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>63</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>True</td>\n",
       "      <td>590</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>121027.351562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>107</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>1000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>315907.531250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>125</td>\n",
       "      <td>4</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>890</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>298760.375000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>130</td>\n",
       "      <td>5</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>885</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>300748.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>135</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>912</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>317857.093750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>76</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>False</td>\n",
       "      <td>970</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>275794.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>87</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>701</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>162815.890625</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   sq_mt_built  n_rooms  n_bathrooms  is_floor_under  rent_price  buy_price  \\\n",
       "0           65        2          1.0           False         420        NaN   \n",
       "1           72        3          1.0           False         650        NaN   \n",
       "2           97        2          2.0           False         730        NaN   \n",
       "3           63        2          1.0            True         590        NaN   \n",
       "4          107        2          2.0           False        1000        NaN   \n",
       "5          125        4          2.0           False         890        NaN   \n",
       "6          130        5          2.0           False         885        NaN   \n",
       "7          135        3          2.0           False         912        NaN   \n",
       "8           76        3          2.0           False         970        NaN   \n",
       "9           87        2          1.0           False         701        NaN   \n",
       "\n",
       "   has_parking     PREDICCIÓN  \n",
       "0        False   80540.460938  \n",
       "1        False  137216.750000  \n",
       "2        False  223319.765625  \n",
       "3        False  121027.351562  \n",
       "4         True  315907.531250  \n",
       "5         True  298760.375000  \n",
       "6         True  300748.250000  \n",
       "7        False  317857.093750  \n",
       "8         True  275794.500000  \n",
       "9        False  162815.890625  "
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Agregamos la columna con las predicciones al dataframe\n",
    "df_pred['PREDICCIÓN'] = pd.DataFrame(resultado)\n",
    "\n",
    "df_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Guardar y cargar modelos para usos posteriores"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Guardar modelo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Guardamos el modelo, éste se guarda en la carpeta raíz\n",
    "model.save('modelo_pred_vivienda.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cargar modelo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importamos el método load_model desde keras.models\n",
    "from keras.models import load_model\n",
    "\n",
    "# Cargamos el modelo desde el archivo guardado\n",
    "modelo_viviendas = load_model('modelo_pred_vivienda.h5')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10.2 64-bit",
   "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": "3.10.2"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "26de051ba29f2982a8de78e945f0abaf191376122a1563185a90213a26c5da77"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}