ML-para-Data-Science/Notebooks/Clase4.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "TratDatos_pandas.ipynb",
      "provenance": [],
      "toc_visible": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7OI7tLnm2esU"
      },
      "source": [
        "# Tratamiento de datos con Pandas 🐼"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ytsICs7T3bmO"
      },
      "source": [
        "## 00.Carga de datos (BigQuery Public Dataset program)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5mdxXlSiEplk"
      },
      "source": [
        "Cargamos los datos del [BigQuery Public Dataset Program](https://console.cloud.google.com/marketplace/product/gencat/cell_coverage)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k12wJvU15pqL",
        "outputId": "6189a339-6209-473f-f8fb-5bae6c6f0356"
      },
      "source": [
        "from google.colab import auth\n",
        "auth.authenticate_user()    #autenticar google cloud\n",
        "print('Autenticacion exitosa')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Autenticacion exitosa\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gxZCtc2gFAte"
      },
      "source": [
        "Cargamos las librerias necesarias: bigquery, pandas y numpy"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "d063uwR_5NDj"
      },
      "source": [
        "from google.cloud import bigquery\n",
        "import pandas as pd\n",
        "import numpy as np"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uOuhj2tcH-sc"
      },
      "source": [
        "Leemos los datos de BigQuery y los dejamos en un Dataframe de pandas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 529
        },
        "id": "azsk7Y352kbc",
        "outputId": "735f53e3-968b-4447-c2b1-ffeda03dd19b"
      },
      "source": [
        "project_reader = \"clase4-313909\" #nombre de proyecto en Google Cloud\n",
        "project_public = \"bigquery-public-data\" #datasets públicos de BigQuery\n",
        "dataset_id = \"catalonian_mobile_coverage_eu\" #nombre del dataset\n",
        "table_id = \"mobile_data_2015_2017\"\n",
        "bq_client = bigquery.Client(project_reader)\n",
        "dataset_ref = bigquery.DatasetReference(project_public, dataset_id)\n",
        "table_ref = dataset_ref.table(table_id)\n",
        "table = bq_client.get_table(table_ref)\n",
        "df = bq_client.list_rows(table,max_results=100000).to_dataframe() #primeros 100,000 registros\n",
        "df.head(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>date</th>\n",
              "      <th>hour</th>\n",
              "      <th>lat</th>\n",
              "      <th>long</th>\n",
              "      <th>signal</th>\n",
              "      <th>network</th>\n",
              "      <th>operator</th>\n",
              "      <th>status</th>\n",
              "      <th>description</th>\n",
              "      <th>net</th>\n",
              "      <th>speed</th>\n",
              "      <th>satellites</th>\n",
              "      <th>precission</th>\n",
              "      <th>provider</th>\n",
              "      <th>activity</th>\n",
              "      <th>downloadSpeed</th>\n",
              "      <th>uploadSpeed</th>\n",
              "      <th>postal_code</th>\n",
              "      <th>town_name</th>\n",
              "      <th>position_geom</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>21:01:29</td>\n",
              "      <td>41.21993</td>\n",
              "      <td>1.75448</td>\n",
              "      <td>0</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>82.2</td>\n",
              "      <td>2.0</td>\n",
              "      <td>35.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(1.75448 41.21993)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>06:31:35</td>\n",
              "      <td>41.24392</td>\n",
              "      <td>1.87257</td>\n",
              "      <td>3</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>73.1</td>\n",
              "      <td>1.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(1.87257 41.24392)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>15:15:19</td>\n",
              "      <td>41.27102</td>\n",
              "      <td>2.02823</td>\n",
              "      <td>10</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>100.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>26.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>080898</td>\n",
              "      <td>Gavà</td>\n",
              "      <td>POINT(2.02823 41.27102)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>06:53:59</td>\n",
              "      <td>41.33530</td>\n",
              "      <td>2.10246</td>\n",
              "      <td>4</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>34.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>081691</td>\n",
              "      <td>el Prat de Llobregat</td>\n",
              "      <td>POINT(2.10246 41.3353)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>17:56:01</td>\n",
              "      <td>41.23979</td>\n",
              "      <td>1.86027</td>\n",
              "      <td>7</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>74.6</td>\n",
              "      <td>3.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>082704</td>\n",
              "      <td>Sitges</td>\n",
              "      <td>POINT(1.86027 41.23979)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>20:55:41</td>\n",
              "      <td>41.23733</td>\n",
              "      <td>1.82107</td>\n",
              "      <td>0</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>90.5</td>\n",
              "      <td>2.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>082704</td>\n",
              "      <td>Sitges</td>\n",
              "      <td>POINT(1.82107 41.23733)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>17:57:05</td>\n",
              "      <td>41.24427</td>\n",
              "      <td>1.87139</td>\n",
              "      <td>0</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>59.4</td>\n",
              "      <td>0.0</td>\n",
              "      <td>43.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>082704</td>\n",
              "      <td>Sitges</td>\n",
              "      <td>POINT(1.87139 41.24427)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>20:52:58</td>\n",
              "      <td>41.24342</td>\n",
              "      <td>1.86916</td>\n",
              "      <td>5</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>60.9</td>\n",
              "      <td>0.0</td>\n",
              "      <td>12.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>082704</td>\n",
              "      <td>Sitges</td>\n",
              "      <td>POINT(1.86916 41.24342)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>16:44:33</td>\n",
              "      <td>41.43556</td>\n",
              "      <td>2.19186</td>\n",
              "      <td>17</td>\n",
              "      <td>orange</td>\n",
              "      <td>Orange</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>2G</td>\n",
              "      <td>14.3</td>\n",
              "      <td>0.0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>STILL</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>080193</td>\n",
              "      <td>Barcelona</td>\n",
              "      <td>POINT(2.19186 41.43556)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>16:47:13</td>\n",
              "      <td>41.44103</td>\n",
              "      <td>2.19065</td>\n",
              "      <td>17</td>\n",
              "      <td>orange</td>\n",
              "      <td>Orange</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>2G</td>\n",
              "      <td>13.8</td>\n",
              "      <td>3.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>STILL</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>080193</td>\n",
              "      <td>Barcelona</td>\n",
              "      <td>POINT(2.19065 41.44103)</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         date      hour  ...             town_name            position_geom\n",
              "0  2016-02-22  21:01:29  ...                  None  POINT(1.75448 41.21993)\n",
              "1  2016-02-22  06:31:35  ...                  None  POINT(1.87257 41.24392)\n",
              "2  2016-02-22  15:15:19  ...                  Gavà  POINT(2.02823 41.27102)\n",
              "3  2016-02-22  06:53:59  ...  el Prat de Llobregat   POINT(2.10246 41.3353)\n",
              "4  2016-02-22  17:56:01  ...                Sitges  POINT(1.86027 41.23979)\n",
              "5  2016-02-22  20:55:41  ...                Sitges  POINT(1.82107 41.23733)\n",
              "6  2016-02-22  17:57:05  ...                Sitges  POINT(1.87139 41.24427)\n",
              "7  2016-02-22  20:52:58  ...                Sitges  POINT(1.86916 41.24342)\n",
              "8  2016-02-22  16:44:33  ...             Barcelona  POINT(2.19186 41.43556)\n",
              "9  2016-02-22  16:47:13  ...             Barcelona  POINT(2.19065 41.44103)\n",
              "\n",
              "[10 rows x 20 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nNoAFJYRFY-b"
      },
      "source": [
        "Datos estadisticos del dataframe\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "-jmW6V5h8fiV",
        "outputId": "65ac9288-ca26-42bb-952a-46576b8abd46"
      },
      "source": [
        "df.describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>lat</th>\n",
              "      <th>long</th>\n",
              "      <th>signal</th>\n",
              "      <th>status</th>\n",
              "      <th>speed</th>\n",
              "      <th>satellites</th>\n",
              "      <th>precission</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>100000.000000</td>\n",
              "      <td>100000.000000</td>\n",
              "      <td>100000.00000</td>\n",
              "      <td>100000.000000</td>\n",
              "      <td>100000.000000</td>\n",
              "      <td>100000.000000</td>\n",
              "      <td>100000.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>41.157284</td>\n",
              "      <td>0.682763</td>\n",
              "      <td>13.51655</td>\n",
              "      <td>1.775490</td>\n",
              "      <td>24.522199</td>\n",
              "      <td>4.870640</td>\n",
              "      <td>58.088670</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>3.392317</td>\n",
              "      <td>9.852919</td>\n",
              "      <td>7.48710</td>\n",
              "      <td>0.636325</td>\n",
              "      <td>33.713067</td>\n",
              "      <td>3.826211</td>\n",
              "      <td>303.237498</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>-9.405610</td>\n",
              "      <td>-115.151690</td>\n",
              "      <td>0.00000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>41.385160</td>\n",
              "      <td>1.803968</td>\n",
              "      <td>8.00000</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>1.300000</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>10.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>41.491140</td>\n",
              "      <td>2.131445</td>\n",
              "      <td>13.00000</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>6.700000</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>17.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>41.689563</td>\n",
              "      <td>2.236403</td>\n",
              "      <td>18.00000</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>35.900000</td>\n",
              "      <td>7.000000</td>\n",
              "      <td>30.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>48.354450</td>\n",
              "      <td>121.578020</td>\n",
              "      <td>99.00000</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>252.800000</td>\n",
              "      <td>24.000000</td>\n",
              "      <td>2753.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                 lat           long  ...     satellites     precission\n",
              "count  100000.000000  100000.000000  ...  100000.000000  100000.000000\n",
              "mean       41.157284       0.682763  ...       4.870640      58.088670\n",
              "std         3.392317       9.852919  ...       3.826211     303.237498\n",
              "min        -9.405610    -115.151690  ...       0.000000       0.000000\n",
              "25%        41.385160       1.803968  ...       2.000000      10.000000\n",
              "50%        41.491140       2.131445  ...       4.000000      17.000000\n",
              "75%        41.689563       2.236403  ...       7.000000      30.000000\n",
              "max        48.354450     121.578020  ...      24.000000    2753.000000\n",
              "\n",
              "[8 rows x 7 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Av9cmzNM8tjL"
      },
      "source": [
        "##  01.Deteccion de datos nulos e imputacion de **valores**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kn-so66TDAr3"
      },
      "source": [
        "Vamos a buscar cuantos campos tienen algun valor nulo (NaN)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VONimV-T82JB",
        "outputId": "47991605-daa9-4746-b173-2143f5d27e3d"
      },
      "source": [
        "df.isnull().sum()  #cantidad de nulos por columna"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "date                  0\n",
              "hour                  0\n",
              "lat                   0\n",
              "long                  0\n",
              "signal                0\n",
              "network             338\n",
              "operator            338\n",
              "status                0\n",
              "description           0\n",
              "net                1979\n",
              "speed                 0\n",
              "satellites            0\n",
              "precission            0\n",
              "provider              0\n",
              "activity              0\n",
              "downloadSpeed    100000\n",
              "uploadSpeed      100000\n",
              "postal_code        8467\n",
              "town_name          8467\n",
              "position_geom         0\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_2WVX7X-DGAx"
      },
      "source": [
        "Seleccionamos las filas concretas que tienen los valores NaN, por ejemplo network"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JkNDKMwv9JGi",
        "outputId": "883dfdd0-5676-4403-9aa8-9afeff1b53c3"
      },
      "source": [
        "fila_con_nan = df.isnull()['network'] #indices de las filas con null\n",
        "filas_con_nan = df[fila_con_nan] #filas con null\n",
        "print(filas_con_nan)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "             date      hour  ...               town_name            position_geom\n",
            "42     2016-02-22  20:04:29  ...  Esplugues de Llobregat  POINT(2.08614 41.37022)\n",
            "43     2016-02-22  20:06:29  ...  Esplugues de Llobregat  POINT(2.08539 41.37012)\n",
            "1092   2016-02-22  07:41:56  ...             Bellaguarda  POINT(0.74629 41.34041)\n",
            "15487  2015-12-05  08:45:50  ...       Vilallonga de Ter  POINT(2.31248 42.33092)\n",
            "15488  2015-12-05  17:10:30  ...       Vilallonga de Ter  POINT(2.31239 42.33096)\n",
            "...           ...       ...  ...                     ...                      ...\n",
            "92982  2015-07-20  12:40:39  ...                Cubelles  POINT(1.67821 41.20151)\n",
            "92983  2015-07-20  12:41:36  ...                Cubelles  POINT(1.67897 41.20102)\n",
            "92984  2015-07-20  12:39:36  ...                Cubelles  POINT(1.67782 41.20166)\n",
            "92985  2015-07-20  09:45:10  ...                Cubelles  POINT(1.67609 41.20255)\n",
            "92986  2015-07-20  12:49:29  ...                Cubelles  POINT(1.67885 41.20084)\n",
            "\n",
            "[338 rows x 20 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n05Gzxd8Dmf2"
      },
      "source": [
        "¿Que valor introducimos? Podemos poner el valor mas frecuente, la media, un cero ....\n",
        "Para net vamos a poner por ejemplo el valor mas repetido, para ello pintamos un histograma"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "background_save": true,
          "base_uri": "https://localhost:8080/",
          "height": 699
        },
        "id": "n2_qv2_UDb9h",
        "outputId": "7b664dbc-53e5-4ef4-affe-800519ab6d8b"
      },
      "source": [
        "hist = df['network'].hist(xlabelsize=7, xrot=90, figsize=(20,10)) #ocurrencia de cada red"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9VND_4rAFwsf"
      },
      "source": [
        "df['network'].fillna('yoigo', inplace=True) #reemplazar null"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tMrPPREcFnyd"
      },
      "source": [
        "Por ejemplo si fuese una variable continua, podemos imputar la media en los valores nulos con `df.fillna(df.mean())`"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ne0x-hpHF6G6",
        "outputId": "28587c36-7f60-47d9-e4e8-5a1b080375e5"
      },
      "source": [
        "df.isnull().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "date                  0\n",
              "hour                  0\n",
              "lat                   0\n",
              "long                  0\n",
              "signal                0\n",
              "network               0\n",
              "operator            338\n",
              "status                0\n",
              "description           0\n",
              "net                1979\n",
              "speed                 0\n",
              "satellites            0\n",
              "precission            0\n",
              "provider              0\n",
              "activity              0\n",
              "downloadSpeed    100000\n",
              "uploadSpeed      100000\n",
              "postal_code        8467\n",
              "town_name          8467\n",
              "position_geom         0\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZIDkMlfYCuNG"
      },
      "source": [
        "## 02.Reformateo de datos"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uJ7BdbV-F1nz"
      },
      "source": [
        "Cambiar el formato de las fechas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "geIuNWY8GyyO"
      },
      "source": [
        "columna_fecha = pd.to_datetime(df['date'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7QN9Jf1LHKW-",
        "outputId": "dd462849-f05b-48ad-88c4-4d34400609e9"
      },
      "source": [
        "columna_fecha"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0       2016-02-22\n",
              "1       2016-02-22\n",
              "2       2016-02-22\n",
              "3       2016-02-22\n",
              "4       2016-02-22\n",
              "           ...    \n",
              "99995   2015-07-20\n",
              "99996   2015-07-20\n",
              "99997   2015-07-20\n",
              "99998   2015-07-20\n",
              "99999   2015-07-20\n",
              "Name: date, Length: 100000, dtype: datetime64[ns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HenW59FhF906"
      },
      "source": [
        "Utilizamos la funcion `dt.strftime` para definir una nueva mascara de fechas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lLQySdG7GaiC"
      },
      "source": [
        "cambio_fecha_df = columna_fecha.dt.strftime(\"%m/%d/%y\")     ########buscar más mascaras"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fR-qeB2sHP-X",
        "outputId": "4a67de8c-ea89-43f0-94e7-44c892bd3518"
      },
      "source": [
        "cambio_fecha_df"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0        02/22/16\n",
              "1        02/22/16\n",
              "2        02/22/16\n",
              "3        02/22/16\n",
              "4        02/22/16\n",
              "           ...   \n",
              "99995    07/20/15\n",
              "99996    07/20/15\n",
              "99997    07/20/15\n",
              "99998    07/20/15\n",
              "99999    07/20/15\n",
              "Name: date, Length: 100000, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 41
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8BH67VLTGFED"
      },
      "source": [
        "Otro clasico es cambiar los . por , o viceversa como separador de cifras"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UGeZlu-QHcAj"
      },
      "source": [
        "columna_lat = df['lat'].astype(str).str.replace('.', ',') #cambiar punto por coma"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "56WrydqKH1yl",
        "outputId": "01c5504c-293e-4863-84f3-f2b1e4b75297"
      },
      "source": [
        "columna_lat"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0        41,21993\n",
              "1        41,24392\n",
              "2        41,27102\n",
              "3         41,3353\n",
              "4        41,23979\n",
              "           ...   \n",
              "99995     41,3872\n",
              "99996    41,38741\n",
              "99997    41,38041\n",
              "99998    41,38468\n",
              "99999     41,3777\n",
              "Name: lat, Length: 100000, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JDJrQ0LjCx0B"
      },
      "source": [
        "## 03.Normalizacion, centrado y escalado"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 588
        },
        "id": "Blk3OLvZI5bd",
        "outputId": "b46b645c-1e25-4fb6-df0a-fdcb6535c19e"
      },
      "source": [
        "df"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>date</th>\n",
              "      <th>hour</th>\n",
              "      <th>lat</th>\n",
              "      <th>long</th>\n",
              "      <th>signal</th>\n",
              "      <th>network</th>\n",
              "      <th>operator</th>\n",
              "      <th>status</th>\n",
              "      <th>description</th>\n",
              "      <th>net</th>\n",
              "      <th>speed</th>\n",
              "      <th>satellites</th>\n",
              "      <th>precission</th>\n",
              "      <th>provider</th>\n",
              "      <th>activity</th>\n",
              "      <th>downloadSpeed</th>\n",
              "      <th>uploadSpeed</th>\n",
              "      <th>postal_code</th>\n",
              "      <th>town_name</th>\n",
              "      <th>position_geom</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>21:01:29</td>\n",
              "      <td>41.21993</td>\n",
              "      <td>1.75448</td>\n",
              "      <td>0</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>82.2</td>\n",
              "      <td>2.0</td>\n",
              "      <td>35.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(1.75448 41.21993)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>06:31:35</td>\n",
              "      <td>41.24392</td>\n",
              "      <td>1.87257</td>\n",
              "      <td>3</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>73.1</td>\n",
              "      <td>1.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(1.87257 41.24392)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>15:15:19</td>\n",
              "      <td>41.27102</td>\n",
              "      <td>2.02823</td>\n",
              "      <td>10</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>100.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>26.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>080898</td>\n",
              "      <td>Gavà</td>\n",
              "      <td>POINT(2.02823 41.27102)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>06:53:59</td>\n",
              "      <td>41.33530</td>\n",
              "      <td>2.10246</td>\n",
              "      <td>4</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>34.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>081691</td>\n",
              "      <td>el Prat de Llobregat</td>\n",
              "      <td>POINT(2.10246 41.3353)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>17:56:01</td>\n",
              "      <td>41.23979</td>\n",
              "      <td>1.86027</td>\n",
              "      <td>7</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>74.6</td>\n",
              "      <td>3.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>082704</td>\n",
              "      <td>Sitges</td>\n",
              "      <td>POINT(1.86027 41.23979)</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",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99995</th>\n",
              "      <td>2015-07-20</td>\n",
              "      <td>18:52:37</td>\n",
              "      <td>41.38720</td>\n",
              "      <td>2.13133</td>\n",
              "      <td>6</td>\n",
              "      <td>vodafone</td>\n",
              "      <td>Vodafone ES</td>\n",
              "      <td>2</td>\n",
              "      <td>STATE_EMERGENCY_ONLY</td>\n",
              "      <td>4G</td>\n",
              "      <td>3.6</td>\n",
              "      <td>5.0</td>\n",
              "      <td>48.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>ON_FOOT</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>080193</td>\n",
              "      <td>Barcelona</td>\n",
              "      <td>POINT(2.13133 41.3872)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99996</th>\n",
              "      <td>2015-07-20</td>\n",
              "      <td>18:51:14</td>\n",
              "      <td>41.38741</td>\n",
              "      <td>2.13038</td>\n",
              "      <td>6</td>\n",
              "      <td>vodafone</td>\n",
              "      <td>Vodafone ES</td>\n",
              "      <td>2</td>\n",
              "      <td>STATE_EMERGENCY_ONLY</td>\n",
              "      <td>4G</td>\n",
              "      <td>4.6</td>\n",
              "      <td>11.0</td>\n",
              "      <td>64.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>ON_FOOT</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>080193</td>\n",
              "      <td>Barcelona</td>\n",
              "      <td>POINT(2.13038 41.38741)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99997</th>\n",
              "      <td>2015-07-20</td>\n",
              "      <td>19:57:42</td>\n",
              "      <td>41.38041</td>\n",
              "      <td>2.16373</td>\n",
              "      <td>18</td>\n",
              "      <td>vodafone</td>\n",
              "      <td>Vodafone ES</td>\n",
              "      <td>2</td>\n",
              "      <td>STATE_EMERGENCY_ONLY</td>\n",
              "      <td>4G</td>\n",
              "      <td>1.2</td>\n",
              "      <td>3.0</td>\n",
              "      <td>89.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>ON_FOOT</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>080193</td>\n",
              "      <td>Barcelona</td>\n",
              "      <td>POINT(2.16373 41.38041)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99998</th>\n",
              "      <td>2015-07-20</td>\n",
              "      <td>21:17:46</td>\n",
              "      <td>41.38468</td>\n",
              "      <td>2.16385</td>\n",
              "      <td>17</td>\n",
              "      <td>vodafone</td>\n",
              "      <td>Vodafone ES</td>\n",
              "      <td>2</td>\n",
              "      <td>STATE_EMERGENCY_ONLY</td>\n",
              "      <td>4G</td>\n",
              "      <td>3.6</td>\n",
              "      <td>7.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>ON_FOOT</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>080193</td>\n",
              "      <td>Barcelona</td>\n",
              "      <td>POINT(2.16385 41.38468)</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99999</th>\n",
              "      <td>2015-07-20</td>\n",
              "      <td>20:29:18</td>\n",
              "      <td>41.37770</td>\n",
              "      <td>2.16381</td>\n",
              "      <td>25</td>\n",
              "      <td>vodafone</td>\n",
              "      <td>Vodafone ES</td>\n",
              "      <td>2</td>\n",
              "      <td>STATE_EMERGENCY_ONLY</td>\n",
              "      <td>4G</td>\n",
              "      <td>3.6</td>\n",
              "      <td>5.0</td>\n",
              "      <td>54.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>ON_FOOT</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>080193</td>\n",
              "      <td>Barcelona</td>\n",
              "      <td>POINT(2.16381 41.3777)</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>100000 rows × 20 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "             date      hour  ...             town_name            position_geom\n",
              "0      2016-02-22  21:01:29  ...                  None  POINT(1.75448 41.21993)\n",
              "1      2016-02-22  06:31:35  ...                  None  POINT(1.87257 41.24392)\n",
              "2      2016-02-22  15:15:19  ...                  Gavà  POINT(2.02823 41.27102)\n",
              "3      2016-02-22  06:53:59  ...  el Prat de Llobregat   POINT(2.10246 41.3353)\n",
              "4      2016-02-22  17:56:01  ...                Sitges  POINT(1.86027 41.23979)\n",
              "...           ...       ...  ...                   ...                      ...\n",
              "99995  2015-07-20  18:52:37  ...             Barcelona   POINT(2.13133 41.3872)\n",
              "99996  2015-07-20  18:51:14  ...             Barcelona  POINT(2.13038 41.38741)\n",
              "99997  2015-07-20  19:57:42  ...             Barcelona  POINT(2.16373 41.38041)\n",
              "99998  2015-07-20  21:17:46  ...             Barcelona  POINT(2.16385 41.38468)\n",
              "99999  2015-07-20  20:29:18  ...             Barcelona   POINT(2.16381 41.3777)\n",
              "\n",
              "[100000 rows x 20 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ej_fSZNhGQu4"
      },
      "source": [
        "Vamos a trabajar con las columnas speed, satellites y precision, aqui vamos a utilizar una funcion de sklearn para hacer este preprocesado. Hay muchas formas de escalar y estandarizar valores, en este caso vamos a utilizar una tecnica para dejar los valores entre -1 y 1. El framework sklearn tiene una buena [paleta de funciones](https://https://scikit-learn.org/stable/modules/preprocessing.html) para realizar preparacion de datos"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "do8-FVsMSYj1",
        "outputId": "699776fd-2bbd-4351-b595-02214ed9a2fb"
      },
      "source": [
        "from sklearn import preprocessing\n",
        "\n",
        "columna_speed = df['speed'].values #Devuelve numpy\n",
        "min_max_scaler = preprocessing.MinMaxScaler()\n",
        "columna_speed_scaled = min_max_scaler.fit_transform(columna_speed.reshape(-1, 1)) #escalar datos de -1 a 1\n",
        "columna_speed_scaled"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.32515823],\n",
              "       [0.28916139],\n",
              "       [0.39556962],\n",
              "       ...,\n",
              "       [0.00474684],\n",
              "       [0.01424051],\n",
              "       [0.01424051]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WPUMdnNBC1yc"
      },
      "source": [
        "## 04.Integracion de datos"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UO_dGwRuHUWT"
      },
      "source": [
        "Vamos a genera un dataframe dummy para realizar un JOIN entre ellos, por ejemplo vamos a imaginar que tenemos datos con informacion (columnas A,B,C,D con datos aleatorios) de las compañias telefonicas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PopNvqzmHph9"
      },
      "source": [
        "df_dummy = pd.DataFrame(np.random.randint(0,1000,size=(100, 4)), columns=list('ABCD'))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oyKYjg2DJ2oQ"
      },
      "source": [
        "df_dummy['network'] = df['network'].sample(n=100).tolist()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 399
        },
        "id": "05m0W0m0IeuQ",
        "outputId": "572418c0-87b9-4180-c1ef-2e81bfdb3b83"
      },
      "source": [
        "df_dummy"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>A</th>\n",
              "      <th>B</th>\n",
              "      <th>C</th>\n",
              "      <th>D</th>\n",
              "      <th>network</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>179</td>\n",
              "      <td>642</td>\n",
              "      <td>263</td>\n",
              "      <td>435</td>\n",
              "      <td>movistar</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>667</td>\n",
              "      <td>134</td>\n",
              "      <td>184</td>\n",
              "      <td>693</td>\n",
              "      <td>movistar</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>391</td>\n",
              "      <td>224</td>\n",
              "      <td>157</td>\n",
              "      <td>188</td>\n",
              "      <td>movistar</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>435</td>\n",
              "      <td>621</td>\n",
              "      <td>513</td>\n",
              "      <td>349</td>\n",
              "      <td>movistar</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>711</td>\n",
              "      <td>312</td>\n",
              "      <td>110</td>\n",
              "      <td>791</td>\n",
              "      <td>orange</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>95</th>\n",
              "      <td>284</td>\n",
              "      <td>739</td>\n",
              "      <td>999</td>\n",
              "      <td>362</td>\n",
              "      <td>vodafone</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>96</th>\n",
              "      <td>378</td>\n",
              "      <td>323</td>\n",
              "      <td>976</td>\n",
              "      <td>740</td>\n",
              "      <td>movistar</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>97</th>\n",
              "      <td>509</td>\n",
              "      <td>549</td>\n",
              "      <td>508</td>\n",
              "      <td>805</td>\n",
              "      <td>movistar</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>98</th>\n",
              "      <td>15</td>\n",
              "      <td>518</td>\n",
              "      <td>743</td>\n",
              "      <td>358</td>\n",
              "      <td>vodafone</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99</th>\n",
              "      <td>955</td>\n",
              "      <td>787</td>\n",
              "      <td>638</td>\n",
              "      <td>426</td>\n",
              "      <td>movistar</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>100 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      A    B    C    D   network\n",
              "0   179  642  263  435  movistar\n",
              "1   667  134  184  693  movistar\n",
              "2   391  224  157  188  movistar\n",
              "3   435  621  513  349  movistar\n",
              "4   711  312  110  791    orange\n",
              "..  ...  ...  ...  ...       ...\n",
              "95  284  739  999  362  vodafone\n",
              "96  378  323  976  740  movistar\n",
              "97  509  549  508  805  movistar\n",
              "98   15  518  743  358  vodafone\n",
              "99  955  787  638  426  movistar\n",
              "\n",
              "[100 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DeNeCK7fMWKu"
      },
      "source": [
        "Por ultimo, podemos realizar JOINs al estilo SQL con la funcion merge de pandas, en este caso vamos a realizar por ejemplo un INNER JOIN entre los dataframes df y df_dummy"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u9pCkBqcL_3Z"
      },
      "source": [
        "df_joined = pd.merge(df, df_dummy, on='network', how='inner')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 588
        },
        "id": "CONJl1q1MIlx",
        "outputId": "122a0d07-d951-4223-fccd-a987b5c8f2d7"
      },
      "source": [
        "df_joined"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>date</th>\n",
              "      <th>hour</th>\n",
              "      <th>lat</th>\n",
              "      <th>long</th>\n",
              "      <th>signal</th>\n",
              "      <th>network</th>\n",
              "      <th>operator</th>\n",
              "      <th>status</th>\n",
              "      <th>description</th>\n",
              "      <th>net</th>\n",
              "      <th>speed</th>\n",
              "      <th>satellites</th>\n",
              "      <th>precission</th>\n",
              "      <th>provider</th>\n",
              "      <th>activity</th>\n",
              "      <th>downloadSpeed</th>\n",
              "      <th>uploadSpeed</th>\n",
              "      <th>postal_code</th>\n",
              "      <th>town_name</th>\n",
              "      <th>position_geom</th>\n",
              "      <th>A</th>\n",
              "      <th>B</th>\n",
              "      <th>C</th>\n",
              "      <th>D</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>21:01:29</td>\n",
              "      <td>41.21993</td>\n",
              "      <td>1.75448</td>\n",
              "      <td>0</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>82.2</td>\n",
              "      <td>2.0</td>\n",
              "      <td>35.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(1.75448 41.21993)</td>\n",
              "      <td>425</td>\n",
              "      <td>915</td>\n",
              "      <td>796</td>\n",
              "      <td>164</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>21:01:29</td>\n",
              "      <td>41.21993</td>\n",
              "      <td>1.75448</td>\n",
              "      <td>0</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>82.2</td>\n",
              "      <td>2.0</td>\n",
              "      <td>35.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(1.75448 41.21993)</td>\n",
              "      <td>687</td>\n",
              "      <td>905</td>\n",
              "      <td>296</td>\n",
              "      <td>491</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>06:31:35</td>\n",
              "      <td>41.24392</td>\n",
              "      <td>1.87257</td>\n",
              "      <td>3</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>73.1</td>\n",
              "      <td>1.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(1.87257 41.24392)</td>\n",
              "      <td>425</td>\n",
              "      <td>915</td>\n",
              "      <td>796</td>\n",
              "      <td>164</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>06:31:35</td>\n",
              "      <td>41.24392</td>\n",
              "      <td>1.87257</td>\n",
              "      <td>3</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>73.1</td>\n",
              "      <td>1.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(1.87257 41.24392)</td>\n",
              "      <td>687</td>\n",
              "      <td>905</td>\n",
              "      <td>296</td>\n",
              "      <td>491</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2016-02-22</td>\n",
              "      <td>15:15:19</td>\n",
              "      <td>41.27102</td>\n",
              "      <td>2.02823</td>\n",
              "      <td>10</td>\n",
              "      <td>yoigo</td>\n",
              "      <td>YOIGO</td>\n",
              "      <td>3</td>\n",
              "      <td>STATE_POWER_OFF</td>\n",
              "      <td>None</td>\n",
              "      <td>100.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>26.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>080898</td>\n",
              "      <td>Gavà</td>\n",
              "      <td>POINT(2.02823 41.27102)</td>\n",
              "      <td>425</td>\n",
              "      <td>915</td>\n",
              "      <td>796</td>\n",
              "      <td>164</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",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3052119</th>\n",
              "      <td>2015-07-20</td>\n",
              "      <td>12:22:07</td>\n",
              "      <td>26.66852</td>\n",
              "      <td>-81.93980</td>\n",
              "      <td>30</td>\n",
              "      <td>MetroPCS</td>\n",
              "      <td>MetroPCS</td>\n",
              "      <td>2</td>\n",
              "      <td>STATE_EMERGENCY_ONLY</td>\n",
              "      <td>4G</td>\n",
              "      <td>43.3</td>\n",
              "      <td>8.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(-81.9398 26.66852)</td>\n",
              "      <td>337</td>\n",
              "      <td>34</td>\n",
              "      <td>154</td>\n",
              "      <td>38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3052120</th>\n",
              "      <td>2015-07-20</td>\n",
              "      <td>12:11:12</td>\n",
              "      <td>26.65641</td>\n",
              "      <td>-81.84830</td>\n",
              "      <td>24</td>\n",
              "      <td>MetroPCS</td>\n",
              "      <td>MetroPCS</td>\n",
              "      <td>2</td>\n",
              "      <td>STATE_EMERGENCY_ONLY</td>\n",
              "      <td>4G</td>\n",
              "      <td>62.8</td>\n",
              "      <td>8.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(-81.8483 26.65641)</td>\n",
              "      <td>337</td>\n",
              "      <td>34</td>\n",
              "      <td>154</td>\n",
              "      <td>38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3052121</th>\n",
              "      <td>2015-07-20</td>\n",
              "      <td>12:14:10</td>\n",
              "      <td>26.65438</td>\n",
              "      <td>-81.86906</td>\n",
              "      <td>27</td>\n",
              "      <td>MetroPCS</td>\n",
              "      <td>MetroPCS</td>\n",
              "      <td>2</td>\n",
              "      <td>STATE_EMERGENCY_ONLY</td>\n",
              "      <td>4G</td>\n",
              "      <td>65.4</td>\n",
              "      <td>8.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(-81.86906 26.65438)</td>\n",
              "      <td>337</td>\n",
              "      <td>34</td>\n",
              "      <td>154</td>\n",
              "      <td>38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3052122</th>\n",
              "      <td>2015-07-20</td>\n",
              "      <td>14:19:52</td>\n",
              "      <td>26.65277</td>\n",
              "      <td>-81.97128</td>\n",
              "      <td>27</td>\n",
              "      <td>MetroPCS</td>\n",
              "      <td>MetroPCS</td>\n",
              "      <td>2</td>\n",
              "      <td>STATE_EMERGENCY_ONLY</td>\n",
              "      <td>4G</td>\n",
              "      <td>13.8</td>\n",
              "      <td>3.0</td>\n",
              "      <td>56.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>IN_VEHICLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(-81.97128 26.65277)</td>\n",
              "      <td>337</td>\n",
              "      <td>34</td>\n",
              "      <td>154</td>\n",
              "      <td>38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3052123</th>\n",
              "      <td>2015-07-20</td>\n",
              "      <td>14:51:11</td>\n",
              "      <td>26.59405</td>\n",
              "      <td>-81.87668</td>\n",
              "      <td>15</td>\n",
              "      <td>MetroPCS</td>\n",
              "      <td>MetroPCS</td>\n",
              "      <td>2</td>\n",
              "      <td>STATE_EMERGENCY_ONLY</td>\n",
              "      <td>4G</td>\n",
              "      <td>6.3</td>\n",
              "      <td>12.0</td>\n",
              "      <td>10.0</td>\n",
              "      <td>gps</td>\n",
              "      <td>ON_BICYCLE</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>POINT(-81.87668 26.59405)</td>\n",
              "      <td>337</td>\n",
              "      <td>34</td>\n",
              "      <td>154</td>\n",
              "      <td>38</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3052124 rows × 24 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "               date      hour       lat      long  ...    A    B    C    D\n",
              "0        2016-02-22  21:01:29  41.21993   1.75448  ...  425  915  796  164\n",
              "1        2016-02-22  21:01:29  41.21993   1.75448  ...  687  905  296  491\n",
              "2        2016-02-22  06:31:35  41.24392   1.87257  ...  425  915  796  164\n",
              "3        2016-02-22  06:31:35  41.24392   1.87257  ...  687  905  296  491\n",
              "4        2016-02-22  15:15:19  41.27102   2.02823  ...  425  915  796  164\n",
              "...             ...       ...       ...       ...  ...  ...  ...  ...  ...\n",
              "3052119  2015-07-20  12:22:07  26.66852 -81.93980  ...  337   34  154   38\n",
              "3052120  2015-07-20  12:11:12  26.65641 -81.84830  ...  337   34  154   38\n",
              "3052121  2015-07-20  12:14:10  26.65438 -81.86906  ...  337   34  154   38\n",
              "3052122  2015-07-20  14:19:52  26.65277 -81.97128  ...  337   34  154   38\n",
              "3052123  2015-07-20  14:51:11  26.59405 -81.87668  ...  337   34  154   38\n",
              "\n",
              "[3052124 rows x 24 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SevM-wU1C6mi"
      },
      "source": [
        "##  05.Discretizacion y 1-hot-encoding"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hGM0TKCnMj8l"
      },
      "source": [
        "Por ultimo, dos tecnicas muy usadas en la preparacion de features para nuestros modelos de ML son la discretizacion y el 1 hot encoding.\n",
        "El 1 hot encoding intenta paliar el posible efecto adverso de convertir valores no numericos en una interpretaciion numerica: el 1 esta numericamente \"mas cerca\" del 2 que el 3.\n",
        "Podemos utilizar la funcion get_dummies de pandas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BCcxuj4xXavR"
      },
      "source": [
        "y = pd.get_dummies(df['activity'], prefix='1HotEnc')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 399
        },
        "id": "jsvTEqDqYrW-",
        "outputId": "40a9fd38-2591-4115-ec6d-be586d1250c8"
      },
      "source": [
        "y"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>1HotEnc_IN_VEHICLE</th>\n",
              "      <th>1HotEnc_ON_BICYCLE</th>\n",
              "      <th>1HotEnc_ON_FOOT</th>\n",
              "      <th>1HotEnc_STILL</th>\n",
              "      <th>1HotEnc_TILTING</th>\n",
              "      <th>1HotEnc_UNKNOWN</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </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>99995</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99996</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99997</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99998</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99999</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>100000 rows × 6 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "       1HotEnc_IN_VEHICLE  1HotEnc_ON_BICYCLE  ...  1HotEnc_TILTING  1HotEnc_UNKNOWN\n",
              "0                       1                   0  ...                0                0\n",
              "1                       1                   0  ...                0                0\n",
              "2                       1                   0  ...                0                0\n",
              "3                       1                   0  ...                0                0\n",
              "4                       1                   0  ...                0                0\n",
              "...                   ...                 ...  ...              ...              ...\n",
              "99995                   0                   0  ...                0                0\n",
              "99996                   0                   0  ...                0                0\n",
              "99997                   0                   0  ...                0                0\n",
              "99998                   0                   0  ...                0                0\n",
              "99999                   0                   0  ...                0                0\n",
              "\n",
              "[100000 rows x 6 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nMIve8VENNfx"
      },
      "source": [
        "Para generar una cuantizacion vamos primero a generar una columna dummy con una variable continua"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ezZ2zEzraO7a"
      },
      "source": [
        "df_dist = pd.DataFrame(np.random.randint(0,3456,size=(100, 1)), columns=list('A'))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 399
        },
        "id": "S5OEYDz3alUE",
        "outputId": "30a653a4-1592-42e7-eb3c-59d3f2eb7602"
      },
      "source": [
        "df_dist"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>A</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2989</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>313</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2455</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3019</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1791</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>95</th>\n",
              "      <td>3342</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>96</th>\n",
              "      <td>3146</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>97</th>\n",
              "      <td>123</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>98</th>\n",
              "      <td>2066</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>99</th>\n",
              "      <td>1387</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>100 rows × 1 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "       A\n",
              "0   2989\n",
              "1    313\n",
              "2   2455\n",
              "3   3019\n",
              "4   1791\n",
              "..   ...\n",
              "95  3342\n",
              "96  3146\n",
              "97   123\n",
              "98  2066\n",
              "99  1387\n",
              "\n",
              "[100 rows x 1 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s1O59dheNVA5"
      },
      "source": [
        "Y ahora podemos utilizar la funcion ´qcut´ de pandas para discretizar estos valores por ejemplo en 3 buckets"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nefqyfNga1UW",
        "outputId": "414fe2e9-6019-489d-9c5d-fbcec48fd0f6"
      },
      "source": [
        "pd.qcut(df_dist.A, 3, labels=[\"Bajo\",\"Medio\",\"Alto\"])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0      Alto\n",
              "1      Bajo\n",
              "2     Medio\n",
              "3      Alto\n",
              "4     Medio\n",
              "      ...  \n",
              "95     Alto\n",
              "96     Alto\n",
              "97     Bajo\n",
              "98    Medio\n",
              "99    Medio\n",
              "Name: A, Length: 100, dtype: category\n",
              "Categories (3, object): ['Bajo' < 'Medio' < 'Alto']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EEITGRyJT3O4"
      },
      "source": [
        "## One-hot en Pytorch"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JhqVqktfT3O5",
        "outputId": "3c6595ef-26bc-46c6-dc6e-1da3c50dfb3b"
      },
      "source": [
        "import torch\n",
        "x = torch.arange(3, 7)\n",
        "print(x)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([3, 4, 5, 6])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "n1_ha1iVT3O6",
        "outputId": "52855664-4a2b-4038-d7d7-2e502c378c72"
      },
      "source": [
        "x_hot = torch.nn.functional.one_hot(x)\n",
        "print(x_hot)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tensor([[0, 0, 0, 1, 0, 0, 0],\n",
            "        [0, 0, 0, 0, 1, 0, 0],\n",
            "        [0, 0, 0, 0, 0, 1, 0],\n",
            "        [0, 0, 0, 0, 0, 0, 1]])\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}