KNN 1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.preprocessing import Imputer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Archivos de train y test\n",
    "df_train = pd.read_csv(\"data/train.csv\", index_col=0)\n",
    "df_submit = pd.read_csv(\"data/test.csv\", index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train.tipodepropiedad = df_train.tipodepropiedad.astype('category').cat.codes\n",
    "df_train.ciudad = df_train.ciudad.astype('category').cat.codes\n",
    "df_train.provincia = df_train.provincia.astype('category').cat.codes\n",
    "df_train.fecha = pd.to_datetime(df_train.fecha)\n",
    "df_train.fecha = df_train.fecha.dt.year\n",
    "\n",
    "df_submit.tipodepropiedad = df_submit.tipodepropiedad.astype('category').cat.codes\n",
    "df_submit.ciudad = df_submit.ciudad.astype('category').cat.codes\n",
    "df_submit.provincia = df_submit.provincia.astype('category').cat.codes\n",
    "df_submit.fecha = pd.to_datetime(df_submit.fecha)\n",
    "df_submit.fecha = df_submit.fecha.dt.year\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>nulls</th>\n",
       "      <th>porcentaje</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>lng</th>\n",
       "      <td>123488</td>\n",
       "      <td>51.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lat</th>\n",
       "      <td>123488</td>\n",
       "      <td>51.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>direccion</th>\n",
       "      <td>53072</td>\n",
       "      <td>22.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metrostotales</th>\n",
       "      <td>51467</td>\n",
       "      <td>21.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>antiguedad</th>\n",
       "      <td>43555</td>\n",
       "      <td>18.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>garages</th>\n",
       "      <td>37765</td>\n",
       "      <td>15.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>idzona</th>\n",
       "      <td>28621</td>\n",
       "      <td>11.93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>banos</th>\n",
       "      <td>26221</td>\n",
       "      <td>10.93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>habitaciones</th>\n",
       "      <td>22471</td>\n",
       "      <td>9.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metroscubiertos</th>\n",
       "      <td>17400</td>\n",
       "      <td>7.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>titulo</th>\n",
       "      <td>5387</td>\n",
       "      <td>2.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>descripcion</th>\n",
       "      <td>1619</td>\n",
       "      <td>0.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>centroscomercialescercanos</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fecha</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gimnasio</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>usosmultiples</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>piscina</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>provincia</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ciudad</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>escuelascercanas</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tipodepropiedad</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precio</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             nulls  porcentaje\n",
       "lng                         123488       51.45\n",
       "lat                         123488       51.45\n",
       "direccion                    53072       22.11\n",
       "metrostotales                51467       21.44\n",
       "antiguedad                   43555       18.15\n",
       "garages                      37765       15.74\n",
       "idzona                       28621       11.93\n",
       "banos                        26221       10.93\n",
       "habitaciones                 22471        9.36\n",
       "metroscubiertos              17400        7.25\n",
       "titulo                        5387        2.24\n",
       "descripcion                   1619        0.67\n",
       "centroscomercialescercanos       0        0.00\n",
       "fecha                            0        0.00\n",
       "gimnasio                         0        0.00\n",
       "usosmultiples                    0        0.00\n",
       "piscina                          0        0.00\n",
       "provincia                        0        0.00\n",
       "ciudad                           0        0.00\n",
       "escuelascercanas                 0        0.00\n",
       "tipodepropiedad                  0        0.00\n",
       "precio                           0        0.00"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nulls = pd.DataFrame(df_train.isnull().sum().sort_values(ascending=False), columns=['nulls'])\n",
    "nulls['porcentaje'] = round(100*nulls['nulls'] / len(df_train), 2)\n",
    "nulls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train=df_train.drop(columns=['titulo','descripcion','direccion', 'lat', 'lng'])\n",
    "df_train.antiguedad = df_train.antiguedad.fillna(0)\n",
    "df_train.habitaciones = df_train.habitaciones.fillna(0)\n",
    "df_train.garages = df_train.garages.fillna(0)\n",
    "df_train.banos = df_train.banos.fillna(0)\n",
    "df_train.dropna(subset=['idzona'],inplace=True)\n",
    "df_train['metrostotales'].fillna(df_train['metroscubiertos'], inplace=True)\n",
    "df_train['metroscubiertos'].fillna(0, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>nulls</th>\n",
       "      <th>porcentaje</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>precio</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metroscubiertos</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ciudad</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>provincia</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>antiguedad</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>habitaciones</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>garages</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>banos</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metrostotales</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>centroscomercialescercanos</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>idzona</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fecha</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gimnasio</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>usosmultiples</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>piscina</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>escuelascercanas</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tipodepropiedad</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            nulls  porcentaje\n",
       "precio                          0         0.0\n",
       "metroscubiertos                 0         0.0\n",
       "ciudad                          0         0.0\n",
       "provincia                       0         0.0\n",
       "antiguedad                      0         0.0\n",
       "habitaciones                    0         0.0\n",
       "garages                         0         0.0\n",
       "banos                           0         0.0\n",
       "metrostotales                   0         0.0\n",
       "centroscomercialescercanos      0         0.0\n",
       "idzona                          0         0.0\n",
       "fecha                           0         0.0\n",
       "gimnasio                        0         0.0\n",
       "usosmultiples                   0         0.0\n",
       "piscina                         0         0.0\n",
       "escuelascercanas                0         0.0\n",
       "tipodepropiedad                 0         0.0"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nulls = pd.DataFrame(df_train.isnull().sum().sort_values(ascending=False), columns=['nulls'])\n",
    "nulls['porcentaje'] = round(100*nulls['nulls'] / len(df_train), 2)\n",
    "nulls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df_train.iloc[:, :-1].values\n",
    "y = df_train.precio.values\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.20)\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(X_train)\n",
    "\n",
    "X_train = scaler.transform(X_train)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n"
     ]
    }
   ],
   "source": [
    "error = []\n",
    "calcRange = (2, 10)\n",
    "# Calculating error for K values between 5 and 15\n",
    "for i in range(calcRange[0],calcRange[1]):\n",
    "    print(i)\n",
    "    knn = KNeighborsClassifier(n_neighbors=i)\n",
    "    knn.fit(X_train, y_train)\n",
    "    pred_i = knn.predict(X_test)\n",
    "    error.append(RMSLE(y_test, pred_i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'RMSLE Error')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 6))\n",
    "plt.plot(range(calcRange[0],calcRange[1]), error, color='red', linestyle='dashed', marker='o',\n",
    "         markerfacecolor='blue', markersize=10)\n",
    "plt.title('Error Rate K Value')\n",
    "plt.xlabel('K Value')\n",
    "plt.ylabel('RMSLE Error')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(42276,)\n",
      "(42276, 16)\n",
      "(42276,)\n"
     ]
    }
   ],
   "source": [
    "print(pred_i.shape)\n",
    "print(X_test.shape)\n",
    "print(y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>nulls</th>\n",
       "      <th>porcentaje</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>lng</th>\n",
       "      <td>30695</td>\n",
       "      <td>51.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lat</th>\n",
       "      <td>30695</td>\n",
       "      <td>51.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>direccion</th>\n",
       "      <td>13191</td>\n",
       "      <td>21.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metrostotales</th>\n",
       "      <td>12655</td>\n",
       "      <td>21.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>antiguedad</th>\n",
       "      <td>10714</td>\n",
       "      <td>17.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>garages</th>\n",
       "      <td>9323</td>\n",
       "      <td>15.54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>idzona</th>\n",
       "      <td>7179</td>\n",
       "      <td>11.96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>banos</th>\n",
       "      <td>6554</td>\n",
       "      <td>10.92</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>habitaciones</th>\n",
       "      <td>5628</td>\n",
       "      <td>9.38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metroscubiertos</th>\n",
       "      <td>4299</td>\n",
       "      <td>7.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>titulo</th>\n",
       "      <td>1378</td>\n",
       "      <td>2.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>descripcion</th>\n",
       "      <td>401</td>\n",
       "      <td>0.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>escuelascercanas</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fecha</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gimnasio</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>usosmultiples</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>provincia</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ciudad</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>piscina</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tipodepropiedad</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>centroscomercialescercanos</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            nulls  porcentaje\n",
       "lng                         30695       51.16\n",
       "lat                         30695       51.16\n",
       "direccion                   13191       21.98\n",
       "metrostotales               12655       21.09\n",
       "antiguedad                  10714       17.86\n",
       "garages                      9323       15.54\n",
       "idzona                       7179       11.96\n",
       "banos                        6554       10.92\n",
       "habitaciones                 5628        9.38\n",
       "metroscubiertos              4299        7.16\n",
       "titulo                       1378        2.30\n",
       "descripcion                   401        0.67\n",
       "escuelascercanas                0        0.00\n",
       "fecha                           0        0.00\n",
       "gimnasio                        0        0.00\n",
       "usosmultiples                   0        0.00\n",
       "provincia                       0        0.00\n",
       "ciudad                          0        0.00\n",
       "piscina                         0        0.00\n",
       "tipodepropiedad                 0        0.00\n",
       "centroscomercialescercanos      0        0.00"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nulls = pd.DataFrame(df_submit.isnull().sum().sort_values(ascending=False), columns=['nulls'])\n",
    "nulls['porcentaje'] = round(100*nulls['nulls'] / len(df_submit), 2)\n",
    "nulls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_submit=df_submit.drop(columns=['titulo','descripcion','direccion', 'lat', 'lng'])\n",
    "df_submit.antiguedad = df_submit.antiguedad.fillna(0)\n",
    "df_submit.habitaciones = df_submit.habitaciones.fillna(0)\n",
    "df_submit.garages = df_submit.garages.fillna(0)\n",
    "df_submit.banos = df_submit.banos.fillna(0)\n",
    "#df_submit.dropna(subset=['idzona'],inplace=True)\n",
    "df_submit['metrostotales'].fillna(df_submit['metroscubiertos'], inplace=True)\n",
    "df_submit['metroscubiertos'].fillna(0, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/rama/anaconda3/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"
     ]
    }
   ],
   "source": [
    "imp = Imputer()\n",
    "df_submit['idzona'] = imp.fit_transform(df_submit[['idzona']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>nulls</th>\n",
       "      <th>porcentaje</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>centroscomercialescercanos</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>escuelascercanas</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>piscina</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>usosmultiples</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>gimnasio</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fecha</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>idzona</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metrostotales</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>metroscubiertos</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>banos</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>garages</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>habitaciones</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>antiguedad</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>provincia</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ciudad</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tipodepropiedad</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            nulls  porcentaje\n",
       "centroscomercialescercanos      0         0.0\n",
       "escuelascercanas                0         0.0\n",
       "piscina                         0         0.0\n",
       "usosmultiples                   0         0.0\n",
       "gimnasio                        0         0.0\n",
       "fecha                           0         0.0\n",
       "idzona                          0         0.0\n",
       "metrostotales                   0         0.0\n",
       "metroscubiertos                 0         0.0\n",
       "banos                           0         0.0\n",
       "garages                         0         0.0\n",
       "habitaciones                    0         0.0\n",
       "antiguedad                      0         0.0\n",
       "provincia                       0         0.0\n",
       "ciudad                          0         0.0\n",
       "tipodepropiedad                 0         0.0"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nulls = pd.DataFrame(df_submit.isnull().sum().sort_values(ascending=False), columns=['nulls'])\n",
    "nulls['porcentaje'] = round(100*nulls['nulls'] / len(df_submit), 2)\n",
    "nulls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df_train.iloc[:, :-1].values\n",
    "y = df_train.precio.values\n",
    "X_test = df_submit.iloc[:, :].values\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(X)\n",
    "\n",
    "\n",
    "X = scaler.transform(X)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "knn = KNeighborsClassifier(n_neighbors=2)\n",
    "knn.fit(X, y)\n",
    "pred_KNN = knn.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>target</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4941</th>\n",
       "      <td>6500000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51775</th>\n",
       "      <td>490000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115253</th>\n",
       "      <td>2610000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>299321</th>\n",
       "      <td>1575000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>173570</th>\n",
       "      <td>450000.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           target\n",
       "id               \n",
       "4941    6500000.0\n",
       "51775    490000.0\n",
       "115253  2610000.0\n",
       "299321  1575000.0\n",
       "173570   450000.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "res = pd.DataFrame(pred_KNN, index=df_submit.index, columns=['target'])\n",
    "display(res.head())\n",
    "res.to_csv(\"data/workshop-submission-knn.csv\", header=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df_train.iloc[:, :-1].values\n",
    "y = df_train.precio.values\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)\n",
    "scaler = StandardScaler()\n",
    "scaler.fit(X_train)\n",
    "\n",
    "X_train = scaler.transform(X_train)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.3664579992620725\n"
     ]
    }
   ],
   "source": [
    "pred_i = knn.predict(X_test)\n",
    "print(RMSLE(y_test, pred_i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n"
     ]
    }
   ],
   "source": [
    "error = []\n",
    "calcRange = (2, 10)\n",
    "# Calculating error for K values between 5 and 15\n",
    "for i in range(calcRange[0],calcRange[1]):\n",
    "    print(i)\n",
    "    knn = KNeighborsClassifier(n_neighbors=i)\n",
    "    knn.fit(X_train, y_train)\n",
    "    pred_i = knn.predict(X_test)\n",
    "    error.append(np.mean(y_test != pred_i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Mean Error')"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 6))\n",
    "plt.plot(range(calcRange[0],calcRange[1]), error, color='red', linestyle='dashed', marker='o',\n",
    "         markerfacecolor='blue', markersize=10)\n",
    "plt.title('Error Rate K Value')\n",
    "plt.xlabel('K Value')\n",
    "plt.ylabel('Mean Error')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "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.antiguedad = df.antiguedad.fillna(0) #Asumo que si no tiene antiguedad entonces es nuevo\n",
    "    df.drop([\"direccion\", 'descripcion', 'lat', 'lng', 'fecha', 'ciudad', 'provincia', 'titulo'], 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(np.isnan(banos[0]) and np.isnan(habitaciones[0]))\n",
    "    \n",
    "    df['metros'] = df['metrostotales'] + df['metroscubiertos']\n",
    "    df.habitaciones = df.habitaciones.fillna(0)\n",
    "    df.garages = df.garages.fillna(0)\n",
    "    df.banos = df.banos.fillna(0)\n",
    "    df.idzona = df.idzona.fillna(50003995.0) # IDZona calculado que es la moda\n",
    "\n",
    "    \n",
    "    if(train):\n",
    "\n",
    "        df.dropna(subset=['tipodepropiedad'],inplace=True)\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",
    "        idDel = df_train[df_train.tipodepropiedad == 'Garage'].index\n",
    "        df = df.drop(idDel)\n",
    "        idDel = df_train[df_train.tipodepropiedad == 'Hospedaje'].index\n",
    "        df = df.drop(idDel)\n",
    "        print('Despues de filtrar: ', df.shape)\n",
    "    else:\n",
    "        df.tipodepropiedad = df.tipodepropiedad.fillna('Casa')   #El tipo de propiedad mas comun\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/rama/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:29: 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/indexing.html#indexing-view-versus-copy\n",
      "/home/rama/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:30: 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/indexing.html#indexing-view-versus-copy\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Despues de filtrar:  (223429, 18)\n"
     ]
    }
   ],
   "source": [
    "#Archivos de train y test\n",
    "\n",
    "df_train = pd.read_csv('data/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('data/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)\n",
    "\n",
    "tiposdepropiedades = df_train.tipodepropiedad.cat.categories.to_list()\n",
    "\n",
    "cols = list(df_train.columns)\n",
    "cols =  cols[:13] + cols[14:] +[cols[13]]\n",
    "df_train = df_train[cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n"
     ]
    }
   ],
   "source": [
    "error = []\n",
    "calcRange = (2, 10)\n",
    "# Calculating error for K values between 5 and 15\n",
    "for i in range(calcRange[0],calcRange[1]):\n",
    "    print(i)\n",
    "    knn = KNeighborsClassifier(n_neighbors=i)\n",
    "    knn.fit(X_train, y_train)\n",
    "    pred_i = knn.predict(X_test)\n",
    "    error.append(np.mean(y_test != pred_i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Mean Error')"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 6))\n",
    "plt.plot(range(calcRange[0],calcRange[1]), error, color='red', linestyle='dashed', marker='o',\n",
    "         markerfacecolor='blue', markersize=10)\n",
    "plt.title('Error Rate K Value')\n",
    "plt.xlabel('K Value')\n",
    "plt.ylabel('Mean Error')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n"
     ]
    }
   ],
   "source": [
    "error = []\n",
    "calcRange = (2, 10)\n",
    "# Calculating error for K values between 5 and 15\n",
    "for i in range(calcRange[0],calcRange[1]):\n",
    "    print(i)\n",
    "    knn = KNeighborsClassifier(n_neighbors=i)\n",
    "    knn.fit(X_train, y_train)\n",
    "    pred_i = knn.predict(X_test)\n",
    "    error.append(RMSLE(y_test,pred_i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'RMSLE Error')"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 6))\n",
    "plt.plot(range(calcRange[0],calcRange[1]), error, color='red', linestyle='dashed', marker='o',\n",
    "         markerfacecolor='blue', markersize=10)\n",
    "plt.title('Error Rate K Value')\n",
    "plt.xlabel('K Value')\n",
    "plt.ylabel('RMSLE Error')"
   ]
  },
  {
   "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
}