pruebas_tomi.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 429,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "# Importamos utilidades y modelos de sklearn\n",
    "from sklearn.preprocessing import Imputer\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.dummy import DummyRegressor\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.linear_model import Ridge\n",
    "from sklearn.linear_model import BayesianRidge\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "from sklearn.tree import DecisionTreeRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 392,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Métrica de evaluación\n",
    "def RMSLE(actual, pred):\n",
    "    return (np.mean((np.log(actual + 1) - np.log(pred + 1)) ** 2)) **.5\n",
    "\n",
    "def normailize_df(refDf, train):\n",
    "    df = refDf.copy()\n",
    "    df['año'] = df['fecha'].dt.year\n",
    "    df['antiguedad'] = df['antiguedad'] + (2016 - df['año']) # Se normaliza la antiguedad.\n",
    "    df.drop([\"direccion\", 'descripcion', 'lat', 'lng', 'fecha'], axis=1, inplace=True)\n",
    "    \n",
    "    nulltotales = df[df['metrostotales'].isnull()]\n",
    "    nullcubiertos = df[df['metroscubiertos'].isnull()]\n",
    "    notnullapart = df[(~ df['metrostotales'].isnull()) & (df['metrostotales'] < df['metroscubiertos'])]\n",
    "    notnullapart2 = df[(~ df['metrostotales'].isnull()) & (df['metrostotales'] > df['metroscubiertos'])]\n",
    "    notnullapart3 = df[(~ df['metrostotales'].isnull()) & (df['metrostotales'] == df['metroscubiertos'])]\n",
    " \n",
    "    df['habitable'] = False\n",
    "    tipodepropiedades = df.tipodepropiedad.cat.categories.to_list()\n",
    "    for tipodepropiedad in tipodepropiedades: \n",
    "\n",
    "        idsNullMetrosTotales = (df.tipodepropiedad == tipodepropiedad) & (df.metrostotales.isnull())\n",
    "        idsNullMetrosCubiertos = (df.tipodepropiedad == tipodepropiedad) & (df.metroscubiertos.isnull())\n",
    "        banos = df[df.tipodepropiedad == tipodepropiedad].banos.mode(dropna=False);\n",
    "        habitaciones = df[df.tipodepropiedad == tipodepropiedad].habitaciones.mode(dropna=False);\n",
    "\n",
    "        ## Verificamos si la cantidad de registros con metros cubiertos nulos es mayor a 2/5 de los totales. Si es asi \n",
    "        ## los consideramos propiedades no habitables. Y los tratamos de manera diferente\n",
    "        if(len(nullcubiertos[nullcubiertos.tipodepropiedad == tipodepropiedad]) >= 2/5* len(df[df.tipodepropiedad == tipodepropiedad])):\n",
    "            df.metrostotales.fillna(0, inplace=True)\n",
    "            df.metroscubiertos.fillna(0, inplace=True)\n",
    "        else:\n",
    "            df[idsNullMetrosTotales]['metrostotales'] =  df[idsNullMetrosTotales]['metroscubiertos']\n",
    "            df[idsNullMetrosCubiertos]['metroscubiertos'] =  df[idsNullMetrosCubiertos]['metrostotales']\n",
    "\n",
    "        #Si la moda del tipo de propiedad de banos y habitaciones son ambas distintas de nan entonces la propiedad es habitable.\n",
    "\n",
    "        df.loc[(df.tipodepropiedad == tipodepropiedad), 'habitable'] = not(numpy.isnan(banos[0]) and numpy.isnan(habitaciones[0]))\n",
    "    \n",
    "    df['metros'] = df['metrostotales'] + df['metroscubiertos']\n",
    "    if(train):\n",
    "        def is_outlier(group):\n",
    "            Q1 = group.quantile(0.25)\n",
    "            Q3 = group.quantile(0.75)\n",
    "            IQR = Q3 - Q1\n",
    "            precio_min = Q1 - 1.5 * IQR\n",
    "            precio_max = Q3 + 1.5 * IQR\n",
    "            return ~group.between(precio_min, precio_max)\n",
    "        df['precio_mt2'] = df['precio'] / df['metros']\n",
    "        print()\n",
    "        \n",
    "        df = df[~df.groupby('tipodepropiedad')['precio_mt2'].apply(is_outlier).fillna(False)]\n",
    "        print('Despues de filtrar: ', df.shape)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 440,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/ipykernel_launcher.py:32: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/ipykernel_launcher.py:33: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Despues de filtrar:  (223477, 21)\n"
     ]
    }
   ],
   "source": [
    "df_train = pd.read_csv('train.csv',\n",
    "        index_col='id',\n",
    "        dtype={'gimnasio': int,\n",
    "                'usosmultiples': int,\n",
    "                'escuelascercanas': int,\n",
    "                'piscina': int,\n",
    "                'centroscomercialescercanos': int,\n",
    "                'tipodepropiedad': 'category',\n",
    "                'provincia': 'category',\n",
    "                'ciudad': 'category'\n",
    "            },\n",
    "        parse_dates=['fecha'])\n",
    "df_test = pd.read_csv('test.csv',\n",
    "        index_col='id',\n",
    "        dtype={'gimnasio': int,\n",
    "                'usosmultiples': int,\n",
    "                'escuelascercanas': int,\n",
    "                'piscina': int,\n",
    "                'centroscomercialescercanos': int,\n",
    "                'tipodepropiedad': 'category',\n",
    "                'provincia': 'category',\n",
    "                'ciudad': 'category'\n",
    "            },\n",
    "        parse_dates=['fecha'])\n",
    "\n",
    "df_train = normailize_df(df_train, True)\n",
    "df_test = normailize_df(df_test, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 441,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/sklearn/utils/deprecation.py:66: DeprecationWarning: Class Imputer is deprecated; Imputer was deprecated in version 0.20 and will be removed in 0.22. Import impute.SimpleImputer from sklearn instead.\n",
      "  warnings.warn(msg, category=DeprecationWarning)\n"
     ]
    },
    {
     "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>titulo</th>\n",
       "      <th>tipodepropiedad</th>\n",
       "      <th>ciudad</th>\n",
       "      <th>provincia</th>\n",
       "      <th>antiguedad</th>\n",
       "      <th>habitaciones</th>\n",
       "      <th>garages</th>\n",
       "      <th>banos</th>\n",
       "      <th>metroscubiertos</th>\n",
       "      <th>metrostotales</th>\n",
       "      <th>idzona</th>\n",
       "      <th>gimnasio</th>\n",
       "      <th>usosmultiples</th>\n",
       "      <th>piscina</th>\n",
       "      <th>escuelascercanas</th>\n",
       "      <th>centroscomercialescercanos</th>\n",
       "      <th>año</th>\n",
       "      <th>habitable</th>\n",
       "      <th>metros</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [titulo, tipodepropiedad, ciudad, provincia, antiguedad, habitaciones, garages, banos, metroscubiertos, metrostotales, idzona, gimnasio, usosmultiples, piscina, escuelascercanas, centroscomercialescercanos, año, habitable, metros, target]\n",
       "Index: []"
      ]
     },
     "execution_count": 441,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "df_train['metros'] = df_train['metrostotales'] + df_train['metroscubiertos']\n",
    "df_test['metros'] = df_test['metrostotales'] + df_test['metroscubiertos']\n",
    "\n",
    "\n",
    "tipodepropiedades = df_train.tipodepropiedad.cat.categories.to_list()\n",
    "\n",
    "\n",
    "def predict_by_prop(idsTrainFil, idsTestFil): \n",
    "    for tipodepropiedad in tipodepropiedades: \n",
    "        #Solo las propiedades que queremos\n",
    "        idsTrain = (df_train.tipodepropiedad == tipodepropiedad) & idsTrainFil\n",
    "        idsTest = (df_test.tipodepropiedad == tipodepropiedad) & idsTestFil\n",
    "        trainDf = df_train[idsTrain].groupby(['antiguedad']).agg(precio_mt2=('precio_mt2', 'median')).reset_index()\n",
    "\n",
    "        X = trainDf[['antiguedad']]\n",
    "        y = trainDf['precio_mt2']\n",
    "#         print('------', tipodepropiedad)\n",
    "#         print(X.shape, len(y))\n",
    "        if(len(y) == 0 or len(df_test[idsTest][['antiguedad']]) == 0): \n",
    "            continue\n",
    "        if(len(y) == 1):\n",
    "    #         print(        df_test[df_test.tipodepropiedad == tipodepropiedad, 'target'])\n",
    "            df_test.loc[df_test.tipodepropiedad == tipodepropiedad, 'target'] = y.to_list()[0]\n",
    "            continue\n",
    "\n",
    "#         linear_model = Ridge(alpha=0.5).fit(X, y)\n",
    "        linear_model = make_pipeline(PolynomialFeatures(degree=3), Ridge())\n",
    "        linear_model.fit(X, y)\n",
    "\n",
    "        pred = linear_model.predict(df_test[idsTest][['antiguedad']])\n",
    "\n",
    "        df_test.loc[idsTest, 'target'] = pred\n",
    "\n",
    "        \n",
    "\n",
    "imp = Imputer(strategy='most_frequent')\n",
    "df_train['antiguedad'] = imp.fit_transform(df_train[['antiguedad']])        \n",
    "df_test['antiguedad'] = imp.transform(df_test[['antiguedad']])\n",
    "\n",
    "dfAntiguedadMenor10Train = (df_train[\"antiguedad\"] < 10)\n",
    "dfAntiguedadMenor10Test = (df_test[\"antiguedad\"] < 10)\n",
    "\n",
    "dfAntiguedadMayor10Train = (df_train[\"antiguedad\"] >= 10) & (df_train[\"antiguedad\"] < 40)\n",
    "dfAntiguedadMayor10Test = (df_test[\"antiguedad\"] >= 10) & (df_test[\"antiguedad\"] < 40)\n",
    "\n",
    "dfAntiguedadMayor40Train = (df_train[\"antiguedad\"] >= 40)\n",
    "dfAntiguedadMayor40Test = (df_test[\"antiguedad\"] >= 40)\n",
    "\n",
    "predict_by_prop(dfAntiguedadMenor10Train, dfAntiguedadMenor10Test)\n",
    "\n",
    "predict_by_prop(dfAntiguedadMayor10Train, dfAntiguedadMayor10Test)\n",
    "\n",
    "predict_by_prop(dfAntiguedadMayor40Train, dfAntiguedadMayor40Test)\n",
    "\n",
    "\n",
    "df_test.loc[df_test.target.isnull(), 'target'] = df_test['target'].median()\n",
    "df_test['target'] = df_test['target'] *  df_test['metros']\n",
    "final_prediction = df_test.copy().drop(df_test.columns.difference(['target']), 1)\n",
    "# # print()\n",
    "final_prediction.to_csv('prediction.csv')\n",
    "df_test[df_test['target'].isnull()]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 443,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(151879, 21)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.05, 'Regresión del costo m2 según tipo de propiedad en base a la antigüedad \\n (Ultimos 10 años)')"
      ]
     },
     "execution_count": 443,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 713.125x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "habitables = ['Casa', 'Casa en condominio', 'Apartamento', 'Local Comercial']\n",
    "dfAntiguedad = df_train[(~ df_train[\"antiguedad\"].isnull()) & (df_train[\"antiguedad\"] <= 10) & (df_train[\"tipodepropiedad\"].isin(habitables)) ]\n",
    "print(dfAntiguedad.shape)\n",
    "# dfAntiguedad.tipodepropiedad.value_counts()\n",
    "dfAntiguedad['tipodepropiedad'].cat.remove_unused_categories(inplace=True)\n",
    "dfAntiguedad = dfAntiguedad.groupby(['antiguedad', 'tipodepropiedad']).agg(prom_precio_mt2=('precio_mt2', 'mean')).reset_index()\n",
    "\n",
    "\n",
    "ax = sns.lmplot(x=\"antiguedad\", y=\"prom_precio_mt2\", data=dfAntiguedad, hue=\"tipodepropiedad\", height=8, order=1);\n",
    "\n",
    "ax.set(xlabel='Antigüedad', ylabel='Precio m2')\n",
    "\n",
    "new_title = 'Tipo de propiedad'\n",
    "ax._legend.set_title(new_title)\n",
    "ax.fig.suptitle('Regresión del costo m2 según tipo de propiedad en base a la antigüedad \\n (Ultimos 10 años)', y= 1.05)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 442,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/sklearn/utils/deprecation.py:66: DeprecationWarning: Class Imputer is deprecated; Imputer was deprecated in version 0.20 and will be removed in 0.22. Import impute.SimpleImputer from sklearn instead.\n",
      "  warnings.warn(msg, category=DeprecationWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSLE Linear Regression (train): 0.52455\n"
     ]
    }
   ],
   "source": [
    "tipodepropiedades = df_train.tipodepropiedad.cat.categories.to_list()\n",
    "\n",
    "df_test2 = df_train.copy()\n",
    "\n",
    "def predict_by_prop(idsTrainFil, idsTestFil): \n",
    "    for tipodepropiedad in tipodepropiedades: \n",
    "        #Solo las propiedades que queremos\n",
    "        idsTrain = (df_train.tipodepropiedad == tipodepropiedad) & idsTrainFil\n",
    "        idsTest = (df_test2.tipodepropiedad == tipodepropiedad) & idsTestFil\n",
    "        trainDf = df_train[idsTrain].groupby(['antiguedad']).agg(precio_mt2=('precio_mt2', 'median')).reset_index()\n",
    "\n",
    "        X = trainDf[['antiguedad']]\n",
    "        y = trainDf['precio_mt2']\n",
    "#         print('------', tipodepropiedad)\n",
    "#         print(X.shape, len(y))\n",
    "        if(len(y) == 0 or len(df_test2[idsTest][['antiguedad']]) == 0): \n",
    "            continue\n",
    "        if(len(y) == 1):\n",
    "    #         print(        df_test[df_test.tipodepropiedad == tipodepropiedad, 'target'])\n",
    "            df_test2.loc[df_test2.tipodepropiedad == tipodepropiedad, 'target'] = y.to_list()[0]\n",
    "            continue\n",
    "\n",
    "#         linear_model = Ridge(alpha=0.5).fit(X, y)\n",
    "        linear_model = make_pipeline(PolynomialFeatures(degree=3), Ridge())\n",
    "        linear_model.fit(X, y)\n",
    "\n",
    "        pred = linear_model.predict(df_test2[idsTest][['antiguedad']])\n",
    "\n",
    "        df_test2.loc[idsTest, 'target'] = pred\n",
    "\n",
    "\n",
    "imp = Imputer(strategy='most_frequent')\n",
    "df_train['antiguedad'] = imp.fit_transform(df_train[['antiguedad']])        \n",
    "df_test2['antiguedad'] = imp.transform(df_train[['antiguedad']])\n",
    "\n",
    "dfAntiguedadMenor10Train = (df_train[\"antiguedad\"] < 10)\n",
    "dfAntiguedadMenor10Test = (df_test2[\"antiguedad\"] < 10)\n",
    "\n",
    "dfAntiguedadMayor10Train = (df_train[\"antiguedad\"] >= 10) & (df_train[\"antiguedad\"] < 40)\n",
    "dfAntiguedadMayor10Test = (df_test2[\"antiguedad\"] >= 10) & (df_test2[\"antiguedad\"] < 40)\n",
    "\n",
    "dfAntiguedadMayor40Train = (df_train[\"antiguedad\"].isnull())\n",
    "dfAntiguedadMayor40Test = (df_test2[\"antiguedad\"].isnull()) | (df_test2[\"antiguedad\"] >= 40)\n",
    "\n",
    "predict_by_prop(dfAntiguedadMenor10Train, dfAntiguedadMenor10Test)\n",
    "\n",
    "predict_by_prop(dfAntiguedadMayor10Train, dfAntiguedadMayor10Test)\n",
    "\n",
    "predict_by_prop(dfAntiguedadMayor40Train, dfAntiguedadMayor40Test)\n",
    "\n",
    "linear_rmsle_train = RMSLE(df_train['precio_mt2'], df_test2['target'])\n",
    "print(f\"RMSLE Linear Regression (train): {linear_rmsle_train:.5f}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}