machine-learning/Analisis_de_datos_Regresion_lineal_Machine_Learning_II_.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Analisis de datos - Regresion lineal - Machine Learning II .ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyOMbvkF+ODRQDXwtOrACRv3",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/rasecotineb/datascience/blob/master/machine-learning/Analisis_de_datos_Regresion_lineal_Machine_Learning_II_.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NaboeXJthh7d",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "2304da16-9262-4c03-d1fe-d97c0eeba008"
      },
      "source": [
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy  as np\n",
        "%ls\n",
        "\n",
        "#El método \"read_csv\" nos permite leer un excel\n",
        "#El link compartido original es: https://drive.google.com/open?id=1kTPOY0ku4vbHF_4dxOXmxVALjFawc2MF\n",
        "#Se modifica el id por uc (user control) y se agrega al final: &export=download\n",
        "dataset = pd.read_csv('https://drive.google.com/uc?id=1kTPOY0ku4vbHF_4dxOXmxVALjFawc2MF&export=download')\n",
        "#dataset = pd.read_csv(file_name)\n",
        "dataset.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[01;34msample_data\u001b[0m/\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1000, 16)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 32
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Md8vAZ8SjLwg",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        },
        "outputId": "d044181c-ff58-4da2-8fb5-ef9a8c3e16ee"
      },
      "source": [
        "dataset.dtypes"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Region Comercial text         object\n",
              "Subregion Comercial text      object\n",
              "Sector Comercial text         object\n",
              "Locacion Comercial text       object\n",
              "Mesa Comercial text           object\n",
              "Ruta Comercial text           object\n",
              "Canal ACT text                object\n",
              "Canal ISSCOM text             object\n",
              "Sub Canal ISSCOM text         object\n",
              "Tipo Venta Cliente text       object\n",
              "Cliente text                  object\n",
              "Volumen C.U                  float64\n",
              "Cuota C.U                    float64\n",
              "Transacciones                  int64\n",
              "Precio Neto C.U S/.          float64\n",
              "Importe Total Factura S/.    float64\n",
              "dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 33
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A08ilGqpYFBv",
        "colab_type": "text"
      },
      "source": [
        "#Pre-Procesamiento de datos"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rgG18xmIjgC7",
        "colab_type": "text"
      },
      "source": [
        "###Determinamos cuantos filas (registros) existen por cada columna (variable) sin contar repeticiones. Si una columna (variable) tuviera los valores [1,1,2,3,3,4] diremos que tiene 4 registros únicos que son [1,2,3,4]:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1X8dszn2jutu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        },
        "outputId": "decec443-d2e8-4824-edf4-888332d78fe4"
      },
      "source": [
        "dataset.nunique()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Region Comercial text          1\n",
              "Subregion Comercial text       8\n",
              "Sector Comercial text         11\n",
              "Locacion Comercial text       21\n",
              "Mesa Comercial text           41\n",
              "Ruta Comercial text          285\n",
              "Canal ACT text                10\n",
              "Canal ISSCOM text             23\n",
              "Sub Canal ISSCOM text         57\n",
              "Tipo Venta Cliente text        1\n",
              "Cliente text                 999\n",
              "Volumen C.U                  885\n",
              "Cuota C.U                    922\n",
              "Transacciones                475\n",
              "Precio Neto C.U S/.          404\n",
              "Importe Total Factura S/.    960\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "le_-ufKXkFMq",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        },
        "outputId": "be73628a-38d0-4e91-f7c8-27f208b37d03"
      },
      "source": [
        "#Porcentaje de filas (registros) únicas por cada columna (variable). Mientras más cercano el valor a 1 significa que tiene más filas (registros) únicos\n",
        "dataset.nunique() / len(dataset)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Region Comercial text        0.001\n",
              "Subregion Comercial text     0.008\n",
              "Sector Comercial text        0.011\n",
              "Locacion Comercial text      0.021\n",
              "Mesa Comercial text          0.041\n",
              "Ruta Comercial text          0.285\n",
              "Canal ACT text               0.010\n",
              "Canal ISSCOM text            0.023\n",
              "Sub Canal ISSCOM text        0.057\n",
              "Tipo Venta Cliente text      0.001\n",
              "Cliente text                 0.999\n",
              "Volumen C.U                  0.885\n",
              "Cuota C.U                    0.922\n",
              "Transacciones                0.475\n",
              "Precio Neto C.U S/.          0.404\n",
              "Importe Total Factura S/.    0.960\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 35
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LAhNmgGIkQLC",
        "colab_type": "text"
      },
      "source": [
        "###Revisamos si las filas (registros) tienen nulos o no:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4pQ42YkBkS3K",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 490
        },
        "outputId": "2bb760ba-c0f8-4046-b9b8-f42ef46f6e58"
      },
      "source": [
        "dataset.isna()==True"
      ],
      "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>Region Comercial text</th>\n",
              "      <th>Subregion Comercial text</th>\n",
              "      <th>Sector Comercial text</th>\n",
              "      <th>Locacion Comercial text</th>\n",
              "      <th>Mesa Comercial text</th>\n",
              "      <th>Ruta Comercial text</th>\n",
              "      <th>Canal ACT text</th>\n",
              "      <th>Canal ISSCOM text</th>\n",
              "      <th>Sub Canal ISSCOM text</th>\n",
              "      <th>Tipo Venta Cliente text</th>\n",
              "      <th>Cliente text</th>\n",
              "      <th>Volumen C.U</th>\n",
              "      <th>Cuota C.U</th>\n",
              "      <th>Transacciones</th>\n",
              "      <th>Precio Neto C.U S/.</th>\n",
              "      <th>Importe Total Factura S/.</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <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>995</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>996</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>997</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>998</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>999</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1000 rows × 16 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     Region Comercial text  ...  Importe Total Factura S/.\n",
              "0                    False  ...                      False\n",
              "1                    False  ...                      False\n",
              "2                    False  ...                      False\n",
              "3                    False  ...                      False\n",
              "4                    False  ...                      False\n",
              "..                     ...  ...                        ...\n",
              "995                  False  ...                      False\n",
              "996                  False  ...                       True\n",
              "997                  False  ...                      False\n",
              "998                  False  ...                      False\n",
              "999                  False  ...                      False\n",
              "\n",
              "[1000 rows x 16 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 36
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4lXlDe_Jkt6z",
        "colab_type": "text"
      },
      "source": [
        "###Determinamos cuantas filas (registros) vacías existen en cada columna (variable) sumando todos sus valores \"True\" que indican nulidad:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7m5yf5BYkv6-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        },
        "outputId": "6052d417-f9dc-45dd-8690-a3ad215e900d"
      },
      "source": [
        "#Cantidad de filas (registros) vacías por cada columna (variable)\n",
        "dataset.isna().sum()\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Region Comercial text         0\n",
              "Subregion Comercial text      0\n",
              "Sector Comercial text         0\n",
              "Locacion Comercial text       0\n",
              "Mesa Comercial text           0\n",
              "Ruta Comercial text           0\n",
              "Canal ACT text                0\n",
              "Canal ISSCOM text             0\n",
              "Sub Canal ISSCOM text         0\n",
              "Tipo Venta Cliente text       0\n",
              "Cliente text                  0\n",
              "Volumen C.U                  34\n",
              "Cuota C.U                     0\n",
              "Transacciones                 0\n",
              "Precio Neto C.U S/.           0\n",
              "Importe Total Factura S/.    34\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kzrPtdIZl2eF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        },
        "outputId": "b058ebf8-7016-4ca1-d7f2-7e7b43d2fcfc"
      },
      "source": [
        "#Porcentaje de filas (registros) vacías por cada columna (variable). Cuanto más cercano el valor a 1 significa que tiene más filas (registros) vacías\n",
        "dataset.isna().sum() / len(dataset)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Region Comercial text        0.000\n",
              "Subregion Comercial text     0.000\n",
              "Sector Comercial text        0.000\n",
              "Locacion Comercial text      0.000\n",
              "Mesa Comercial text          0.000\n",
              "Ruta Comercial text          0.000\n",
              "Canal ACT text               0.000\n",
              "Canal ISSCOM text            0.000\n",
              "Sub Canal ISSCOM text        0.000\n",
              "Tipo Venta Cliente text      0.000\n",
              "Cliente text                 0.000\n",
              "Volumen C.U                  0.034\n",
              "Cuota C.U                    0.000\n",
              "Transacciones                0.000\n",
              "Precio Neto C.U S/.          0.000\n",
              "Importe Total Factura S/.    0.034\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 38
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kPB9VDwpnPI6",
        "colab_type": "text"
      },
      "source": [
        "###Filtramos las columnas (variables) cuyo porcentaje de filas (registros) únicos sea cercano a 1 porque lo más probable es que sean identificadores (IDs), hecho que normalmente no aporta información respecto a un objetivo como puede ser predecir los valores de una columna (variable) y por ende son considerados como ruido:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qZKtB0hinYLf",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "d3424e19-a154-49a6-98b2-5356be50b35a"
      },
      "source": [
        "#Determinamos aquellas filas (registros) que tengan un porcentaje de unicos mayor al valor de 0.7\n",
        "threshold = 0.7\n",
        "\n",
        "unique_percentages = dataset.nunique() / len(dataset)\n",
        "\n",
        "criteria = unique_percentages > threshold\n",
        "\n",
        "columns_to_filter = unique_percentages[criteria].keys()\n",
        "columns_to_filter"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['Cliente text', 'Volumen C.U', 'Cuota C.U',\n",
              "       'Importe Total Factura S/.'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AVJsy7_zjJ7V",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "db5ce158-2f8b-4a22-9c65-7c1c8005a691"
      },
      "source": [
        "arreglo = (x for x in columns_to_filter)\n",
        "print (list(arreglo)[0:2])\n",
        "#index , y = enumerate(columns_to_filter)\n",
        "list(columns_to_filter[0:2])\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['Cliente text', 'Volumen C.U']\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['Cliente text', 'Volumen C.U']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 40
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DsXe4MkSo0nx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 830
        },
        "outputId": "c7790468-6f1d-445a-bae0-41f2c6980525"
      },
      "source": [
        "#Eliminamos las columnas (variables) seleccionadas en el paso anterior. Al aplicar \"drop\" obtenemos un nuevo \"DataFrame\", pero con el parámetro\n",
        "#inplace igual a \"True\" podemos realizar dicha operación de eliminación y sobre-escribir el \"DataFrame\" del cual estamos eliminando con el\n",
        "#nuevo \"DataFrame\" obtenido. El parámetro axis permite seleccionar el eje respecto al cual aplicar la operación \"drop\". Los valores válidos\n",
        "#son [0 o \"index\"] y [1 o \"Después de eliminar las columnas (variables) cuyo porcentaje de filas (registros) únicos sea cercano a 1 procedemos a \n",
        "#eliminar las filas (registros) duplicadas.\n",
        "\n",
        "#Si ejecutáramos esto antes del paso anterior terminaríamos eliminando menos o ninguna fila (registro) debido que estas se diferenciarían entre sí \n",
        "#mediante las columnas (variables) cuyo porcentaje de filas (registros) únicos sea cercano a 1. Por ejemplo, si tuvieramos el siguiente dataset y \n",
        "#tratáramos de eliminar los duplicados no pasaría nada:\n",
        "\n",
        "# A\tB\tC\n",
        "# 1\th\ti\n",
        "# 2\tj\tk\n",
        "# 3\th\ti\n",
        "# Pero si elimináramos la columna A y eliminarámos duplicados obtendríamos el siguiente dataset:\n",
        "\n",
        "# B\tC\n",
        "# h\ti\n",
        "# j\tk\n",
        "# El procedimiento de eliminar duplicados también considera los nulos para determinar si una fila es distinta o no de otra:columns\"]\n",
        "\n",
        "#dataset.drop(columns_to_filter[0], axis=1, inplace=True)\n",
        "dataset.drop(list(columns_to_filter[0:2]), axis=1, inplace=True)  #Elimina Cliente_text y Volumen_CU\n",
        "\n",
        "dataset"
      ],
      "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>Region Comercial text</th>\n",
              "      <th>Subregion Comercial text</th>\n",
              "      <th>Sector Comercial text</th>\n",
              "      <th>Locacion Comercial text</th>\n",
              "      <th>Mesa Comercial text</th>\n",
              "      <th>Ruta Comercial text</th>\n",
              "      <th>Canal ACT text</th>\n",
              "      <th>Canal ISSCOM text</th>\n",
              "      <th>Sub Canal ISSCOM text</th>\n",
              "      <th>Tipo Venta Cliente text</th>\n",
              "      <th>Cuota C.U</th>\n",
              "      <th>Transacciones</th>\n",
              "      <th>Precio Neto C.U S/.</th>\n",
              "      <th>Importe Total Factura S/.</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Centro</td>\n",
              "      <td>Centro 2</td>\n",
              "      <td>71 BAJOPONTINA HUANCAYO</td>\n",
              "      <td>71SV05</td>\n",
              "      <td>71H57</td>\n",
              "      <td>Tradicional</td>\n",
              "      <td>VIV.LOCALES TRADICIONALES</td>\n",
              "      <td>BODEGAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>22.02</td>\n",
              "      <td>227</td>\n",
              "      <td>10.34</td>\n",
              "      <td>429.90</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Centro</td>\n",
              "      <td>Centro 2</td>\n",
              "      <td>71 BAJOPONTINA HUANCAYO</td>\n",
              "      <td>71SV05</td>\n",
              "      <td>71H52</td>\n",
              "      <td>Comer y Beber</td>\n",
              "      <td>RESTAURANT</td>\n",
              "      <td>POLLERIAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>11.89</td>\n",
              "      <td>147</td>\n",
              "      <td>11.70</td>\n",
              "      <td>341.73</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Centro</td>\n",
              "      <td>Centro 2</td>\n",
              "      <td>57 BAJOPONTINA HUANCAVELICA</td>\n",
              "      <td>57SV01</td>\n",
              "      <td>57H03</td>\n",
              "      <td>Comer y Beber</td>\n",
              "      <td>VENTA AL PASO</td>\n",
              "      <td>CARRETILLAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>0.00</td>\n",
              "      <td>151</td>\n",
              "      <td>9.72</td>\n",
              "      <td>247.60</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Ayacucho</td>\n",
              "      <td>AYACUCHO 1</td>\n",
              "      <td>46 MORVISA AYACUCHO</td>\n",
              "      <td>46SV02</td>\n",
              "      <td>46H12</td>\n",
              "      <td>Comercios Minoristas</td>\n",
              "      <td>LICORERIAS</td>\n",
              "      <td>BEBIDAS EN GENERAL</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>9.37</td>\n",
              "      <td>126</td>\n",
              "      <td>10.47</td>\n",
              "      <td>276.58</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. ICA</td>\n",
              "      <td>Ica 1</td>\n",
              "      <td>I3 ECOBESA ICA</td>\n",
              "      <td>I3SV02</td>\n",
              "      <td>I3H12</td>\n",
              "      <td>Tradicional</td>\n",
              "      <td>VIV.LOCALES TRADICIONALES</td>\n",
              "      <td>BODEGAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>24.60</td>\n",
              "      <td>267</td>\n",
              "      <td>9.41</td>\n",
              "      <td>649.88</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",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>995</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. ICA</td>\n",
              "      <td>Ica 1</td>\n",
              "      <td>I3 ECOBESA ICA</td>\n",
              "      <td>I3SV01</td>\n",
              "      <td>I3H07</td>\n",
              "      <td>Tradicional</td>\n",
              "      <td>VIV.LOCALES TRADICIONALES</td>\n",
              "      <td>HOGARES/PARTICULARES</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>43.75</td>\n",
              "      <td>164</td>\n",
              "      <td>10.71</td>\n",
              "      <td>498.13</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>996</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Selva Central</td>\n",
              "      <td>Selva Central 2</td>\n",
              "      <td>47 TRAHIS LA MERCED</td>\n",
              "      <td>47SV01</td>\n",
              "      <td>47H03</td>\n",
              "      <td>Comercios Minoristas</td>\n",
              "      <td>LICORERIAS</td>\n",
              "      <td>BEBIDAS EN GENERAL</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>7.43</td>\n",
              "      <td>0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>997</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. ICA</td>\n",
              "      <td>Ica 1</td>\n",
              "      <td>I3 ECOBESA ICA</td>\n",
              "      <td>I3SV01</td>\n",
              "      <td>I3H03</td>\n",
              "      <td>Comercios Minoristas</td>\n",
              "      <td>VENTAS SERVICIOS ESPECIALIZADO</td>\n",
              "      <td>OTRAS VENTAS Y SERVICIOS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>20.53</td>\n",
              "      <td>60</td>\n",
              "      <td>8.76</td>\n",
              "      <td>142.17</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>998</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Selva Central</td>\n",
              "      <td>Selva Central 1</td>\n",
              "      <td>27 HUALLPA PASCO</td>\n",
              "      <td>27SV01</td>\n",
              "      <td>27H02</td>\n",
              "      <td>Tradicional</td>\n",
              "      <td>VIV.LOCALES TRADICIONALES</td>\n",
              "      <td>BODEGAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>28.16</td>\n",
              "      <td>547</td>\n",
              "      <td>11.10</td>\n",
              "      <td>885.35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>999</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. ICA</td>\n",
              "      <td>Ica 1</td>\n",
              "      <td>I6 ECOBESA CHINCHA</td>\n",
              "      <td>I6SV02</td>\n",
              "      <td>I6H13</td>\n",
              "      <td>Tradicional</td>\n",
              "      <td>VIV.LOCALES TRADICIONALES</td>\n",
              "      <td>BODEGAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>11.40</td>\n",
              "      <td>149</td>\n",
              "      <td>10.09</td>\n",
              "      <td>297.35</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1000 rows × 14 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     Region Comercial text  ... Importe Total Factura S/.\n",
              "0    REGION CENTRO ORIENTE  ...                    429.90\n",
              "1    REGION CENTRO ORIENTE  ...                    341.73\n",
              "2    REGION CENTRO ORIENTE  ...                    247.60\n",
              "3    REGION CENTRO ORIENTE  ...                    276.58\n",
              "4    REGION CENTRO ORIENTE  ...                    649.88\n",
              "..                     ...  ...                       ...\n",
              "995  REGION CENTRO ORIENTE  ...                    498.13\n",
              "996  REGION CENTRO ORIENTE  ...                       NaN\n",
              "997  REGION CENTRO ORIENTE  ...                    142.17\n",
              "998  REGION CENTRO ORIENTE  ...                    885.35\n",
              "999  REGION CENTRO ORIENTE  ...                    297.35\n",
              "\n",
              "[1000 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 41
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q8SNfQ_epTji",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "bfdb11cd-a1b5-416f-9931-96046fec93b6"
      },
      "source": [
        "#Al aplicar \"drop_duplicates\" obtenemos un nuevo \"DataFrame\", pero con el parámetro inplace igual a \"True\" podemos realizar dicha operación de\n",
        "#eliminación y sobre-escribir el \"DataFrame\" del cual estamos eliminando con el nuevo \"DataFrame\" obtenido\n",
        "dataset.drop_duplicates(keep=\"first\", inplace=True)\n",
        "\n",
        "#Elimina Transacciones\n",
        "#dataset.drop(['Transacciones'],axis=1,inplace=True)\n",
        "\n",
        "#Revisamos cuantas filas y columnas tiene el dataset después de eliminar los duplicados\n",
        "dataset.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1000, 14)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4hGqshippvi5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Determinamos que columnas (variables) son del tipo númerico y cuales son categóricas para poder usar esta información más adelante: \n",
        "numeric_columns = list()\n",
        "categorical_columns = list()\n",
        "dictionary_of_columns_with_index_to_categorical = dict()\n",
        "dictionary_of_columns_with_categorical_to_index = dict()\n",
        "\n",
        "for column in dataset:\n",
        "  #Determinamos si la columna es numérica o no\n",
        "  if pd.api.types.is_numeric_dtype(dataset[column]):\n",
        "    numeric_columns.append(column)\n",
        "  else:\n",
        "    #Modificamos el tipo de dato de la columna (variable) mediante \"astype\"\n",
        "    dataset[column] = dataset[column].astype(\"category\")\n",
        "\n",
        "    #Verificamos si el tipo de dato de la columna (variable) fue transformado a categórico correctamente\n",
        "    if not pd.api.types.is_categorical_dtype(dataset[column]):\n",
        "      raise Exception(\"La columna {} no se transformó correctamente a categórica\".format(column))\n",
        "\n",
        "    dictionary_of_columns_with_index_to_categorical[column] = dict()\n",
        "    dictionary_of_columns_with_categorical_to_index[column] = dict()\n",
        "    \n",
        "    #Indexamos los valores (categorías), sin tomar en consideración los nulos, de la columna (variable) y guardamos esa información en los diccionarios\n",
        "    for index, category in enumerate(dataset[column].cat.categories):\n",
        "      dictionary_of_columns_with_index_to_categorical[column][index] = category\n",
        "      dictionary_of_columns_with_categorical_to_index[column][category] = index\n",
        "    \n",
        "    categorical_columns.append(column)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jNeMUTW2p9Cy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "7f1f3cf3-b16b-4d2b-f791-2ab0cc08a26c"
      },
      "source": [
        "numeric_columns"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['Cuota C.U',\n",
              " 'Transacciones',\n",
              " 'Precio Neto C.U S/.',\n",
              " 'Importe Total Factura S/.']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u6h7n5xMqC4H",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        },
        "outputId": "5b817a78-d32c-4c6b-f8d2-38b45e51752a"
      },
      "source": [
        "categorical_columns"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['Region Comercial text',\n",
              " 'Subregion Comercial text',\n",
              " 'Sector Comercial text',\n",
              " 'Locacion Comercial text',\n",
              " 'Mesa Comercial text',\n",
              " 'Ruta Comercial text',\n",
              " 'Canal ACT text',\n",
              " 'Canal ISSCOM text',\n",
              " 'Sub Canal ISSCOM text',\n",
              " 'Tipo Venta Cliente text']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o5GEQSJHqTMP",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        },
        "outputId": "5fbaf1bf-cd62-4644-c35d-498b2a6e03e8"
      },
      "source": [
        "#Visualizamos el tipo de dato de las columnas (variables) para validar que las no numéricas se transformaron a categóricas\n",
        "dataset.dtypes"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Region Comercial text        category\n",
              "Subregion Comercial text     category\n",
              "Sector Comercial text        category\n",
              "Locacion Comercial text      category\n",
              "Mesa Comercial text          category\n",
              "Ruta Comercial text          category\n",
              "Canal ACT text               category\n",
              "Canal ISSCOM text            category\n",
              "Sub Canal ISSCOM text        category\n",
              "Tipo Venta Cliente text      category\n",
              "Cuota C.U                     float64\n",
              "Transacciones                   int64\n",
              "Precio Neto C.U S/.           float64\n",
              "Importe Total Factura S/.     float64\n",
              "dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 46
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ASVAP7QnqbR4",
        "colab_type": "text"
      },
      "source": [
        "###Nos encargamos de las filas (registros) con nulos para las variables numéricas. Para esto podemos eliminar dichas filas (registros) o reemplazarlas con algún valor que puede ser la media, mediana, moda.\n",
        "\n",
        "Un criterio que suele usarse para decidir eliminar las filas (registros) es que estas representen un porcentaje pequeño de todo el dataset:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qY5OR9nXqdQv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e6c8be31-60a2-432f-9ee8-e4aea979b563"
      },
      "source": [
        "#Porcentaje de filas (registros) vacías por cada columna (variable) numérica\n",
        "dataset[numeric_columns].isna().sum() / len(dataset)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Cuota C.U                    0.000\n",
              "Transacciones                0.000\n",
              "Precio Neto C.U S/.          0.000\n",
              "Importe Total Factura S/.    0.034\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Zm7G1SEdqw2U",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cbaf7386-2930-48a0-c0ae-d20d54f394af"
      },
      "source": [
        "#Optamos por reemplazar los nulos con algún valor para la columna (variable) \"Importe Total Factura S/.\" debido que tiene un porcentaje de nulos cercano al 4% en todo el dataset.\n",
        "\n",
        "#Se decide usar la mediana para reemplazar los nulos. Como veremos a continuación, por más que indiquemos obtener la mediana en todo el dataset sólo obtendremos\n",
        "#la mediana de aquellas columnas (variables) que son numéricas\n",
        "dataset[numeric_columns].median()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Cuota C.U                     25.22\n",
              "Transacciones                177.00\n",
              "Precio Neto C.U S/.           11.16\n",
              "Importe Total Factura S/.    381.95\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9e_bmlSOr1uV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Podemos reemplazar los nulos con la mediana en todo el dataset y sólo se reemplazarán las columnas (variables) numéricas. Sin embargo, por formalidad procederemos\n",
        "#a reemplazar los nulos con la mediana sólo de aquellas columnas (variables) numéricas\n",
        "median_of_numeric_columns = dataset[numeric_columns].median()\n",
        "\n",
        "dataset[numeric_columns] = dataset[numeric_columns].fillna(median_of_numeric_columns)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RvcMjPBgsEXc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d03309b1-c576-4408-8032-b23e73bc32cc"
      },
      "source": [
        "#Verificamos que el porcentaje de filas (registros) vacías para las columnas (variables) numéricas es 0\n",
        "dataset[numeric_columns].isna().sum() / len(dataset)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Cuota C.U                    0.0\n",
              "Transacciones                0.0\n",
              "Precio Neto C.U S/.          0.0\n",
              "Importe Total Factura S/.    0.0\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 50
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GEc_ufS5s9cd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#dataset.to_csv(\"weka 1000 registros_raw data_depurado.csv\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZjWGYbsH_MbT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "821f5709-9611-43b0-996f-910b7f481f7f"
      },
      "source": [
        "#Cuantos registros de \"Subregion Comercial text\" = \"Sub. Ayacucho\" hay\n",
        "dataset.loc[(dataset['Subregion Comercial text']=='Sub. Ayacucho')].iloc[:,2].count()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "57"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 52
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R2NiGJXrNTs4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a8553bb2-30ad-4f49-b416-d749859099ae"
      },
      "source": [
        "dataset"
      ],
      "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>Region Comercial text</th>\n",
              "      <th>Subregion Comercial text</th>\n",
              "      <th>Sector Comercial text</th>\n",
              "      <th>Locacion Comercial text</th>\n",
              "      <th>Mesa Comercial text</th>\n",
              "      <th>Ruta Comercial text</th>\n",
              "      <th>Canal ACT text</th>\n",
              "      <th>Canal ISSCOM text</th>\n",
              "      <th>Sub Canal ISSCOM text</th>\n",
              "      <th>Tipo Venta Cliente text</th>\n",
              "      <th>Cuota C.U</th>\n",
              "      <th>Transacciones</th>\n",
              "      <th>Precio Neto C.U S/.</th>\n",
              "      <th>Importe Total Factura S/.</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Centro</td>\n",
              "      <td>Centro 2</td>\n",
              "      <td>71 BAJOPONTINA HUANCAYO</td>\n",
              "      <td>71SV05</td>\n",
              "      <td>71H57</td>\n",
              "      <td>Tradicional</td>\n",
              "      <td>VIV.LOCALES TRADICIONALES</td>\n",
              "      <td>BODEGAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>22.02</td>\n",
              "      <td>227</td>\n",
              "      <td>10.34</td>\n",
              "      <td>429.90</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Centro</td>\n",
              "      <td>Centro 2</td>\n",
              "      <td>71 BAJOPONTINA HUANCAYO</td>\n",
              "      <td>71SV05</td>\n",
              "      <td>71H52</td>\n",
              "      <td>Comer y Beber</td>\n",
              "      <td>RESTAURANT</td>\n",
              "      <td>POLLERIAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>11.89</td>\n",
              "      <td>147</td>\n",
              "      <td>11.70</td>\n",
              "      <td>341.73</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Centro</td>\n",
              "      <td>Centro 2</td>\n",
              "      <td>57 BAJOPONTINA HUANCAVELICA</td>\n",
              "      <td>57SV01</td>\n",
              "      <td>57H03</td>\n",
              "      <td>Comer y Beber</td>\n",
              "      <td>VENTA AL PASO</td>\n",
              "      <td>CARRETILLAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>0.00</td>\n",
              "      <td>151</td>\n",
              "      <td>9.72</td>\n",
              "      <td>247.60</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Ayacucho</td>\n",
              "      <td>AYACUCHO 1</td>\n",
              "      <td>46 MORVISA AYACUCHO</td>\n",
              "      <td>46SV02</td>\n",
              "      <td>46H12</td>\n",
              "      <td>Comercios Minoristas</td>\n",
              "      <td>LICORERIAS</td>\n",
              "      <td>BEBIDAS EN GENERAL</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>9.37</td>\n",
              "      <td>126</td>\n",
              "      <td>10.47</td>\n",
              "      <td>276.58</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. ICA</td>\n",
              "      <td>Ica 1</td>\n",
              "      <td>I3 ECOBESA ICA</td>\n",
              "      <td>I3SV02</td>\n",
              "      <td>I3H12</td>\n",
              "      <td>Tradicional</td>\n",
              "      <td>VIV.LOCALES TRADICIONALES</td>\n",
              "      <td>BODEGAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>24.60</td>\n",
              "      <td>267</td>\n",
              "      <td>9.41</td>\n",
              "      <td>649.88</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",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>995</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. ICA</td>\n",
              "      <td>Ica 1</td>\n",
              "      <td>I3 ECOBESA ICA</td>\n",
              "      <td>I3SV01</td>\n",
              "      <td>I3H07</td>\n",
              "      <td>Tradicional</td>\n",
              "      <td>VIV.LOCALES TRADICIONALES</td>\n",
              "      <td>HOGARES/PARTICULARES</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>43.75</td>\n",
              "      <td>164</td>\n",
              "      <td>10.71</td>\n",
              "      <td>498.13</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>996</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Selva Central</td>\n",
              "      <td>Selva Central 2</td>\n",
              "      <td>47 TRAHIS LA MERCED</td>\n",
              "      <td>47SV01</td>\n",
              "      <td>47H03</td>\n",
              "      <td>Comercios Minoristas</td>\n",
              "      <td>LICORERIAS</td>\n",
              "      <td>BEBIDAS EN GENERAL</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>7.43</td>\n",
              "      <td>0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>381.95</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>997</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. ICA</td>\n",
              "      <td>Ica 1</td>\n",
              "      <td>I3 ECOBESA ICA</td>\n",
              "      <td>I3SV01</td>\n",
              "      <td>I3H03</td>\n",
              "      <td>Comercios Minoristas</td>\n",
              "      <td>VENTAS SERVICIOS ESPECIALIZADO</td>\n",
              "      <td>OTRAS VENTAS Y SERVICIOS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>20.53</td>\n",
              "      <td>60</td>\n",
              "      <td>8.76</td>\n",
              "      <td>142.17</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>998</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. Selva Central</td>\n",
              "      <td>Selva Central 1</td>\n",
              "      <td>27 HUALLPA PASCO</td>\n",
              "      <td>27SV01</td>\n",
              "      <td>27H02</td>\n",
              "      <td>Tradicional</td>\n",
              "      <td>VIV.LOCALES TRADICIONALES</td>\n",
              "      <td>BODEGAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>28.16</td>\n",
              "      <td>547</td>\n",
              "      <td>11.10</td>\n",
              "      <td>885.35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>999</th>\n",
              "      <td>REGION CENTRO ORIENTE</td>\n",
              "      <td>Sub. ICA</td>\n",
              "      <td>Ica 1</td>\n",
              "      <td>I6 ECOBESA CHINCHA</td>\n",
              "      <td>I6SV02</td>\n",
              "      <td>I6H13</td>\n",
              "      <td>Tradicional</td>\n",
              "      <td>VIV.LOCALES TRADICIONALES</td>\n",
              "      <td>BODEGAS</td>\n",
              "      <td>INDIRECTO</td>\n",
              "      <td>11.40</td>\n",
              "      <td>149</td>\n",
              "      <td>10.09</td>\n",
              "      <td>297.35</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1000 rows × 14 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     Region Comercial text  ... Importe Total Factura S/.\n",
              "0    REGION CENTRO ORIENTE  ...                    429.90\n",
              "1    REGION CENTRO ORIENTE  ...                    341.73\n",
              "2    REGION CENTRO ORIENTE  ...                    247.60\n",
              "3    REGION CENTRO ORIENTE  ...                    276.58\n",
              "4    REGION CENTRO ORIENTE  ...                    649.88\n",
              "..                     ...  ...                       ...\n",
              "995  REGION CENTRO ORIENTE  ...                    498.13\n",
              "996  REGION CENTRO ORIENTE  ...                    381.95\n",
              "997  REGION CENTRO ORIENTE  ...                    142.17\n",
              "998  REGION CENTRO ORIENTE  ...                    885.35\n",
              "999  REGION CENTRO ORIENTE  ...                    297.35\n",
              "\n",
              "[1000 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0ZBJ0XLtH_Gu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5e9b5415-2b5a-4ffa-806f-6a8f69037ca3"
      },
      "source": [
        "#Convierte las variables categoricas en numericas ONE-HOT-ENCODING\n",
        "dataset_dummy = pd.get_dummies(data=dataset,columns=categorical_columns, drop_first=False)\n",
        "dataset_dummy.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>Cuota C.U</th>\n",
              "      <th>Transacciones</th>\n",
              "      <th>Precio Neto C.U S/.</th>\n",
              "      <th>Importe Total Factura S/.</th>\n",
              "      <th>Region Comercial text_REGION CENTRO ORIENTE</th>\n",
              "      <th>Subregion Comercial text_Sub. Ayacucho</th>\n",
              "      <th>Subregion Comercial text_Sub. Centro</th>\n",
              "      <th>Subregion Comercial text_Sub. ICA</th>\n",
              "      <th>Subregion Comercial text_Sub. Nor Oriente</th>\n",
              "      <th>Subregion Comercial text_Sub. Norte Chico</th>\n",
              "      <th>Subregion Comercial text_Sub. Oriente</th>\n",
              "      <th>Subregion Comercial text_Sub. Selva Alta</th>\n",
              "      <th>Subregion Comercial text_Sub. Selva Central</th>\n",
              "      <th>Sector Comercial text_AYACUCHO 1</th>\n",
              "      <th>Sector Comercial text_Centro 2</th>\n",
              "      <th>Sector Comercial text_Ica 1</th>\n",
              "      <th>Sector Comercial text_Nor Oriente 1</th>\n",
              "      <th>Sector Comercial text_Nor Oriente 2</th>\n",
              "      <th>Sector Comercial text_Norte Chico 1</th>\n",
              "      <th>Sector Comercial text_Oriente 1</th>\n",
              "      <th>Sector Comercial text_Selva Alta 1</th>\n",
              "      <th>Sector Comercial text_Selva Alta 2</th>\n",
              "      <th>Sector Comercial text_Selva Central 1</th>\n",
              "      <th>Sector Comercial text_Selva Central 2</th>\n",
              "      <th>Locacion Comercial text_07 DISTRAROJ TARMA</th>\n",
              "      <th>Locacion Comercial text_08 TRAHIS PICHANAKI</th>\n",
              "      <th>Locacion Comercial text_14 LAS DUNAS NAZCA</th>\n",
              "      <th>Locacion Comercial text_24 IMPERIOS HUACHO</th>\n",
              "      <th>Locacion Comercial text_27 HUALLPA PASCO</th>\n",
              "      <th>Locacion Comercial text_28 IMPERIOS HUARAL</th>\n",
              "      <th>Locacion Comercial text_46 MORVISA AYACUCHO</th>\n",
              "      <th>Locacion Comercial text_47 TRAHIS LA MERCED</th>\n",
              "      <th>Locacion Comercial text_57 BAJOPONTINA HUANCAVELICA</th>\n",
              "      <th>Locacion Comercial text_63 SANTA MONICA TARAPOTO</th>\n",
              "      <th>Locacion Comercial text_71 BAJOPONTINA HUANCAYO</th>\n",
              "      <th>Locacion Comercial text_72 TRAHIS PUCALLPA</th>\n",
              "      <th>Locacion Comercial text_74 DISELVA TINGO MARIA</th>\n",
              "      <th>Locacion Comercial text_76 DISELVA HUANUCO</th>\n",
              "      <th>Locacion Comercial text_83 CONSUELO CANALES PUQUIO</th>\n",
              "      <th>Locacion Comercial text_97 TRAHIS SATIPO</th>\n",
              "      <th>...</th>\n",
              "      <th>Sub Canal ISSCOM text_EST.SERVIC.INDEPENDI C/MART</th>\n",
              "      <th>Sub Canal ISSCOM text_EST.SERVIC.INDEPENDI S/MART</th>\n",
              "      <th>Sub Canal ISSCOM text_ESTADIOS/COLISEOS</th>\n",
              "      <th>Sub Canal ISSCOM text_FERRETERIA/HERRAMIENTAS</th>\n",
              "      <th>Sub Canal ISSCOM text_FRUTERIA/VERDULERIA</th>\n",
              "      <th>Sub Canal ISSCOM text_HAMBURGUESA</th>\n",
              "      <th>Sub Canal ISSCOM text_HOGARES/PARTICULARES</th>\n",
              "      <th>Sub Canal ISSCOM text_HOTELES/MOTEL/POSADA</th>\n",
              "      <th>Sub Canal ISSCOM text_KIOSKOS</th>\n",
              "      <th>Sub Canal ISSCOM text_LAVANDERIA/TINTORERIA</th>\n",
              "      <th>Sub Canal ISSCOM text_LIBRERIAS/REVISTAS/TABAQ.</th>\n",
              "      <th>Sub Canal ISSCOM text_LICORES/CERVERZA/VINOS</th>\n",
              "      <th>Sub Canal ISSCOM text_MAYORISTAS</th>\n",
              "      <th>Sub Canal ISSCOM text_MINIMARKET INDEPENDIENTE</th>\n",
              "      <th>Sub Canal ISSCOM text_NEGOCIOS EN MERCADOS</th>\n",
              "      <th>Sub Canal ISSCOM text_OFICINAS DE NEGOCIOS PROFESI</th>\n",
              "      <th>Sub Canal ISSCOM text_OTRAS VENTAS Y SERVICIOS</th>\n",
              "      <th>Sub Canal ISSCOM text_OTROS ALOJAMIENTOS</th>\n",
              "      <th>Sub Canal ISSCOM text_OTROS COMIDAS RAPIDAS</th>\n",
              "      <th>Sub Canal ISSCOM text_OTROS RESTAURANTES</th>\n",
              "      <th>Sub Canal ISSCOM text_OTROS SISTEMAS DE TRANSPORTE</th>\n",
              "      <th>Sub Canal ISSCOM text_OTROS VIVERES ESPEC.</th>\n",
              "      <th>Sub Canal ISSCOM text_OTROS(GIMNASIOS)</th>\n",
              "      <th>Sub Canal ISSCOM text_PANADERIA/PASTELERIA</th>\n",
              "      <th>Sub Canal ISSCOM text_PARQUES/PICNIC/BALNEARIOS</th>\n",
              "      <th>Sub Canal ISSCOM text_PELUQUERIA/SALON DE BELLEZA</th>\n",
              "      <th>Sub Canal ISSCOM text_PESCADOS Y MARISCOS</th>\n",
              "      <th>Sub Canal ISSCOM text_PIZZA</th>\n",
              "      <th>Sub Canal ISSCOM text_PLAYAS DE ESTACIONAMIENTO</th>\n",
              "      <th>Sub Canal ISSCOM text_POLLERIAS</th>\n",
              "      <th>Sub Canal ISSCOM text_PUBS</th>\n",
              "      <th>Sub Canal ISSCOM text_RESTAURANTE FINO CATEGORIA</th>\n",
              "      <th>Sub Canal ISSCOM text_RESTAURANTES CATEGORIA MEDIA</th>\n",
              "      <th>Sub Canal ISSCOM text_SECUNDARIA/POLIMODAL</th>\n",
              "      <th>Sub Canal ISSCOM text_SERVIC.PROFESION.PARA OF.</th>\n",
              "      <th>Sub Canal ISSCOM text_SERVICIO AUTOMOTRIZ</th>\n",
              "      <th>Sub Canal ISSCOM text_TERMINAL TERRESTRE</th>\n",
              "      <th>Sub Canal ISSCOM text_TRANSPORTE PUBLICO</th>\n",
              "      <th>Sub Canal ISSCOM text_VIDEO CLUB</th>\n",
              "      <th>Tipo Venta Cliente text_INDIRECTO</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.0</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.00000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>...</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.000000</td>\n",
              "      <td>1000.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>42.538550</td>\n",
              "      <td>307.386000</td>\n",
              "      <td>10.740830</td>\n",
              "      <td>635.306030</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.057000</td>\n",
              "      <td>0.184000</td>\n",
              "      <td>0.262000</td>\n",
              "      <td>0.065000</td>\n",
              "      <td>0.167000</td>\n",
              "      <td>0.067000</td>\n",
              "      <td>0.078000</td>\n",
              "      <td>0.120000</td>\n",
              "      <td>0.057000</td>\n",
              "      <td>0.184000</td>\n",
              "      <td>0.262000</td>\n",
              "      <td>0.047000</td>\n",
              "      <td>0.018000</td>\n",
              "      <td>0.167000</td>\n",
              "      <td>0.067000</td>\n",
              "      <td>0.030000</td>\n",
              "      <td>0.048000</td>\n",
              "      <td>0.050000</td>\n",
              "      <td>0.070000</td>\n",
              "      <td>0.028000</td>\n",
              "      <td>0.015000</td>\n",
              "      <td>0.04600</td>\n",
              "      <td>0.111000</td>\n",
              "      <td>0.022000</td>\n",
              "      <td>0.056000</td>\n",
              "      <td>0.057000</td>\n",
              "      <td>0.031000</td>\n",
              "      <td>0.013000</td>\n",
              "      <td>0.047000</td>\n",
              "      <td>0.171000</td>\n",
              "      <td>0.030000</td>\n",
              "      <td>0.015000</td>\n",
              "      <td>0.033000</td>\n",
              "      <td>0.007000</td>\n",
              "      <td>0.024000</td>\n",
              "      <td>...</td>\n",
              "      <td>0.002000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.002000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.007000</td>\n",
              "      <td>0.002000</td>\n",
              "      <td>0.119000</td>\n",
              "      <td>0.013000</td>\n",
              "      <td>0.030000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.008000</td>\n",
              "      <td>0.014000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.005000</td>\n",
              "      <td>0.023000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.048000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.002000</td>\n",
              "      <td>0.004000</td>\n",
              "      <td>0.002000</td>\n",
              "      <td>0.041000</td>\n",
              "      <td>0.003000</td>\n",
              "      <td>0.014000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.002000</td>\n",
              "      <td>0.016000</td>\n",
              "      <td>0.003000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.011000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.002000</td>\n",
              "      <td>0.003000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.002000</td>\n",
              "      <td>0.006000</td>\n",
              "      <td>0.002000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>0.001000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>60.308946</td>\n",
              "      <td>471.074204</td>\n",
              "      <td>2.347355</td>\n",
              "      <td>943.465308</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.231959</td>\n",
              "      <td>0.387678</td>\n",
              "      <td>0.439943</td>\n",
              "      <td>0.246649</td>\n",
              "      <td>0.373162</td>\n",
              "      <td>0.250147</td>\n",
              "      <td>0.268306</td>\n",
              "      <td>0.325124</td>\n",
              "      <td>0.231959</td>\n",
              "      <td>0.387678</td>\n",
              "      <td>0.439943</td>\n",
              "      <td>0.211745</td>\n",
              "      <td>0.133018</td>\n",
              "      <td>0.373162</td>\n",
              "      <td>0.250147</td>\n",
              "      <td>0.170673</td>\n",
              "      <td>0.213873</td>\n",
              "      <td>0.218054</td>\n",
              "      <td>0.255275</td>\n",
              "      <td>0.165055</td>\n",
              "      <td>0.121613</td>\n",
              "      <td>0.20959</td>\n",
              "      <td>0.314289</td>\n",
              "      <td>0.146757</td>\n",
              "      <td>0.230037</td>\n",
              "      <td>0.231959</td>\n",
              "      <td>0.173404</td>\n",
              "      <td>0.113331</td>\n",
              "      <td>0.211745</td>\n",
              "      <td>0.376697</td>\n",
              "      <td>0.170673</td>\n",
              "      <td>0.121613</td>\n",
              "      <td>0.178726</td>\n",
              "      <td>0.083414</td>\n",
              "      <td>0.153126</td>\n",
              "      <td>...</td>\n",
              "      <td>0.044699</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.044699</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.083414</td>\n",
              "      <td>0.044699</td>\n",
              "      <td>0.323951</td>\n",
              "      <td>0.113331</td>\n",
              "      <td>0.170673</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.089129</td>\n",
              "      <td>0.117549</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.070569</td>\n",
              "      <td>0.149978</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.213873</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.044699</td>\n",
              "      <td>0.063151</td>\n",
              "      <td>0.044699</td>\n",
              "      <td>0.198389</td>\n",
              "      <td>0.054717</td>\n",
              "      <td>0.117549</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.044699</td>\n",
              "      <td>0.125538</td>\n",
              "      <td>0.054717</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.104355</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.044699</td>\n",
              "      <td>0.054717</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.044699</td>\n",
              "      <td>0.077266</td>\n",
              "      <td>0.044699</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.031623</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>8.770000</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>13.587500</td>\n",
              "      <td>89.750000</td>\n",
              "      <td>10.390000</td>\n",
              "      <td>204.335000</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>25.220000</td>\n",
              "      <td>177.000000</td>\n",
              "      <td>11.160000</td>\n",
              "      <td>381.950000</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>47.120000</td>\n",
              "      <td>346.000000</td>\n",
              "      <td>11.780000</td>\n",
              "      <td>709.032500</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</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",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>...</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>696.720000</td>\n",
              "      <td>6357.000000</td>\n",
              "      <td>16.900000</td>\n",
              "      <td>10914.320000</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.00000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>...</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>8 rows × 462 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "         Cuota C.U  ...  Tipo Venta Cliente text_INDIRECTO\n",
              "count  1000.000000  ...                             1000.0\n",
              "mean     42.538550  ...                                1.0\n",
              "std      60.308946  ...                                0.0\n",
              "min       0.000000  ...                                1.0\n",
              "25%      13.587500  ...                                1.0\n",
              "50%      25.220000  ...                                1.0\n",
              "75%      47.120000  ...                                1.0\n",
              "max     696.720000  ...                                1.0\n",
              "\n",
              "[8 rows x 462 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 54
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-YUwhzrzRzqd",
        "colab_type": "text"
      },
      "source": [
        "#Feature Engineering"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H9J5R-jONqsN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "31a26a0f-0cb4-49e8-9933-3d0a228b0984"
      },
      "source": [
        "# Columnas de REGION COMERCIAL y TIPO VENTA CLIENTE tienen un unico valor y se eliminan\n",
        "# dataset_dummy.drop(columns=['Region Comercial text_REGION CENTRO ORIENTE','Tipo Venta Cliente text_INDIRECTO'],inplace = True)\n",
        "# df = pd.DataFrame(dataset_dummy).describe()\n",
        "# df2 = pd.DataFrame(df.iloc[2])\n",
        "# df2[df2['std']==0].T\n",
        "from sklearn.feature_selection import VarianceThreshold\n",
        "selector = VarianceThreshold()\n",
        "selector.fit_transform(dataset_dummy)   #Va a eliminar aquellas características con poca o nula varianza (datos con poca variabilidad)\n",
        "indices = selector.get_support()        #Obtiene la matriz booleana con el resultado de seleccion de las variables conservadas (TRUE) y eliminadas (FALSE)\n",
        "\n",
        "eliminados = np.where(indices == False) #Obtiene los indices de las variables (columnas) eliminadas \n",
        "dataset_dummy.columns[eliminados]       #Muestra cuales han sido las columnas eliminadas\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['Region Comercial text_REGION CENTRO ORIENTE', 'Tipo Venta Cliente text_INDIRECTO'], dtype='object')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 55
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qUMCk4TzJE3k",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        },
        "outputId": "fb11e452-2939-4738-8450-d91949969109"
      },
      "source": [
        "from sklearn.feature_selection import mutual_info_regression\n",
        "X = dataset_dummy.drop(columns='Importe Total Factura S/.')\n",
        "y = dataset_dummy['Importe Total Factura S/.']  #Target\n",
        "feature = mutual_info_regression(X,y)   #Feature selection para calculo de variable continua por regresion\n",
        "arr_ind = feature.argsort()[-10:][::-1] #Obtiene los indices de las 10 variables explicativas con mayor valor, las ordena de mayor a menor\n",
        "print(feature[arr_ind],arr_ind)         #Imprime los maximos valores del MUTUAL_INFO_REGRESSION y sus respectivos indices\n",
        "print(X.columns[arr_ind])               #Nombre de las variables (columnas) que finalmente se van a considerar\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[1.23063477 0.64404682 0.12753236 0.02795522 0.02593161 0.02561318\n",
            " 0.0232073  0.02129554 0.0199576  0.01897731] [  1   0   2 370  46  25  17 402  33  32]\n",
            "Index(['Transacciones', 'Cuota C.U', 'Precio Neto C.U S/.',\n",
            "       'Canal ACT text_Comer y Beber', 'Mesa Comercial text_14SV01',\n",
            "       'Locacion Comercial text_14 LAS DUNAS NAZCA',\n",
            "       'Sector Comercial text_Norte Chico 1',\n",
            "       'Canal ISSCOM text_VIV.LOCALES TRADICIONALES',\n",
            "       'Locacion Comercial text_71 BAJOPONTINA HUANCAYO',\n",
            "       'Locacion Comercial text_63 SANTA MONICA TARAPOTO'],\n",
            "      dtype='object')\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LKc8RHLzYQvR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 470
        },
        "outputId": "bf044e91-ac54-47fc-f2b0-025bab733156"
      },
      "source": [
        "X_featured = dataset_dummy[X.columns[arr_ind]]\n",
        "X_featured"
      ],
      "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>Transacciones</th>\n",
              "      <th>Cuota C.U</th>\n",
              "      <th>Precio Neto C.U S/.</th>\n",
              "      <th>Canal ACT text_Comer y Beber</th>\n",
              "      <th>Mesa Comercial text_14SV01</th>\n",
              "      <th>Locacion Comercial text_14 LAS DUNAS NAZCA</th>\n",
              "      <th>Sector Comercial text_Norte Chico 1</th>\n",
              "      <th>Canal ISSCOM text_VIV.LOCALES TRADICIONALES</th>\n",
              "      <th>Locacion Comercial text_71 BAJOPONTINA HUANCAYO</th>\n",
              "      <th>Locacion Comercial text_63 SANTA MONICA TARAPOTO</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>227</td>\n",
              "      <td>22.02</td>\n",
              "      <td>10.34</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>147</td>\n",
              "      <td>11.89</td>\n",
              "      <td>11.70</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>151</td>\n",
              "      <td>0.00</td>\n",
              "      <td>9.72</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>126</td>\n",
              "      <td>9.37</td>\n",
              "      <td>10.47</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>267</td>\n",
              "      <td>24.60</td>\n",
              "      <td>9.41</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </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",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>995</th>\n",
              "      <td>164</td>\n",
              "      <td>43.75</td>\n",
              "      <td>10.71</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>996</th>\n",
              "      <td>0</td>\n",
              "      <td>7.43</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>997</th>\n",
              "      <td>60</td>\n",
              "      <td>20.53</td>\n",
              "      <td>8.76</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>998</th>\n",
              "      <td>547</td>\n",
              "      <td>28.16</td>\n",
              "      <td>11.10</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>999</th>\n",
              "      <td>149</td>\n",
              "      <td>11.40</td>\n",
              "      <td>10.09</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1000 rows × 10 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     Transacciones  ...  Locacion Comercial text_63 SANTA MONICA TARAPOTO\n",
              "0              227  ...                                                 0\n",
              "1              147  ...                                                 0\n",
              "2              151  ...                                                 0\n",
              "3              126  ...                                                 0\n",
              "4              267  ...                                                 0\n",
              "..             ...  ...                                               ...\n",
              "995            164  ...                                                 0\n",
              "996              0  ...                                                 0\n",
              "997             60  ...                                                 0\n",
              "998            547  ...                                                 0\n",
              "999            149  ...                                                 0\n",
              "\n",
              "[1000 rows x 10 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 57
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GUnWrbWzYbEb",
        "colab_type": "text"
      },
      "source": [
        "#Modelo de Regresion Lineal Múltiple"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VfUGbWQlYghG",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn import datasets,linear_model, preprocessing\n",
        "from sklearn.model_selection import train_test_split\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LCXeoX2WsOwv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        },
        "outputId": "c07dea60-6bdc-4cae-bd80-73cbbf7e2b25"
      },
      "source": [
        "#Separamos data de entrenamiento y de test\n",
        "X_train, X_test, y_train,y_test = train_test_split(X_featured,y,test_size=0.33)\n",
        "#Normalizamos datos (features)\n",
        "scaler = preprocessing.StandardScaler().fit(X_train)\n",
        "X_train= scaler.transform(X_train)\n",
        "X_test = scaler.transform(X_test)\n",
        "pd.DataFrame(X_test).describe()\n",
        "X_test = pd.DataFrame(X_test)\n",
        "X_train = pd.DataFrame(X_train)\n",
        "X_test.describe()\n"
      ],
      "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>0</th>\n",
              "      <th>1</th>\n",
              "      <th>2</th>\n",
              "      <th>3</th>\n",
              "      <th>4</th>\n",
              "      <th>5</th>\n",
              "      <th>6</th>\n",
              "      <th>7</th>\n",
              "      <th>8</th>\n",
              "      <th>9</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>330.000000</td>\n",
              "      <td>330.000000</td>\n",
              "      <td>330.000000</td>\n",
              "      <td>330.000000</td>\n",
              "      <td>330.000000</td>\n",
              "      <td>330.000000</td>\n",
              "      <td>330.000000</td>\n",
              "      <td>330.000000</td>\n",
              "      <td>330.000000</td>\n",
              "      <td>330.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>-0.084196</td>\n",
              "      <td>-0.060417</td>\n",
              "      <td>0.007971</td>\n",
              "      <td>-0.061060</td>\n",
              "      <td>-0.065531</td>\n",
              "      <td>-0.065531</td>\n",
              "      <td>-0.107109</td>\n",
              "      <td>0.000185</td>\n",
              "      <td>0.055797</td>\n",
              "      <td>0.078880</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.942021</td>\n",
              "      <td>0.892556</td>\n",
              "      <td>1.022482</td>\n",
              "      <td>0.938471</td>\n",
              "      <td>0.854193</td>\n",
              "      <td>0.854193</td>\n",
              "      <td>0.901735</td>\n",
              "      <td>1.001478</td>\n",
              "      <td>1.049471</td>\n",
              "      <td>1.165817</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>-0.668615</td>\n",
              "      <td>-0.701550</td>\n",
              "      <td>-4.607250</td>\n",
              "      <td>-0.423766</td>\n",
              "      <td>-0.231212</td>\n",
              "      <td>-0.231212</td>\n",
              "      <td>-0.469469</td>\n",
              "      <td>-1.240112</td>\n",
              "      <td>-0.443203</td>\n",
              "      <td>-0.208839</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>-0.487240</td>\n",
              "      <td>-0.486235</td>\n",
              "      <td>-0.085710</td>\n",
              "      <td>-0.423766</td>\n",
              "      <td>-0.231212</td>\n",
              "      <td>-0.231212</td>\n",
              "      <td>-0.469469</td>\n",
              "      <td>-1.240112</td>\n",
              "      <td>-0.443203</td>\n",
              "      <td>-0.208839</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>-0.296483</td>\n",
              "      <td>-0.288867</td>\n",
              "      <td>0.173951</td>\n",
              "      <td>-0.423766</td>\n",
              "      <td>-0.231212</td>\n",
              "      <td>-0.231212</td>\n",
              "      <td>-0.469469</td>\n",
              "      <td>0.806379</td>\n",
              "      <td>-0.443203</td>\n",
              "      <td>-0.208839</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>-0.010869</td>\n",
              "      <td>0.000716</td>\n",
              "      <td>0.449707</td>\n",
              "      <td>-0.423766</td>\n",
              "      <td>-0.231212</td>\n",
              "      <td>-0.231212</td>\n",
              "      <td>-0.469469</td>\n",
              "      <td>0.806379</td>\n",
              "      <td>-0.443203</td>\n",
              "      <td>-0.208839</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>11.625299</td>\n",
              "      <td>10.462273</td>\n",
              "      <td>2.049521</td>\n",
              "      <td>2.359794</td>\n",
              "      <td>4.325030</td>\n",
              "      <td>4.325030</td>\n",
              "      <td>2.130068</td>\n",
              "      <td>0.806379</td>\n",
              "      <td>2.256304</td>\n",
              "      <td>4.788379</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                0           1           2  ...           7           8           9\n",
              "count  330.000000  330.000000  330.000000  ...  330.000000  330.000000  330.000000\n",
              "mean    -0.084196   -0.060417    0.007971  ...    0.000185    0.055797    0.078880\n",
              "std      0.942021    0.892556    1.022482  ...    1.001478    1.049471    1.165817\n",
              "min     -0.668615   -0.701550   -4.607250  ...   -1.240112   -0.443203   -0.208839\n",
              "25%     -0.487240   -0.486235   -0.085710  ...   -1.240112   -0.443203   -0.208839\n",
              "50%     -0.296483   -0.288867    0.173951  ...    0.806379   -0.443203   -0.208839\n",
              "75%     -0.010869    0.000716    0.449707  ...    0.806379   -0.443203   -0.208839\n",
              "max     11.625299   10.462273    2.049521  ...    0.806379    2.256304    4.788379\n",
              "\n",
              "[8 rows x 10 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 59
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l5oBHAk_6I8g",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "b5f3b72a-0d1a-4f0b-8c6b-a49530ea0f7e"
      },
      "source": [
        "print('',X_test.shape,y_test.shape,'\\n',X_train.shape,y_train.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            " (330, 10) (330,) \n",
            " (670, 10) (670,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LJdwv19Kd_Pe",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "82958464-0213-4513-87b5-997c74aa635d"
      },
      "source": [
        "#Seteo del modelo de regresion lineal\n",
        "lr_multiple = linear_model.LinearRegression()\n",
        "#entrenamiento del modelo\n",
        "lr_multiple.fit(X_train,y_train)\n",
        "\n",
        "y_prediction   = lr_multiple.predict(X_test)\n",
        "score_r2_train = lr_multiple.score(X_train,y_train)\n",
        "score_r2_test  = lr_multiple.score(X_test,y_test)\n",
        "coeficientes   = lr_multiple.coef_\n",
        "intercepto     = lr_multiple.intercept_\n",
        "\n",
        "# i = 0\n",
        "# print('Real: {}'.format(y_train.iloc[i]))\n",
        "# print('Predicción: {}'.format(y_prediction[i]))\n",
        "# y = np.dot(X_train.iloc[i],coeficientes) + intercepto\n",
        "# print('Verificacion de prediccion: {}'.format(y))\n",
        "\n",
        "print('R2 Train: {:.5f}, R2 Test: {:.5f}'.format(score_r2_train,score_r2_test))\n",
        "print('Intercepto: {:.5f}, Coeficientes: {}'.format(intercepto, coeficientes))\n",
        "\n",
        "# plt.figure(figsize=(6,6))\n",
        "# plt.scatter(y_prediction,y_test)\n",
        "# plt.title('Predicción vs Real')\n",
        "# plt.xlabel(\"Prediccion\")\n",
        "# plt.ylabel(\"Real\")\n",
        "# plt.show()\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "R2 Train: 0.84915, R2 Test: 0.83237\n",
            "Intercepto: 656.43863, Coeficientes: [766.15465417 203.40163486 -65.14122555  19.05319303  10.08469195\n",
            "  10.08469195 -19.56259368  24.69979908  33.11855839  -7.25167069]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8eP6V17VtBwq",
        "colab_type": "text"
      },
      "source": [
        "##Regresión LASSO"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P3ll_fZyhLet",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "210485ba-9617-4fcd-ef8f-4b79160bb9fe"
      },
      "source": [
        "#Seteo del modelo de regresion lineal LASSO\n",
        "lr_lasso = linear_model.Lasso(fit_intercept=True)\n",
        "#entrenamiento del modelo\n",
        "lr_lasso.fit(X_train,y_train)\n",
        "\n",
        "y_prediction   = lr_lasso.predict(X_test)\n",
        "score_r2_train = lr_lasso.score(X_train,y_train)\n",
        "score_r2_test  = lr_lasso.score(X_test,y_test)\n",
        "coeficientes   = lr_lasso.coef_\n",
        "intercepto     = lr_lasso.intercept_\n",
        "\n",
        "# print(y_train.iloc[0])\n",
        "# print(y_prediction[0])\n",
        "# y = np.dot(X_test.iloc[0],coeficientes) + intercepto\n",
        "# print(y)\n",
        "\n",
        "#coeficientes    = coeficientes[coeficientes !=0]\n",
        "# print(y_prediction)\n",
        "print('R2 Train: {:.5f}, R2 Test: {:.5f}'.format(score_r2_train,score_r2_test),'\\n', 'Intercepto: {:.5f}, Coeficientes: {}'.format(intercepto, coeficientes))\n",
        "\n",
        "# plt.figure(figsize=(20,20))\n",
        "# plt.scatter(y_prediction,y_test,)\n",
        "# plt.title('Predicción vs Real')\n",
        "# plt.xlabel(\"Prediccion\")\n",
        "# plt.ylabel(\"Real\")\n",
        "# plt.show()\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "R2 Train: 0.84914, R2 Test: 0.83304 \n",
            " Intercepto: 656.43863, Coeficientes: [ 7.65282993e+02  2.02778875e+02 -6.40810538e+01  1.67094145e+01\n",
            "  1.94475222e+01  8.41621963e-14 -1.84104830e+01  2.20296969e+01\n",
            "  3.16527888e+01 -6.44691622e+00]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Sm4Kx2PUrPbD",
        "colab_type": "text"
      },
      "source": [
        "#Modelo de Regresion SVR"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UtcMjVrerWJQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "80b412e6-e091-4e7c-c9db-ab141b2a0de9"
      },
      "source": [
        "from sklearn.svm import SVR\n",
        "from sklearn.pipeline import make_pipeline\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "Cs = np.logspace(-4,4,9)  # ~ [0.01, 0.03, 0.1, ..., 10000]\n",
        "gammas = np.logspace(-2,2,9)  # [0.0001, 0.001, ..., 100]\n",
        "\n",
        "mejor_svr = None\n",
        "mejor_score = 0\n",
        "for C in Cs:\n",
        "    for gamma in gammas:\n",
        "        \n",
        "        svr = SVR(kernel='rbf', C=C, gamma=gamma, epsilon = 0.2) \n",
        "        svr.fit(X_train, y_train)\n",
        "        \n",
        "        score_val = svr.score(X_test, y_test)\n",
        "        \n",
        "        if score_val > mejor_score:\n",
        "            mejor_score = score_val\n",
        "            mejor_svr = svr\n",
        "\n",
        "            \n",
        "\n",
        "\n",
        "print ('Mejor valor de C: %0.4f' % mejor_svr.get_params()['C'])\n",
        "print ('Mejor valor de gamma: %0.4f' % mejor_svr.get_params()['gamma'])\n",
        "print ('Exactitud en el conjunto de entrenamiento: %0.4f' % mejor_svr.score(X_train, y_train))\n",
        "print ('Exactitud en el conjunto de validación: %0.4f' % mejor_svr.score(X_test, y_test))\n",
        "\n",
        "\n",
        "\n",
        "# model = SVR(kernel='rbf', C=10000.0, epsilon=0.1)\n",
        "# model.fit(X_train, y_train)\n",
        "# score_train = model.score(X_train, y_train)\n",
        "# score_val = model.score(X_test, y_test)\n",
        "\n",
        "# print ('Exactitud en el conjunto de entrenamiento: %0.4f' % score_train)\n",
        "# print ('Exactitud en el conjunto de validación: %0.4f' % score_val)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mejor valor de C: 10000.0000\n",
            "Mejor valor de gamma: 0.0100\n",
            "Exactitud en el conjunto de entrenamiento: 0.8720\n",
            "Exactitud en el conjunto de validación: 0.8892\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AafLrNmMT4IB",
        "colab_type": "text"
      },
      "source": [
        "#Modelo de Regresion Random Forest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w3y0Ht-OUBFq",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "045d515e-b900-4bcf-9606-f1a7e5df9fe9"
      },
      "source": [
        "from sklearn.ensemble import RandomForestRegressor\n",
        "mejor_rf = None\n",
        "mejor_score = 0\n",
        "# Busqueda secuencial de los mejores parametros\n",
        "for n_estimator in range(1,100,1):\n",
        "    for max_depth in range(2,10,1):\n",
        "        \n",
        "        rf = RandomForestRegressor(n_estimators = n_estimator, max_depth=max_depth, random_state=0)\n",
        "        rf.fit(X_train, y_train)\n",
        "        \n",
        "        score_val = rf.score(X_test, y_test)\n",
        "        \n",
        "        if score_val > mejor_score:\n",
        "            mejor_score = score_val\n",
        "            mejor_rf = rf\n",
        "\n",
        "            \n",
        "\n",
        "\n",
        "print ('Mejor valor de n_estimators: %0.4f' % mejor_rf.get_params()['n_estimators'])\n",
        "print ('Mejor valor de max_depth: %0.4f' % mejor_rf.get_params()['max_depth'])\n",
        "print ('Exactitud en el conjunto de entrenamiento: %0.4f' % mejor_rf.score(X_train, y_train))\n",
        "print ('Exactitud en el conjunto de validación: %0.4f' % mejor_rf.score(X_test, y_test))\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mejor valor de n_estimators: 5.0000\n",
            "Mejor valor de max_depth: 5.0000\n",
            "Exactitud en el conjunto de entrenamiento: 0.9389\n",
            "Exactitud en el conjunto de validación: 0.9477\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gg4u1pU7hi5s",
        "colab_type": "text"
      },
      "source": [
        "#Modelos de Ensamble - Regresión"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CwKAahnjhqvP",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        },
        "outputId": "48745b56-78aa-413f-8df2-e12a3e3313a3"
      },
      "source": [
        "from sklearn.ensemble import VotingRegressor\n",
        "\n",
        "\n",
        "#La lista weights contiene los pesos de los clasificadores de regresión\n",
        "eclf = VotingRegressor(estimators=[('lr', lr_multiple), ('lr_lasso', lr_lasso), ('svr', mejor_svr), \n",
        "                                    ('rf', mejor_rf)], weights=[2, 3, 5, 4])\n",
        "eclf.fit(X_train,y_train)\n",
        "\n",
        "#Precision de los modelos sobre el dataset de test\n",
        "print (\"Exactitud del modelo de regresión lineal      :\", lr_multiple.score(X_test, y_test))\n",
        "print (\"Exactitud del modelo de regresión lasso       :\", lr_lasso.score(X_test, y_test))\n",
        "print (\"Exactitud del modelo de SVR                   :\", mejor_svr.score(X_test, y_test))\n",
        "print (\"Exactitud del modelo random forest            :\", mejor_rf.score(X_test, y_test))\n",
        "print (\"Exactitud del modelo de ensamble por votación :\", eclf.score(X_test, y_test))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Exactitud del modelo de regresión lineal      : 0.902076358017937\n",
            "Exactitud del modelo de regresión lasso       : 0.9032810518962526\n",
            "Exactitud del modelo de SVR                   : 0.8891560774465346\n",
            "Exactitud del modelo random forest            : 0.9476728680648052\n",
            "Exactitud del modelo de ensamble por votación : 0.9493594016228307\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_JZ5zxQttXgs",
        "colab_type": "text"
      },
      "source": [
        "#Gradient Boosting - Regresión"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6p0U_qIitlQg",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 598
        },
        "outputId": "7f0e9af5-2cf5-41cb-8820-65893d0ce410"
      },
      "source": [
        "from sklearn.ensemble import GradientBoostingRegressor\n",
        "from sklearn.metrics import median_absolute_error, mean_squared_error, mean_absolute_error\n",
        "\n",
        "from pprint import pprint\n",
        "\n",
        "lr = 1\n",
        "#El criterio paa\n",
        "gb = GradientBoostingRegressor(learning_rate=lr, criterion='mae', n_estimators=100, random_state=0)\n",
        "gb.fit(X_train, y_train)\n",
        "\n",
        "# Se obtiene el error medio absoluto (MAE) de cada una de las iteraciones (n_estimators)\n",
        "errors = [(mean_absolute_error(y_test, y_pred))\n",
        "          for y_pred in gb.staged_predict(X_test)]\n",
        "print(\"Los errores incurridos por los sucesivos ensambles de árboles son:\", errors)\n",
        "\n",
        "bst_n_estimators = np.argmin(errors) + 1   #El mejor modelo es el que tiene menor error absoluto promedio\n",
        "min_error = np.min(errors)                 #Menor valor MAE\n",
        "\n",
        "\n",
        "print('Total de iteraciones: {}'.format(len(errors)))                    #Total de errores, uno por cada ensamble\n",
        "print('Mejor iteración del GradientBoost: {}'.format(bst_n_estimators))  #Mejor iteración\n",
        "print('Menor error: {}'.format(errors[bst_n_estimators-1]))              #Imprime el error\n",
        "print('Menor error: {} \\n'.format(min_error))                            #Imprime el error\n",
        "\n",
        "gb_best = GradientBoostingRegressor(learning_rate=lr, criterion='mae', n_estimators=bst_n_estimators, random_state=0)\n",
        "gb_best.fit(X_train, y_train)\n",
        "\n",
        "print('Los hiperparámetros del mejor modelo son:')\n",
        "pprint(gb_best.get_params())\n",
        "print()\n",
        "\n",
        "print('Exactitud de GradientBoosting en entrenamiento (R2-score):', gb_best.score(X_train, y_train))\n",
        "print('Exactitud de GradientBoosting en validación (R2-score):', gb_best.score(X_test, y_test))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Los errores incurridos por los sucesivos ensambles de árboles son: [150.2370303030303, 147.2309696969697, 141.1606363636364, 143.0499356060606, 142.70807575757576, 138.52323863636363, 139.5148181818182, 142.6420681818182, 142.85723674242425, 143.0779678030303, 143.7034034090909, 144.0610729166667, 143.83008049242423, 143.8887301136364, 143.98064157196973, 144.02354876893938, 144.83148153409093, 147.16984706439396, 146.8258636363636, 148.1106957859848, 148.0752258522727, 148.21289015151513, 148.8485991950757, 148.99626763731058, 148.83110002367422, 150.31465565814398, 149.41277734375004, 149.67604580965912, 149.39118241003789, 149.47786600378788, 149.41911115056817, 149.82920750473488, 149.85750174597538, 150.19600571141103, 150.5881429036459, 151.8094751716383, 151.76702832031253, 152.9231552734374, 152.9140368134468, 152.69846753669506, 151.21506210049714, 153.056490271366, 153.0938639914772, 153.01066039299243, 153.63269868607952, 153.33506480823857, 153.18531659860324, 154.00460857599427, 153.50748596561303, 153.43725713186544, 153.0254092610675, 153.09187338349292, 153.13048852908972, 153.13787951475194, 153.0800384281042, 152.31751465028694, 152.32602912116766, 153.21232431770085, 153.24356077205766, 153.44609943320526, 153.3454590856377, 153.7291733745227, 153.7231303197686, 153.47982327224256, 153.56210324420343, 153.31372706234083, 153.31202636210116, 153.25941645073158, 151.98707589120568, 152.00136755740996, 151.94515472966967, 152.76295354346064, 152.77201938074276, 153.14417048205738, 153.17698478432845, 153.17698478432845, 153.30576089702942, 153.31766935773314, 154.52093078731752, 154.55426536950188, 154.58643697773314, 154.56150270334868, 154.5403509703838, 154.5075472071676, 154.40687580493002, 154.84783342950996, 154.8508636631532, 154.8508636631532, 154.8508636631532, 154.8508636631532, 154.8508636631532, 154.8508636631532, 154.86012252226743, 154.82073173447813, 155.09685909213445, 155.11988367853024, 155.11991887601943, 155.11991887601943, 155.11991887601943, 155.11991887601943]\n",
            "Total de iteraciones: 100\n",
            "Mejor iteración del GradientBoost: 6\n",
            "Menor error: 138.52323863636363\n",
            "Menor error: 138.52323863636363 \n",
            "\n",
            "Los hiperparámetros del mejor modelo son:\n",
            "{'alpha': 0.9,\n",
            " 'ccp_alpha': 0.0,\n",
            " 'criterion': 'mae',\n",
            " 'init': None,\n",
            " 'learning_rate': 1,\n",
            " 'loss': 'ls',\n",
            " 'max_depth': 3,\n",
            " 'max_features': None,\n",
            " 'max_leaf_nodes': None,\n",
            " 'min_impurity_decrease': 0.0,\n",
            " 'min_impurity_split': None,\n",
            " 'min_samples_leaf': 1,\n",
            " 'min_samples_split': 2,\n",
            " 'min_weight_fraction_leaf': 0.0,\n",
            " 'n_estimators': 6,\n",
            " 'n_iter_no_change': None,\n",
            " 'presort': 'deprecated',\n",
            " 'random_state': 0,\n",
            " 'subsample': 1.0,\n",
            " 'tol': 0.0001,\n",
            " 'validation_fraction': 0.1,\n",
            " 'verbose': 0,\n",
            " 'warm_start': False}\n",
            "\n",
            "Exactitud de GradientBoosting en entrenamiento (R2-score): 0.9245556275581821\n",
            "Exactitud de GradientBoosting en validación (R2-score): 0.8826490700669823\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9V6IZass-Z2y",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "outputId": "467717f4-03de-417d-ebb6-9039fae475b7"
      },
      "source": [
        "plt.figure(figsize=(10, 4))\n",
        "\n",
        "plt.plot(errors, \"b.-\")\n",
        "plt.plot([bst_n_estimators, bst_n_estimators], [0, min_error], \"k--\")\n",
        "plt.plot([0, 100], [min_error, min_error], \"k--\")\n",
        "plt.plot(bst_n_estimators, min_error, \"ko\")\n",
        "plt.text(bst_n_estimators, min_error*1.2, \"Mínimo\", ha=\"center\", fontsize=14)\n",
        "plt.xlabel(\"n_estimators\")\n",
        "plt.ylabel(\"Error\", fontsize=16)\n",
        "plt.title(\"Error en el conjunto de validación\", fontsize=14)\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m-Oh26CDBSDp",
        "colab_type": "text"
      },
      "source": [
        "#ADABoost - Regresión"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2qpf2UXVsI0m",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 615
        },
        "outputId": "d7108bca-9e94-4f83-baff-e616c69c0342"
      },
      "source": [
        "import numpy as np\n",
        "from sklearn.ensemble import AdaBoostRegressor\n",
        "from sklearn.tree     import DecisionTreeRegressor\n",
        "from sklearn.datasets import make_regression\n",
        "\n",
        "lr = 0.1\n",
        "ab = AdaBoostRegressor(base_estimator=DecisionTreeRegressor(max_depth=10,criterion='mae'),learning_rate=lr, n_estimators=500,random_state=0)\n",
        "#ab = AdaBoostRegressor(learning_rate=lr, n_estimators=5,random_state=0)\n",
        "ab.fit(X_train, y_train)\n",
        "\n",
        "# Se obtiene el error medio absoluto (MAE) de cada una de las iteraciones (n_estimators)\n",
        "errors = [(median_absolute_error(y_test, y_pred))\n",
        "          for y_pred in ab.staged_predict(X_test)]\n",
        "print(\"Los errores incurridos por los sucesivos ensambles de árboles son:\", errors)\n",
        "\n",
        "bst_n_estimators = np.argmin(errors) + 1   #El mejor modelo es el que tiene menor error absoluto promedio\n",
        "min_error = np.min(errors)                 #Menor valor MAE\n",
        "\n",
        "print(ab.score(X_test,y_test))\n",
        "print('Total de iteraciones: {}'.format(len(errors)))                 #Total de errores, uno por cada ensamble\n",
        "print('Mejor iteración del ADABoost: {}'.format(bst_n_estimators))    #Mejor iteración\n",
        "print('Menor error: {}'.format(errors[bst_n_estimators-1]))           #Imprime el error\n",
        "print('Menor error: {} \\n'.format(min_error))                         #Imprime el error\n",
        "\n",
        "\n",
        "ab_best = AdaBoostRegressor(base_estimator=DecisionTreeRegressor(max_depth=10, criterion='mae'),learning_rate=lr, n_estimators=bst_n_estimators, random_state=0)\n",
        "#ab_best = AdaBoostRegressor(learning_rate=lr, n_estimators=bst_n_estimators, random_state=0)\n",
        "ab_best.fit(X_train, y_train)\n",
        "\n",
        "print('Los hiperparámetros del mejor modelo son:')\n",
        "pprint(ab_best.get_params())\n",
        "print()\n",
        "\n",
        "print('Exactitud de ADABoost en entrenamiento (R2-score):', ab_best.score(X_train, y_train))\n",
        "print('Exactitud de ADABoost en validación (R2-score):', ab_best.score(X_test, y_test))\n",
        "\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Los errores incurridos por los sucesivos ensambles de árboles son: [92.65000000000003, 76.32999999999996, 71.42000000000002, 69.38499999999998, 75.245, 68.55250000000004, 71.33000000000001, 70.21500000000002, 66.62500000000003, 69.97, 65.825, 63.359999999999985, 62.76, 64.06499999999998, 61.0925, 61.97999999999999, 63.155, 63.4, 63.69499999999999, 63.37999999999999, 68.7, 64.36999999999999, 67.23500000000001, 66.62000000000002, 66.65500000000002, 68.73, 66.405, 65.80499999999999, 66.39500000000001, 67.26500000000001, 69.61999999999999, 67.33500000000002, 68.64, 68.405, 64.765, 67.33500000000002, 66.39, 66.65500000000002, 66.65500000000002, 66.65500000000002, 69.195, 63.74249999999999, 68.335, 65.37000000000003, 66.53500000000003, 63.847500000000025, 63.74249999999999, 63.847500000000025, 63.457499999999996, 63.457499999999996, 63.359999999999985, 63.457499999999996, 63.359999999999985, 63.98499999999997, 63.359999999999985, 63.98499999999997, 66.33500000000001, 63.65499999999997, 63.359999999999985, 63.65499999999997, 62.49500000000002, 63.53499999999998, 63.359999999999985, 63.86499999999998, 63.334999999999994, 63.359999999999985, 63.359999999999985, 63.334999999999994, 63.359999999999985, 63.86499999999998, 63.787499999999994, 63.457499999999996, 63.787499999999994, 63.359999999999985, 63.17000000000001, 63.359999999999985, 63.457499999999996, 65.10999999999999, 63.68999999999998, 63.359999999999985, 63.359999999999985, 63.344999999999985, 63.334999999999994, 63.344999999999985, 63.17000000000001, 63.334999999999994, 62.22750000000002, 62.474999999999966, 61.94500000000002, 62.02999999999996, 62.474999999999966, 63.00999999999996, 63.067499999999995, 63.00999999999996, 63.067499999999995, 63.067499999999995, 63.359999999999985, 62.08999999999999, 61.80499999999999, 62.5325, 62.5325, 61.187500000000014, 61.80499999999999, 61.80499999999999, 61.187500000000014, 61.43499999999999, 61.43499999999999, 61.43499999999999, 61.73750000000004, 61.57500000000002, 61.57500000000002, 61.57500000000002, 61.672500000000014, 61.43499999999999, 61.43499999999999, 61.43499999999999, 62.48750000000001, 64.22250000000001, 63.22999999999999, 63.224999999999994, 62.9725, 63.224999999999994, 63.82499999999999, 63.82499999999999, 63.72999999999999, 63.224999999999994, 63.224999999999994, 63.224999999999994, 63.224999999999994, 63.224999999999994, 62.937499999999986, 62.535, 61.769999999999996, 62.58, 62.31999999999999, 62.282499999999985, 62.282499999999985, 61.43499999999999, 62.31999999999999, 61.769999999999996, 62.31999999999999, 61.769999999999996, 61.43499999999999, 62.31999999999999, 62.24499999999999, 62.31999999999999, 61.43499999999999, 62.24499999999999, 61.43499999999999, 61.43499999999999, 61.82249999999999, 62.63249999999999, 63.224999999999994, 63.224999999999994, 63.22999999999999, 64.51499999999999, 66.095, 64.51499999999999, 66.13999999999999, 65.095, 66.91, 66.47999999999999, 67.32000000000001, 67.01, 67.26, 66.47, 66.72, 66.47, 67.26, 66.22999999999999, 66.91, 66.13999999999999, 66.98000000000002, 66.92000000000002, 67.26, 67.26, 66.91, 67.01, 67.465, 66.66, 66.46499999999995, 66.03499999999994, 66.46499999999995, 66.99000000000002, 66.46499999999995, 66.99000000000002, 66.47499999999997, 66.47499999999997, 66.99000000000002, 67.62249999999996, 66.99000000000002, 66.99000000000002, 67.62249999999996, 66.99000000000002, 67.62249999999996, 67.57000000000001, 67.07000000000001, 67.525, 67.96500000000002, 67.07000000000001, 67.32000000000001, 67.32000000000001, 67.8, 66.22999999999999, 67.07000000000001, 67.995, 67.07000000000001, 67.07000000000001, 67.995, 67.995, 67.8, 67.525, 67.525, 67.525, 67.525, 67.525, 67.525, 67.995, 67.8, 67.995, 67.8, 67.8, 66.99000000000002, 67.8, 66.15, 66.99000000000002, 66.99000000000002, 66.22999999999999, 67.87, 68.03999999999999, 67.255, 67.325, 67.94000000000001, 68.305, 68.615, 68.02000000000002, 68.615, 67.675, 68.615, 67.31000000000003, 68.31500000000001, 68.40499999999999, 68.40499999999999, 68.02, 68.31500000000001, 68.31500000000001, 68.31500000000001, 68.02, 68.12, 67.68499999999999, 68.02000000000001, 67.685, 67.685, 67.685, 67.5925, 67.615, 67.685, 68.235, 68.235, 68.235, 67.95000000000002, 67.93, 68.235, 67.95000000000002, 67.95000000000002, 68.00000000000001, 67.93, 67.93, 67.93, 67.93, 67.93, 67.16249999999994, 66.78999999999999, 66.78249999999994, 66.41499999999994, 67.16249999999994, 67.11499999999995, 67.11499999999995, 67.11499999999995, 66.03499999999994, 67.62249999999996, 67.62249999999996, 68.355, 67.62249999999996, 67.89500000000002, 68.095, 68.095, 68.095, 67.69000000000001, 67.205, 67.69000000000001, 66.92999999999999, 67.38000000000007, 67.5275, 67.445, 67.08000000000001, 66.84000000000002, 66.84000000000002, 67.205, 67.08000000000001, 67.08000000000001, 67.5275, 66.0, 66.0, 67.10249999999999, 66.73750000000001, 67.10249999999999, 67.10249999999999, 67.10249999999999, 66.20999999999998, 66.805, 65.91499999999996, 66.805, 65.91499999999996, 65.91499999999996, 65.91, 67.205, 65.91, 67.205, 67.205, 67.445, 67.445, 67.445, 65.54500000000002, 67.445, 65.935, 65.935, 66.08500000000001, 65.935, 65.19, 65.935, 65.935, 65.935, 65.03, 65.935, 65.935, 65.935, 65.03, 65.96999999999994, 65.03, 65.03, 65.03, 65.03, 65.03, 65.03, 65.03, 65.41999999999996, 64.38, 65.03, 65.31999999999994, 65.96999999999994, 65.31999999999994, 65.285, 65.31999999999994, 66.29999999999998, 66.29999999999998, 66.29999999999998, 66.29999999999998, 66.32499999999996, 66.32499999999996, 64.95249999999996, 66.55499999999995, 67.2125, 65.96749999999997, 65.96749999999997, 64.41999999999999, 64.53999999999999, 65.19, 65.19, 65.19, 65.19, 65.19, 65.19, 65.19, 65.19, 65.96999999999994, 65.96999999999994, 65.96999999999994, 65.96999999999994, 65.96999999999994, 66.49999999999994, 65.96999999999994, 65.96999999999994, 65.96999999999994, 65.46249999999995, 66.99499999999995, 66.61500000000001, 66.99499999999995, 65.96999999999994, 66.49999999999994, 65.96999999999994, 66.61500000000001, 66.49999999999994, 67.36000000000001, 67.00500000000005, 66.86500000000001, 66.86500000000001, 65.57999999999996, 65.57999999999996, 65.07249999999996, 65.46249999999995, 65.46249999999995, 67.49499999999995, 66.8549999999999, 65.8524999999999, 66.47499999999997, 66.47499999999997, 68.24249999999995, 68.24249999999995, 68.71499999999999, 67.22249999999997, 66.47499999999997, 66.47499999999997, 66.47499999999997, 66.98000000000008, 66.49999999999994, 65.57999999999996, 66.49999999999994, 65.57999999999996, 66.49999999999994, 66.8549999999999, 67.11499999999995, 67.49499999999995, 67.49499999999995, 67.49499999999995, 67.61000000000001, 68.71499999999999, 66.96999999999997, 67.04999999999995, 66.47499999999997, 66.695, 66.47499999999997, 66.695, 66.40999999999994, 66.49999999999994, 66.49999999999994, 66.49999999999994, 66.47499999999997, 66.47499999999997, 66.49999999999994, 66.55499999999995, 66.49999999999994, 66.96999999999997, 67.30000000000003, 66.8549999999999, 67.18499999999996, 67.93249999999996, 67.86249999999995, 67.6024999999999, 66.05499999999999, 66.67, 66.46499999999995, 68.25, 66.10999999999999, 68.975, 66.99499999999995, 67.98999999999995, 66.05499999999999, 65.96499999999999, 67.24749999999995, 67.54499999999999, 67.63499999999999, 66.05499999999999, 66.33500000000001, 67.62000000000006, 68.71499999999999, 67.62000000000006, 67.94000000000001, 67.87, 68.405, 69.05999999999999, 68.595, 69.05999999999999, 67.62000000000006, 68.335, 68.335, 68.335, 67.87, 68.335, 69.05999999999999, 69.33499999999995, 68.335, 69.33499999999995, 69.05999999999999, 68.405, 68.405, 68.405, 69.05999999999999, 68.00000000000006, 66.6875, 66.6575, 66.6575, 66.04499999999999, 66.04499999999999, 66.04499999999999, 65.19, 65.19, 65.19, 66.04499999999999, 66.04499999999999]\n",
            "0.9274016939563078\n",
            "Total de iteraciones: 500\n",
            "Mejor iteración del ADABoost: 15\n",
            "Menor error: 61.0925\n",
            "Menor error: 61.0925 \n",
            "\n",
            "Los hiperparámetros del mejor modelo son:\n",
            "{'base_estimator': DecisionTreeRegressor(ccp_alpha=0.0, criterion='mae', max_depth=10,\n",
            "                      max_features=None, max_leaf_nodes=None,\n",
            "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
            "                      min_samples_leaf=1, min_samples_split=2,\n",
            "                      min_weight_fraction_leaf=0.0, presort='deprecated',\n",
            "                      random_state=None, splitter='best'),\n",
            " 'base_estimator__ccp_alpha': 0.0,\n",
            " 'base_estimator__criterion': 'mae',\n",
            " 'base_estimator__max_depth': 10,\n",
            " 'base_estimator__max_features': None,\n",
            " 'base_estimator__max_leaf_nodes': None,\n",
            " 'base_estimator__min_impurity_decrease': 0.0,\n",
            " 'base_estimator__min_impurity_split': None,\n",
            " 'base_estimator__min_samples_leaf': 1,\n",
            " 'base_estimator__min_samples_split': 2,\n",
            " 'base_estimator__min_weight_fraction_leaf': 0.0,\n",
            " 'base_estimator__presort': 'deprecated',\n",
            " 'base_estimator__random_state': None,\n",
            " 'base_estimator__splitter': 'best',\n",
            " 'learning_rate': 0.1,\n",
            " 'loss': 'linear',\n",
            " 'n_estimators': 15,\n",
            " 'random_state': 0}\n",
            "\n",
            "Exactitud de ADABoost en entrenamiento (R2-score): 0.995662836626727\n",
            "Exactitud de ADABoost en validación (R2-score): 0.9281238133808499\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w6GUxZtXBVp0",
        "colab_type": "text"
      },
      "source": [
        "#XGBoost - Regresión"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hqGUDTLUBXNo",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "1ff59c36-0a4e-4893-97e1-a9ccf78a8b2c"
      },
      "source": [
        "import xgboost\n",
        "\n",
        "xgb = xgboost.XGBRegressor(n_estimators=200, random_state=0)\n",
        "\n",
        "#Las metricas a evaluar para el entrenamiento iterativo de la regresión serán: MAE y RMSE (https://github.com/dmlc/xgboost/blob/master/doc/parameter.rst)\n",
        "xgb.fit(X_train, y_train,\n",
        "        eval_set=[(X_train, y_train), (X_test, y_test)], \n",
        "        early_stopping_rounds=10, eval_metric=[\"rmse\", \"mae\"])\n",
        "\n",
        "print('Los hiperparámetros del modelo ajustado son:')\n",
        "pprint(xgb.get_params())\n",
        "print()\n",
        "\n",
        "print('Exactitud de XGBoost en entrenamiento:', xgb.score(X_train, y_train))\n",
        "print('Exactitud de XGBoost en validación:', xgb.score(X_test, y_test))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[02:05:34] WARNING: /workspace/src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
            "[0]\tvalidation_0-rmse:1099.97\tvalidation_0-mae:597.592\tvalidation_1-rmse:910.094\tvalidation_1-mae:523.207\n",
            "Multiple eval metrics have been passed: 'validation_1-mae' will be used for early stopping.\n",
            "\n",
            "Will train until validation_1-mae hasn't improved in 10 rounds.\n",
            "[1]\tvalidation_0-rmse:1009.06\tvalidation_0-mae:540.323\tvalidation_1-rmse:834.622\tvalidation_1-mae:471.811\n",
            "[2]\tvalidation_0-rmse:927.305\tvalidation_0-mae:489.317\tvalidation_1-rmse:765.021\tvalidation_1-mae:425.261\n",
            "[3]\tvalidation_0-rmse:853.723\tvalidation_0-mae:443.264\tvalidation_1-rmse:704.176\tvalidation_1-mae:383.875\n",
            "[4]\tvalidation_0-rmse:787.774\tvalidation_0-mae:401.72\tvalidation_1-rmse:648.696\tvalidation_1-mae:345.949\n",
            "[5]\tvalidation_0-rmse:728.619\tvalidation_0-mae:365.563\tvalidation_1-rmse:601.427\tvalidation_1-mae:316.209\n",
            "[6]\tvalidation_0-rmse:675.388\tvalidation_0-mae:334.141\tvalidation_1-rmse:568.639\tvalidation_1-mae:289.849\n",
            "[7]\tvalidation_0-rmse:627.653\tvalidation_0-mae:306.514\tvalidation_1-rmse:540.391\tvalidation_1-mae:268.128\n",
            "[8]\tvalidation_0-rmse:583.854\tvalidation_0-mae:281.667\tvalidation_1-rmse:498.099\tvalidation_1-mae:245.129\n",
            "[9]\tvalidation_0-rmse:545.353\tvalidation_0-mae:261.4\tvalidation_1-rmse:467.629\tvalidation_1-mae:227.089\n",
            "[10]\tvalidation_0-rmse:510.714\tvalidation_0-mae:244.32\tvalidation_1-rmse:437.741\tvalidation_1-mae:210.78\n",
            "[11]\tvalidation_0-rmse:479.951\tvalidation_0-mae:229.073\tvalidation_1-rmse:420.691\tvalidation_1-mae:197.836\n",
            "[12]\tvalidation_0-rmse:452.514\tvalidation_0-mae:216.126\tvalidation_1-rmse:395.801\tvalidation_1-mae:184.263\n",
            "[13]\tvalidation_0-rmse:428.266\tvalidation_0-mae:204.216\tvalidation_1-rmse:374.802\tvalidation_1-mae:173.738\n",
            "[14]\tvalidation_0-rmse:406.795\tvalidation_0-mae:194.136\tvalidation_1-rmse:363.056\tvalidation_1-mae:165.823\n",
            "[15]\tvalidation_0-rmse:386.9\tvalidation_0-mae:185.642\tvalidation_1-rmse:349.07\tvalidation_1-mae:159.281\n",
            "[16]\tvalidation_0-rmse:368.162\tvalidation_0-mae:178.41\tvalidation_1-rmse:335.805\tvalidation_1-mae:154.628\n",
            "[17]\tvalidation_0-rmse:352.663\tvalidation_0-mae:171.815\tvalidation_1-rmse:329.042\tvalidation_1-mae:150.276\n",
            "[18]\tvalidation_0-rmse:338.789\tvalidation_0-mae:166.658\tvalidation_1-rmse:322.612\tvalidation_1-mae:146.99\n",
            "[19]\tvalidation_0-rmse:322.787\tvalidation_0-mae:162.001\tvalidation_1-rmse:312.672\tvalidation_1-mae:144.031\n",
            "[20]\tvalidation_0-rmse:311.454\tvalidation_0-mae:158.031\tvalidation_1-rmse:307.941\tvalidation_1-mae:141.532\n",
            "[21]\tvalidation_0-rmse:298.06\tvalidation_0-mae:154.011\tvalidation_1-rmse:304.082\tvalidation_1-mae:139.548\n",
            "[22]\tvalidation_0-rmse:286.091\tvalidation_0-mae:150.389\tvalidation_1-rmse:296.405\tvalidation_1-mae:137.668\n",
            "[23]\tvalidation_0-rmse:274.905\tvalidation_0-mae:147.583\tvalidation_1-rmse:289.611\tvalidation_1-mae:136.099\n",
            "[24]\tvalidation_0-rmse:264.926\tvalidation_0-mae:144.76\tvalidation_1-rmse:283.588\tvalidation_1-mae:134.631\n",
            "[25]\tvalidation_0-rmse:255.96\tvalidation_0-mae:142.767\tvalidation_1-rmse:278.262\tvalidation_1-mae:133.795\n",
            "[26]\tvalidation_0-rmse:247.964\tvalidation_0-mae:140.422\tvalidation_1-rmse:272.743\tvalidation_1-mae:132.4\n",
            "[27]\tvalidation_0-rmse:240.703\tvalidation_0-mae:138.678\tvalidation_1-rmse:268.288\tvalidation_1-mae:131.696\n",
            "[28]\tvalidation_0-rmse:233.684\tvalidation_0-mae:136.858\tvalidation_1-rmse:264.722\tvalidation_1-mae:131.242\n",
            "[29]\tvalidation_0-rmse:227.754\tvalidation_0-mae:134.877\tvalidation_1-rmse:265.872\tvalidation_1-mae:131.852\n",
            "[30]\tvalidation_0-rmse:222.002\tvalidation_0-mae:133.258\tvalidation_1-rmse:265.074\tvalidation_1-mae:131.578\n",
            "[31]\tvalidation_0-rmse:216.59\tvalidation_0-mae:131.784\tvalidation_1-rmse:262.24\tvalidation_1-mae:131.338\n",
            "[32]\tvalidation_0-rmse:211.916\tvalidation_0-mae:130.217\tvalidation_1-rmse:261.049\tvalidation_1-mae:131.509\n",
            "[33]\tvalidation_0-rmse:207.55\tvalidation_0-mae:129.028\tvalidation_1-rmse:257.752\tvalidation_1-mae:130.815\n",
            "[34]\tvalidation_0-rmse:203.868\tvalidation_0-mae:128.096\tvalidation_1-rmse:257.216\tvalidation_1-mae:131.18\n",
            "[35]\tvalidation_0-rmse:200.107\tvalidation_0-mae:126.475\tvalidation_1-rmse:254.665\tvalidation_1-mae:130.303\n",
            "[36]\tvalidation_0-rmse:196.793\tvalidation_0-mae:125.346\tvalidation_1-rmse:255.771\tvalidation_1-mae:130.864\n",
            "[37]\tvalidation_0-rmse:193.904\tvalidation_0-mae:124.53\tvalidation_1-rmse:255.324\tvalidation_1-mae:130.719\n",
            "[38]\tvalidation_0-rmse:191.119\tvalidation_0-mae:123.499\tvalidation_1-rmse:253.13\tvalidation_1-mae:130.257\n",
            "[39]\tvalidation_0-rmse:188.389\tvalidation_0-mae:122.065\tvalidation_1-rmse:251.391\tvalidation_1-mae:129.662\n",
            "[40]\tvalidation_0-rmse:185.569\tvalidation_0-mae:121.242\tvalidation_1-rmse:249.931\tvalidation_1-mae:129.324\n",
            "[41]\tvalidation_0-rmse:183.007\tvalidation_0-mae:120.587\tvalidation_1-rmse:251.754\tvalidation_1-mae:129.889\n",
            "[42]\tvalidation_0-rmse:180.905\tvalidation_0-mae:119.961\tvalidation_1-rmse:251.522\tvalidation_1-mae:129.658\n",
            "[43]\tvalidation_0-rmse:178.681\tvalidation_0-mae:118.986\tvalidation_1-rmse:250.178\tvalidation_1-mae:129.081\n",
            "[44]\tvalidation_0-rmse:176.853\tvalidation_0-mae:118.251\tvalidation_1-rmse:249.558\tvalidation_1-mae:128.805\n",
            "[45]\tvalidation_0-rmse:175.02\tvalidation_0-mae:116.957\tvalidation_1-rmse:247.906\tvalidation_1-mae:128.051\n",
            "[46]\tvalidation_0-rmse:173.528\tvalidation_0-mae:116.232\tvalidation_1-rmse:247.644\tvalidation_1-mae:127.795\n",
            "[47]\tvalidation_0-rmse:171.77\tvalidation_0-mae:115.516\tvalidation_1-rmse:247.035\tvalidation_1-mae:127.709\n",
            "[48]\tvalidation_0-rmse:169.691\tvalidation_0-mae:114.822\tvalidation_1-rmse:249.672\tvalidation_1-mae:128.497\n",
            "[49]\tvalidation_0-rmse:168.271\tvalidation_0-mae:114.36\tvalidation_1-rmse:248.848\tvalidation_1-mae:128.238\n",
            "[50]\tvalidation_0-rmse:166.576\tvalidation_0-mae:113.791\tvalidation_1-rmse:251.29\tvalidation_1-mae:128.942\n",
            "[51]\tvalidation_0-rmse:165.272\tvalidation_0-mae:112.584\tvalidation_1-rmse:250.747\tvalidation_1-mae:128.018\n",
            "[52]\tvalidation_0-rmse:163.836\tvalidation_0-mae:111.851\tvalidation_1-rmse:249.394\tvalidation_1-mae:127.486\n",
            "[53]\tvalidation_0-rmse:163.032\tvalidation_0-mae:110.858\tvalidation_1-rmse:249.221\tvalidation_1-mae:127.105\n",
            "[54]\tvalidation_0-rmse:161.954\tvalidation_0-mae:110.264\tvalidation_1-rmse:248.85\tvalidation_1-mae:126.96\n",
            "[55]\tvalidation_0-rmse:160.985\tvalidation_0-mae:109.332\tvalidation_1-rmse:248.442\tvalidation_1-mae:126.294\n",
            "[56]\tvalidation_0-rmse:160.271\tvalidation_0-mae:108.778\tvalidation_1-rmse:247.68\tvalidation_1-mae:126.063\n",
            "[57]\tvalidation_0-rmse:159.384\tvalidation_0-mae:108.283\tvalidation_1-rmse:247.266\tvalidation_1-mae:125.82\n",
            "[58]\tvalidation_0-rmse:158.571\tvalidation_0-mae:107.357\tvalidation_1-rmse:246.986\tvalidation_1-mae:125.042\n",
            "[59]\tvalidation_0-rmse:157.734\tvalidation_0-mae:106.94\tvalidation_1-rmse:247.802\tvalidation_1-mae:124.803\n",
            "[60]\tvalidation_0-rmse:157.193\tvalidation_0-mae:106.357\tvalidation_1-rmse:247.749\tvalidation_1-mae:124.634\n",
            "[61]\tvalidation_0-rmse:156.723\tvalidation_0-mae:106.154\tvalidation_1-rmse:247.798\tvalidation_1-mae:124.645\n",
            "[62]\tvalidation_0-rmse:156.227\tvalidation_0-mae:105.455\tvalidation_1-rmse:247.641\tvalidation_1-mae:124.193\n",
            "[63]\tvalidation_0-rmse:155.466\tvalidation_0-mae:104.901\tvalidation_1-rmse:248.124\tvalidation_1-mae:124.236\n",
            "[64]\tvalidation_0-rmse:155.048\tvalidation_0-mae:104.73\tvalidation_1-rmse:248.156\tvalidation_1-mae:124.242\n",
            "[65]\tvalidation_0-rmse:154.384\tvalidation_0-mae:104.451\tvalidation_1-rmse:248.737\tvalidation_1-mae:124.072\n",
            "[66]\tvalidation_0-rmse:153.864\tvalidation_0-mae:104.073\tvalidation_1-rmse:248.636\tvalidation_1-mae:123.98\n",
            "[67]\tvalidation_0-rmse:152.981\tvalidation_0-mae:103.69\tvalidation_1-rmse:249.306\tvalidation_1-mae:124.02\n",
            "[68]\tvalidation_0-rmse:152.421\tvalidation_0-mae:103.39\tvalidation_1-rmse:250.02\tvalidation_1-mae:124.002\n",
            "[69]\tvalidation_0-rmse:151.871\tvalidation_0-mae:102.763\tvalidation_1-rmse:249.856\tvalidation_1-mae:123.455\n",
            "[70]\tvalidation_0-rmse:151.474\tvalidation_0-mae:102.45\tvalidation_1-rmse:249.36\tvalidation_1-mae:123.344\n",
            "[71]\tvalidation_0-rmse:150.957\tvalidation_0-mae:102.076\tvalidation_1-rmse:249.315\tvalidation_1-mae:123.34\n",
            "[72]\tvalidation_0-rmse:150.651\tvalidation_0-mae:101.934\tvalidation_1-rmse:249.325\tvalidation_1-mae:123.34\n",
            "[73]\tvalidation_0-rmse:150.379\tvalidation_0-mae:101.8\tvalidation_1-rmse:249.334\tvalidation_1-mae:123.339\n",
            "[74]\tvalidation_0-rmse:150.058\tvalidation_0-mae:101.498\tvalidation_1-rmse:249.305\tvalidation_1-mae:123.322\n",
            "[75]\tvalidation_0-rmse:149.608\tvalidation_0-mae:101.251\tvalidation_1-rmse:250.051\tvalidation_1-mae:123.356\n",
            "[76]\tvalidation_0-rmse:149.158\tvalidation_0-mae:100.659\tvalidation_1-rmse:249.943\tvalidation_1-mae:122.9\n",
            "[77]\tvalidation_0-rmse:148.288\tvalidation_0-mae:100.215\tvalidation_1-rmse:249.816\tvalidation_1-mae:122.962\n",
            "[78]\tvalidation_0-rmse:147.857\tvalidation_0-mae:99.9251\tvalidation_1-rmse:250.507\tvalidation_1-mae:122.916\n",
            "[79]\tvalidation_0-rmse:147.58\tvalidation_0-mae:99.4879\tvalidation_1-rmse:250.458\tvalidation_1-mae:122.626\n",
            "[80]\tvalidation_0-rmse:146.636\tvalidation_0-mae:99.0161\tvalidation_1-rmse:251.936\tvalidation_1-mae:123.011\n",
            "[81]\tvalidation_0-rmse:146.432\tvalidation_0-mae:98.8859\tvalidation_1-rmse:251.963\tvalidation_1-mae:123.009\n",
            "[82]\tvalidation_0-rmse:146.041\tvalidation_0-mae:98.6213\tvalidation_1-rmse:251.931\tvalidation_1-mae:123.064\n",
            "[83]\tvalidation_0-rmse:145.86\tvalidation_0-mae:98.513\tvalidation_1-rmse:251.935\tvalidation_1-mae:123.061\n",
            "[84]\tvalidation_0-rmse:145.57\tvalidation_0-mae:98.1684\tvalidation_1-rmse:251.903\tvalidation_1-mae:123.066\n",
            "[85]\tvalidation_0-rmse:145.312\tvalidation_0-mae:97.7702\tvalidation_1-rmse:251.805\tvalidation_1-mae:122.634\n",
            "[86]\tvalidation_0-rmse:143.962\tvalidation_0-mae:97.0481\tvalidation_1-rmse:252.593\tvalidation_1-mae:123.018\n",
            "[87]\tvalidation_0-rmse:142.987\tvalidation_0-mae:96.6533\tvalidation_1-rmse:252.738\tvalidation_1-mae:123.115\n",
            "[88]\tvalidation_0-rmse:142.152\tvalidation_0-mae:96.2925\tvalidation_1-rmse:252.635\tvalidation_1-mae:123.001\n",
            "[89]\tvalidation_0-rmse:140.969\tvalidation_0-mae:95.6389\tvalidation_1-rmse:252.671\tvalidation_1-mae:123.015\n",
            "Stopping. Best iteration:\n",
            "[79]\tvalidation_0-rmse:147.58\tvalidation_0-mae:99.4879\tvalidation_1-rmse:250.458\tvalidation_1-mae:122.626\n",
            "\n",
            "Los hiperparámetros del modelo ajustado son:\n",
            "{'base_score': 0.5,\n",
            " 'booster': 'gbtree',\n",
            " 'colsample_bylevel': 1,\n",
            " 'colsample_bynode': 1,\n",
            " 'colsample_bytree': 1,\n",
            " 'gamma': 0,\n",
            " 'importance_type': 'gain',\n",
            " 'learning_rate': 0.1,\n",
            " 'max_delta_step': 0,\n",
            " 'max_depth': 3,\n",
            " 'min_child_weight': 1,\n",
            " 'missing': None,\n",
            " 'n_estimators': 200,\n",
            " 'n_jobs': 1,\n",
            " 'nthread': None,\n",
            " 'objective': 'reg:linear',\n",
            " 'random_state': 0,\n",
            " 'reg_alpha': 0,\n",
            " 'reg_lambda': 1,\n",
            " 'scale_pos_weight': 1,\n",
            " 'seed': None,\n",
            " 'silent': None,\n",
            " 'subsample': 1,\n",
            " 'verbosity': 1}\n",
            "\n",
            "Exactitud de XGBoost en entrenamiento: 0.9783139514352203\n",
            "Exactitud de XGBoost en validación: 0.9036739350100773\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Q4NtfLJXr1hI",
        "colab_type": "text"
      },
      "source": [
        "#Stacking - Regresión"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j2yif8kdr82-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "7d96ddcb-8a86-42fa-87e3-274da09c5bb7"
      },
      "source": [
        "from sklearn.ensemble import StackingRegressor\n",
        "estimators = [('lr_multiple', lr_multiple),\n",
        "              ('lr_lasso', lr_lasso),\n",
        "#               ('adaboost',ab),\n",
        "#              ('gradientboost',gb_best),\n",
        "#             ('xgb',xgb),\n",
        "              ('svr', mejor_svr)]\n",
        "\n",
        "reg = StackingRegressor( estimators      = estimators,\n",
        "                         final_estimator = RandomForestRegressor(n_estimators=17,\n",
        "                                                                 random_state=42))\n",
        "reg.fit(X_train, y_train).score(X_test,y_test)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8831254035796172"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 132
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2sDg-EdLQ2iK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "4bff6a3a-5225-48df-858b-4d4390d2940f"
      },
      "source": [
        "difer=np.abs(reg.predict(X_test) - y_test)  #valor absoluto de las diferencias\n",
        "#difer=reg.predict(X_test) - y_test\n",
        "print(difer.argmax(), difer.iloc[difer.argmax()])\n",
        "print(difer.argmin(), difer.iloc[difer.argmin()])\n",
        "df = pd.DataFrame(data=difer)\n",
        "\n",
        "\n",
        "df.rename(columns={'Importe Total Factura S/.':'Diferencia: Predict - Real'},inplace=True)\n",
        "df.reset_index(drop=True, inplace=True)\n",
        "df.sort_values('Diferencia: Predict - Real', ascending=False, inplace=True)\n",
        "#Total de suma por columna\n",
        "df.loc['Total',:]=df.sum(axis=0)\n",
        "pd.set_option(\"display.max_rows\", None, \"display.max_columns\", None)\n",
        "df.reset_index(inplace=True)\n",
        "df.index = df.index + 1\n",
        "print(df)\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "71 3144.0958823529427\n",
            "137 0.29235294117654576\n",
            "     index  Diferencia: Predict - Real\n",
            "1       71                 3144.095882\n",
            "2       61                 1627.682941\n",
            "3       49                  987.710000\n",
            "4      280                  793.321765\n",
            "5       52                  756.591176\n",
            "6        2                  702.177647\n",
            "7      281                  646.065294\n",
            "8      298                  629.115882\n",
            "9      195                  624.851765\n",
            "10     169                  624.606471\n",
            "11     123                  610.638824\n",
            "12     303                  590.902941\n",
            "13     308                  588.194706\n",
            "14      83                  564.667059\n",
            "15     101                  549.445882\n",
            "16      57                  548.324118\n",
            "17      53                  484.408824\n",
            "18     254                  448.038824\n",
            "19     172                  445.778824\n",
            "20     201                  439.774706\n",
            "21     247                  433.638824\n",
            "22      23                  412.704706\n",
            "23     245                  397.430588\n",
            "24      40                  385.761765\n",
            "25     199                  384.270000\n",
            "26     124                  375.495882\n",
            "27     126                  369.031176\n",
            "28     246                  363.060000\n",
            "29      82                  360.572353\n",
            "30      90                  349.247647\n",
            "31     296                  343.892941\n",
            "32     265                  341.175294\n",
            "33     241                  339.364706\n",
            "34     216                  328.040000\n",
            "35     323                  319.122353\n",
            "36     168                  310.465882\n",
            "37     273                  305.397647\n",
            "38     142                  302.985294\n",
            "39     146                  296.135882\n",
            "40     222                  281.157059\n",
            "41     106                  278.402941\n",
            "42      19                  276.450000\n",
            "43     118                  271.030000\n",
            "44      91                  256.935882\n",
            "45     292                  249.724706\n",
            "46      35                  249.419412\n",
            "47     268                  247.387647\n",
            "48      16                  243.618235\n",
            "49     154                  234.430588\n",
            "50     329                  227.624118\n",
            "51     158                  226.589412\n",
            "52     226                  221.231765\n",
            "53     230                  220.488235\n",
            "54      11                  218.452353\n",
            "55     220                  212.295294\n",
            "56     289                  209.606471\n",
            "57      74                  206.894118\n",
            "58       5                  203.682941\n",
            "59      65                  201.414706\n",
            "60      42                  196.472353\n",
            "61       9                  194.966471\n",
            "62     114                  194.891765\n",
            "63     301                  194.775882\n",
            "64      18                  191.985294\n",
            "65     219                  190.975294\n",
            "66     288                  190.961765\n",
            "67       6                  187.765882\n",
            "68     321                  186.674118\n",
            "69     319                  185.060000\n",
            "70      21                  181.382353\n",
            "71     171                  181.082941\n",
            "72     294                  178.718235\n",
            "73     306                  177.586471\n",
            "74     192                  177.160588\n",
            "75      31                  174.814706\n",
            "76     304                  172.934706\n",
            "77      97                  170.472353\n",
            "78     297                  163.881176\n",
            "79     165                  163.387647\n",
            "80     249                  156.012353\n",
            "81     238                  154.240000\n",
            "82     188                  152.227059\n",
            "83     253                  150.023529\n",
            "84     309                  149.931176\n",
            "85     328                  148.232941\n",
            "86      98                  147.548824\n",
            "87      77                  146.447647\n",
            "88     215                  145.240588\n",
            "89     258                  144.335882\n",
            "90      30                  143.814706\n",
            "91     196                  142.568824\n",
            "92     159                  138.874118\n",
            "93     107                  136.211176\n",
            "94     135                  132.754706\n",
            "95     193                  129.647059\n",
            "96      96                  126.551765\n",
            "97      45                  124.713529\n",
            "98      54                  122.140588\n",
            "99     163                  121.948824\n",
            "100    207                  121.870000\n",
            "101    285                  120.993529\n",
            "102    221                  120.890588\n",
            "103    145                  120.303529\n",
            "104    286                  119.190588\n",
            "105    240                  118.161176\n",
            "106    267                  117.131765\n",
            "107    208                  116.272353\n",
            "108    310                  115.433529\n",
            "109    244                  113.605882\n",
            "110    136                  111.201176\n",
            "111    218                  111.128824\n",
            "112    181                  110.913529\n",
            "113     87                  110.612353\n",
            "114    119                  110.184118\n",
            "115    177                  110.055882\n",
            "116    234                  109.361765\n",
            "117    214                  107.571765\n",
            "118    259                  107.276471\n",
            "119     24                  106.352941\n",
            "120    202                  106.252941\n",
            "121     34                  104.581176\n",
            "122    186                  104.113529\n",
            "123     48                  103.755294\n",
            "124    103                  103.273529\n",
            "125    278                  103.118824\n",
            "126    130                  101.984706\n",
            "127    317                  100.088235\n",
            "128    256                   98.270588\n",
            "129    109                   97.363529\n",
            "130     70                   96.531765\n",
            "131    243                   96.480000\n",
            "132    295                   95.307059\n",
            "133    132                   93.383529\n",
            "134      7                   92.351176\n",
            "135     41                   91.541765\n",
            "136     15                   91.238235\n",
            "137      3                   91.028824\n",
            "138    191                   90.291765\n",
            "139      0                   90.184118\n",
            "140    160                   90.107059\n",
            "141     17                   89.302941\n",
            "142    190                   89.019412\n",
            "143    266                   87.987647\n",
            "144    236                   87.687059\n",
            "145    120                   87.643529\n",
            "146    300                   84.164118\n",
            "147    237                   83.932941\n",
            "148     78                   83.555294\n",
            "149     51                   82.689412\n",
            "150    299                   81.846471\n",
            "151    167                   80.441176\n",
            "152    115                   79.995882\n",
            "153    302                   79.856471\n",
            "154    261                   79.407059\n",
            "155     32                   79.372941\n",
            "156    200                   79.154706\n",
            "157    113                   79.077647\n",
            "158    312                   78.514706\n",
            "159     72                   77.477059\n",
            "160     62                   75.687647\n",
            "161     93                   75.268235\n",
            "162     79                   73.916471\n",
            "163      4                   73.317647\n",
            "164    116                   73.233529\n",
            "165    316                   73.221765\n",
            "166    183                   72.785294\n",
            "167    127                   71.637059\n",
            "168    252                   69.935882\n",
            "169    235                   69.909412\n",
            "170    152                   69.741765\n",
            "171    148                   68.177059\n",
            "172    166                   67.087059\n",
            "173    287                   67.068235\n",
            "174      8                   66.371765\n",
            "175    157                   65.798235\n",
            "176     28                   65.758235\n",
            "177    263                   65.692353\n",
            "178    178                   65.678824\n",
            "179    223                   65.529412\n",
            "180    229                   65.210000\n",
            "181     58                   64.923529\n",
            "182    151                   64.508235\n",
            "183    155                   64.158824\n",
            "184    180                   63.862353\n",
            "185    250                   63.692941\n",
            "186     92                   63.672941\n",
            "187     76                   63.369412\n",
            "188    325                   63.168235\n",
            "189    144                   63.144706\n",
            "190    170                   62.981765\n",
            "191     84                   62.958824\n",
            "192     69                   62.484118\n",
            "193     85                   62.291176\n",
            "194    129                   61.944118\n",
            "195     68                   61.239412\n",
            "196    179                   60.619412\n",
            "197     80                   60.580000\n",
            "198     95                   60.546471\n",
            "199    128                   60.386471\n",
            "200    147                   59.957059\n",
            "201    269                   59.062353\n",
            "202    326                   58.734118\n",
            "203     22                   58.688235\n",
            "204    203                   58.681765\n",
            "205    290                   58.638235\n",
            "206    242                   58.288824\n",
            "207    255                   58.191765\n",
            "208    125                   57.585294\n",
            "209     81                   56.430588\n",
            "210     47                   56.356471\n",
            "211    105                   54.694706\n",
            "212    264                   54.156471\n",
            "213    248                   53.767059\n",
            "214     73                   53.077647\n",
            "215    322                   50.433529\n",
            "216    228                   50.057647\n",
            "217    274                   49.168824\n",
            "218     20                   49.100000\n",
            "219    311                   48.071765\n",
            "220    231                   46.878235\n",
            "221    140                   46.694118\n",
            "222     75                   45.940000\n",
            "223    324                   44.327647\n",
            "224    153                   43.030000\n",
            "225    173                   42.818235\n",
            "226    204                   42.463529\n",
            "227    206                   41.920000\n",
            "228    185                   41.623529\n",
            "229     56                   40.965294\n",
            "230    205                   40.235882\n",
            "231     29                   40.151765\n",
            "232    307                   40.091765\n",
            "233    225                   39.197647\n",
            "234     12                   39.056471\n",
            "235    251                   38.643529\n",
            "236     64                   36.966471\n",
            "237    189                   36.790000\n",
            "238     50                   36.464706\n",
            "239    121                   36.034118\n",
            "240    227                   35.826471\n",
            "241    108                   35.778235\n",
            "242    270                   35.556471\n",
            "243    239                   35.274118\n",
            "244    210                   34.802353\n",
            "245    262                   34.120000\n",
            "246    139                   34.110588\n",
            "247     13                   32.494706\n",
            "248    194                   32.426471\n",
            "249     66                   32.370588\n",
            "250    272                   32.127647\n",
            "251      1                   31.625882\n",
            "252    100                   31.330588\n",
            "253    313                   31.274118\n",
            "254     38                   29.977059\n",
            "255     33                   29.319412\n",
            "256     59                   29.317059\n",
            "257     94                   29.070000\n",
            "258    217                   28.472941\n",
            "259    133                   27.430000\n",
            "260    276                   27.047059\n",
            "261     39                   25.509412\n",
            "262     88                   24.976471\n",
            "263    284                   24.777647\n",
            "264    277                   24.707647\n",
            "265    122                   24.504118\n",
            "266     60                   23.405294\n",
            "267    141                   23.372941\n",
            "268    305                   23.200000\n",
            "269    174                   20.420000\n",
            "270    197                   20.274706\n",
            "271    143                   19.943529\n",
            "272    156                   19.873529\n",
            "273    164                   19.464118\n",
            "274    211                   19.434118\n",
            "275     67                   18.648235\n",
            "276    149                   18.534706\n",
            "277     37                   17.618824\n",
            "278    314                   16.831176\n",
            "279    162                   16.413529\n",
            "280    134                   16.169412\n",
            "281    293                   15.840588\n",
            "282     26                   15.452941\n",
            "283    175                   14.947059\n",
            "284     63                   14.690588\n",
            "285     89                   14.276471\n",
            "286    184                   13.703529\n",
            "287     46                   13.634706\n",
            "288     14                   13.381176\n",
            "289    232                   13.317059\n",
            "290    271                   12.894706\n",
            "291     99                   12.518824\n",
            "292    102                   12.468824\n",
            "293    198                   12.064118\n",
            "294    213                   11.989412\n",
            "295    279                   11.721765\n",
            "296    104                   11.472941\n",
            "297    291                   10.850588\n",
            "298    131                   10.760000\n",
            "299     55                   10.260000\n",
            "300    212                    9.752941\n",
            "301    110                    9.374706\n",
            "302    176                    8.915882\n",
            "303    117                    7.724706\n",
            "304     10                    7.097059\n",
            "305    224                    7.020588\n",
            "306    112                    7.004706\n",
            "307    318                    6.897647\n",
            "308    233                    5.981176\n",
            "309    150                    5.933529\n",
            "310    275                    5.721765\n",
            "311    182                    5.569412\n",
            "312    327                    4.977647\n",
            "313     86                    4.844118\n",
            "314     27                    4.589412\n",
            "315    260                    4.437647\n",
            "316     36                    4.346471\n",
            "317    282                    3.744706\n",
            "318     43                    3.314118\n",
            "319    161                    3.018824\n",
            "320    315                    2.807647\n",
            "321    187                    2.567647\n",
            "322     44                    2.528824\n",
            "323    111                    2.310000\n",
            "324    257                    2.110000\n",
            "325    283                    1.670588\n",
            "326    138                    1.473529\n",
            "327     25                    1.369412\n",
            "328    209                    0.980000\n",
            "329    320                    0.744706\n",
            "330    137                    0.292353\n",
            "331  Total                45669.515882\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "pandas.core.indexes.range.RangeIndex"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 154
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n65zIDwW-53A",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "7d7b052e-7a17-47ea-c411-4f33dc041a75"
      },
      "source": [
        "#y_predict = reg.predict(X_test)\n",
        "print(y_test.sum(axis=0))\n",
        "print(y_predict.sum())\n",
        "print(y_predict[71], np.array(y_test)[71])\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "191807.31\n",
            "202588.65647058823\n",
            "6801.194117647058 9945.29\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fO3U9WUoJJ0I",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "OXQQ0DtRJKUK",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "04dc4bef-628c-456a-8d30-852f77823b05"
      },
      "source": [
        "difer=(reg.predict(X_test) - y_test)  #valor absoluto de las diferencias\n",
        "#difer=reg.predict(X_test) - y_test\n",
        "print(difer.argmax(), difer.iloc[difer.argmax()])\n",
        "print(difer.argmin(), difer.iloc[difer.argmin()])\n",
        "df = pd.DataFrame(data=difer)\n",
        "\n",
        "\n",
        "df.rename(columns={'Importe Total Factura S/.':'Diferencia: Predict - Real'},inplace=True)\n",
        "df.reset_index(drop=True, inplace=True)\n",
        "df.sort_values('Diferencia: Predict - Real', ascending=False, inplace=True)\n",
        "#Total de suma por columna\n",
        "df.loc['Total',:]=df.sum(axis=0)\n",
        "pd.set_option(\"display.max_rows\", None, \"display.max_columns\", None)\n",
        "df.reset_index(inplace=True)\n",
        "df.index = df.index + 1\n",
        "print(df)\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "61 1627.68294117647\n",
            "71 -3144.0958823529427\n",
            "     index  Diferencia: Predict - Real\n",
            "1       61                 1627.682941\n",
            "2       49                  987.710000\n",
            "3       52                  756.591176\n",
            "4        2                  702.177647\n",
            "5      281                  646.065294\n",
            "6      298                  629.115882\n",
            "7      169                  624.606471\n",
            "8      123                  610.638824\n",
            "9      308                  588.194706\n",
            "10      83                  564.667059\n",
            "11     101                  549.445882\n",
            "12     254                  448.038824\n",
            "13     172                  445.778824\n",
            "14     201                  439.774706\n",
            "15      23                  412.704706\n",
            "16     245                  397.430588\n",
            "17     199                  384.270000\n",
            "18     124                  375.495882\n",
            "19      82                  360.572353\n",
            "20      90                  349.247647\n",
            "21     241                  339.364706\n",
            "22     216                  328.040000\n",
            "23     168                  310.465882\n",
            "24     273                  305.397647\n",
            "25     142                  302.985294\n",
            "26      19                  276.450000\n",
            "27     118                  271.030000\n",
            "28      35                  249.419412\n",
            "29     268                  247.387647\n",
            "30      16                  243.618235\n",
            "31     329                  227.624118\n",
            "32     158                  226.589412\n",
            "33     230                  220.488235\n",
            "34      11                  218.452353\n",
            "35     220                  212.295294\n",
            "36     289                  209.606471\n",
            "37      74                  206.894118\n",
            "38       5                  203.682941\n",
            "39      65                  201.414706\n",
            "40      42                  196.472353\n",
            "41       9                  194.966471\n",
            "42     114                  194.891765\n",
            "43      18                  191.985294\n",
            "44     219                  190.975294\n",
            "45     288                  190.961765\n",
            "46       6                  187.765882\n",
            "47     319                  185.060000\n",
            "48      21                  181.382353\n",
            "49     171                  181.082941\n",
            "50     294                  178.718235\n",
            "51     306                  177.586471\n",
            "52     192                  177.160588\n",
            "53     304                  172.934706\n",
            "54      97                  170.472353\n",
            "55     297                  163.881176\n",
            "56     249                  156.012353\n",
            "57     238                  154.240000\n",
            "58     188                  152.227059\n",
            "59     253                  150.023529\n",
            "60     309                  149.931176\n",
            "61     328                  148.232941\n",
            "62      98                  147.548824\n",
            "63      77                  146.447647\n",
            "64      30                  143.814706\n",
            "65     159                  138.874118\n",
            "66     107                  136.211176\n",
            "67     135                  132.754706\n",
            "68      96                  126.551765\n",
            "69      54                  122.140588\n",
            "70     163                  121.948824\n",
            "71     285                  120.993529\n",
            "72     221                  120.890588\n",
            "73     240                  118.161176\n",
            "74     208                  116.272353\n",
            "75     244                  113.605882\n",
            "76     136                  111.201176\n",
            "77     218                  111.128824\n",
            "78     181                  110.913529\n",
            "79      87                  110.612353\n",
            "80     119                  110.184118\n",
            "81     259                  107.276471\n",
            "82      24                  106.352941\n",
            "83     202                  106.252941\n",
            "84     186                  104.113529\n",
            "85     103                  103.273529\n",
            "86     278                  103.118824\n",
            "87     317                  100.088235\n",
            "88     109                   97.363529\n",
            "89      70                   96.531765\n",
            "90       7                   92.351176\n",
            "91      41                   91.541765\n",
            "92      15                   91.238235\n",
            "93       3                   91.028824\n",
            "94     191                   90.291765\n",
            "95     160                   90.107059\n",
            "96      17                   89.302941\n",
            "97     266                   87.987647\n",
            "98     300                   84.164118\n",
            "99     237                   83.932941\n",
            "100     51                   82.689412\n",
            "101    299                   81.846471\n",
            "102    167                   80.441176\n",
            "103     72                   77.477059\n",
            "104     62                   75.687647\n",
            "105      4                   73.317647\n",
            "106    127                   71.637059\n",
            "107    252                   69.935882\n",
            "108    152                   69.741765\n",
            "109    148                   68.177059\n",
            "110      8                   66.371765\n",
            "111    157                   65.798235\n",
            "112     28                   65.758235\n",
            "113    178                   65.678824\n",
            "114    223                   65.529412\n",
            "115    229                   65.210000\n",
            "116    151                   64.508235\n",
            "117    155                   64.158824\n",
            "118    144                   63.144706\n",
            "119    170                   62.981765\n",
            "120     84                   62.958824\n",
            "121    129                   61.944118\n",
            "122    179                   60.619412\n",
            "123     80                   60.580000\n",
            "124    269                   59.062353\n",
            "125    326                   58.734118\n",
            "126     22                   58.688235\n",
            "127    290                   58.638235\n",
            "128    242                   58.288824\n",
            "129    255                   58.191765\n",
            "130    125                   57.585294\n",
            "131     81                   56.430588\n",
            "132    105                   54.694706\n",
            "133     73                   53.077647\n",
            "134    322                   50.433529\n",
            "135    228                   50.057647\n",
            "136    274                   49.168824\n",
            "137     20                   49.100000\n",
            "138    311                   48.071765\n",
            "139     75                   45.940000\n",
            "140    324                   44.327647\n",
            "141    153                   43.030000\n",
            "142    173                   42.818235\n",
            "143    204                   42.463529\n",
            "144    185                   41.623529\n",
            "145     29                   40.151765\n",
            "146     12                   39.056471\n",
            "147    251                   38.643529\n",
            "148    189                   36.790000\n",
            "149    121                   36.034118\n",
            "150    227                   35.826471\n",
            "151    262                   34.120000\n",
            "152    139                   34.110588\n",
            "153    272                   32.127647\n",
            "154    313                   31.274118\n",
            "155     38                   29.977059\n",
            "156    133                   27.430000\n",
            "157    276                   27.047059\n",
            "158    284                   24.777647\n",
            "159    122                   24.504118\n",
            "160     60                   23.405294\n",
            "161    143                   19.943529\n",
            "162     37                   17.618824\n",
            "163    162                   16.413529\n",
            "164     26                   15.452941\n",
            "165    175                   14.947059\n",
            "166     89                   14.276471\n",
            "167    184                   13.703529\n",
            "168     46                   13.634706\n",
            "169     14                   13.381176\n",
            "170    102                   12.468824\n",
            "171    198                   12.064118\n",
            "172    279                   11.721765\n",
            "173    104                   11.472941\n",
            "174    291                   10.850588\n",
            "175    212                    9.752941\n",
            "176    110                    9.374706\n",
            "177    117                    7.724706\n",
            "178     86                    4.844118\n",
            "179     27                    4.589412\n",
            "180     36                    4.346471\n",
            "181    282                    3.744706\n",
            "182     43                    3.314118\n",
            "183    315                    2.807647\n",
            "184    187                    2.567647\n",
            "185    111                    2.310000\n",
            "186    257                    2.110000\n",
            "187    283                    1.670588\n",
            "188    138                    1.473529\n",
            "189    137                   -0.292353\n",
            "190    320                   -0.744706\n",
            "191    209                   -0.980000\n",
            "192     25                   -1.369412\n",
            "193     44                   -2.528824\n",
            "194    161                   -3.018824\n",
            "195    260                   -4.437647\n",
            "196    327                   -4.977647\n",
            "197    182                   -5.569412\n",
            "198    275                   -5.721765\n",
            "199    150                   -5.933529\n",
            "200    233                   -5.981176\n",
            "201    318                   -6.897647\n",
            "202    112                   -7.004706\n",
            "203    224                   -7.020588\n",
            "204     10                   -7.097059\n",
            "205    176                   -8.915882\n",
            "206     55                  -10.260000\n",
            "207    131                  -10.760000\n",
            "208    213                  -11.989412\n",
            "209     99                  -12.518824\n",
            "210    271                  -12.894706\n",
            "211    232                  -13.317059\n",
            "212     63                  -14.690588\n",
            "213    293                  -15.840588\n",
            "214    134                  -16.169412\n",
            "215    314                  -16.831176\n",
            "216    149                  -18.534706\n",
            "217     67                  -18.648235\n",
            "218    211                  -19.434118\n",
            "219    164                  -19.464118\n",
            "220    156                  -19.873529\n",
            "221    197                  -20.274706\n",
            "222    174                  -20.420000\n",
            "223    305                  -23.200000\n",
            "224    141                  -23.372941\n",
            "225    277                  -24.707647\n",
            "226     88                  -24.976471\n",
            "227     39                  -25.509412\n",
            "228    217                  -28.472941\n",
            "229     94                  -29.070000\n",
            "230     59                  -29.317059\n",
            "231     33                  -29.319412\n",
            "232    100                  -31.330588\n",
            "233      1                  -31.625882\n",
            "234     66                  -32.370588\n",
            "235    194                  -32.426471\n",
            "236     13                  -32.494706\n",
            "237    210                  -34.802353\n",
            "238    239                  -35.274118\n",
            "239    270                  -35.556471\n",
            "240    108                  -35.778235\n",
            "241     50                  -36.464706\n",
            "242     64                  -36.966471\n",
            "243    225                  -39.197647\n",
            "244    307                  -40.091765\n",
            "245    205                  -40.235882\n",
            "246     56                  -40.965294\n",
            "247    206                  -41.920000\n",
            "248    140                  -46.694118\n",
            "249    231                  -46.878235\n",
            "250    248                  -53.767059\n",
            "251    264                  -54.156471\n",
            "252     47                  -56.356471\n",
            "253    203                  -58.681765\n",
            "254    147                  -59.957059\n",
            "255    128                  -60.386471\n",
            "256     95                  -60.546471\n",
            "257     68                  -61.239412\n",
            "258     85                  -62.291176\n",
            "259     69                  -62.484118\n",
            "260    325                  -63.168235\n",
            "261     76                  -63.369412\n",
            "262     92                  -63.672941\n",
            "263    250                  -63.692941\n",
            "264    180                  -63.862353\n",
            "265     58                  -64.923529\n",
            "266    263                  -65.692353\n",
            "267    287                  -67.068235\n",
            "268    166                  -67.087059\n",
            "269    235                  -69.909412\n",
            "270    183                  -72.785294\n",
            "271    316                  -73.221765\n",
            "272    116                  -73.233529\n",
            "273     79                  -73.916471\n",
            "274     93                  -75.268235\n",
            "275    312                  -78.514706\n",
            "276    113                  -79.077647\n",
            "277    200                  -79.154706\n",
            "278     32                  -79.372941\n",
            "279    261                  -79.407059\n",
            "280    302                  -79.856471\n",
            "281    115                  -79.995882\n",
            "282     78                  -83.555294\n",
            "283    120                  -87.643529\n",
            "284    236                  -87.687059\n",
            "285    190                  -89.019412\n",
            "286      0                  -90.184118\n",
            "287    132                  -93.383529\n",
            "288    295                  -95.307059\n",
            "289    243                  -96.480000\n",
            "290    256                  -98.270588\n",
            "291    130                 -101.984706\n",
            "292     48                 -103.755294\n",
            "293     34                 -104.581176\n",
            "294    214                 -107.571765\n",
            "295    234                 -109.361765\n",
            "296    177                 -110.055882\n",
            "297    310                 -115.433529\n",
            "298    267                 -117.131765\n",
            "299    286                 -119.190588\n",
            "300    145                 -120.303529\n",
            "301    207                 -121.870000\n",
            "302     45                 -124.713529\n",
            "303    193                 -129.647059\n",
            "304    196                 -142.568824\n",
            "305    258                 -144.335882\n",
            "306    215                 -145.240588\n",
            "307    165                 -163.387647\n",
            "308     31                 -174.814706\n",
            "309    321                 -186.674118\n",
            "310    301                 -194.775882\n",
            "311    226                 -221.231765\n",
            "312    154                 -234.430588\n",
            "313    292                 -249.724706\n",
            "314     91                 -256.935882\n",
            "315    106                 -278.402941\n",
            "316    222                 -281.157059\n",
            "317    146                 -296.135882\n",
            "318    323                 -319.122353\n",
            "319    265                 -341.175294\n",
            "320    296                 -343.892941\n",
            "321    246                 -363.060000\n",
            "322    126                 -369.031176\n",
            "323     40                 -385.761765\n",
            "324    247                 -433.638824\n",
            "325     53                 -484.408824\n",
            "326     57                 -548.324118\n",
            "327    303                 -590.902941\n",
            "328    195                 -624.851765\n",
            "329    280                 -793.321765\n",
            "330     71                -3144.095882\n",
            "331  Total                10781.346471\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EmAHZgX-RKvH",
        "colab_type": "text"
      },
      "source": [
        "#Redes Neuronales - Regresión"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "heqaKq7ORO6t",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "outputId": "709f0316-e75c-4716-d768-1a13bf835946"
      },
      "source": [
        "from sklearn.neural_network import MLPRegressor\n",
        "from sklearn.model_selection import RandomizedSearchCV\n",
        "\n",
        "\n",
        "solver = ['lbfgs','sgd','adam']\n",
        "learning_rate_init = np.linspace(start = 0.01, stop = 10, num = 50)\n",
        "max_iter = [int(x) for x in np.linspace(start = 600, stop = 3000, num = 10)]\n",
        "activation = ['identity', 'logistic', 'tanh', 'relu']\n",
        "\n",
        "#Configuración de búsqueda de mejores hiperparámetros\n",
        "random_grid = {'solver': solver,\n",
        "               'learning_rate_init': learning_rate_init,\n",
        "               'max_iter': max_iter,\n",
        "               'activation': activation}\n",
        "\n",
        "\n",
        "regr_MLP = MLPRegressor(random_state=1)\n",
        "regr_MLP_random = RandomizedSearchCV(estimator = regr_MLP, \n",
        "                                     param_distributions = random_grid, \n",
        "                                     n_iter = 6, \n",
        "                                     cv = 10,          # Validación cruzada 10-fold\n",
        "                                     verbose=2, \n",
        "                                     random_state=0, \n",
        "                                     n_jobs = -1      # Paralelizar en todos los cores disponibles\n",
        "                               )\n",
        "regr_MLP_random.fit(X_train, y_train)               \n",
        "\n",
        "regr_MLP_best = regr_MLP_random.best_estimator_\n",
        "\n",
        "print('Los hiperparámetros del mejor modelo son:')\n",
        "pprint(regr_MLP_random.best_params_)\n",
        "print()\n",
        "print('Exactitud luego de búsqueda aleatoria en validación:', regr_MLP_best.score(X_test, y_test))\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Fitting 10 folds for each of 6 candidates, totalling 60 fits\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 2 concurrent workers.\n",
            "[Parallel(n_jobs=-1)]: Done  37 tasks      | elapsed:  2.0min\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Los hiperparámetros del mejor modelo son:\n",
            "{'activation': 'relu',\n",
            " 'learning_rate_init': 2.2526530612244895,\n",
            " 'max_iter': 3000,\n",
            " 'solver': 'adam'}\n",
            "\n",
            "Exactitud luego de búsqueda aleatoria en validación: 0.5065415714073829\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=-1)]: Done  60 out of  60 | elapsed:  2.2min finished\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NLGflKLn6H5i",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "b7fd1de2-aa91-43ee-f257-4d5578a97ce6"
      },
      "source": [
        ">>> from sklearn import svm, datasets\n",
        ">>> from sklearn.model_selection import GridSearchCV\n",
        ">>> iris = datasets.load_iris()\n",
        ">>> parameters = {'kernel':('linear', 'rbf'), 'C':[1, 10,100]}\n",
        ">>> svc = svm.SVC()\n",
        ">>> clf = GridSearchCV(svc, parameters)\n",
        ">>> clf.fit(iris.data, iris.target)\n",
        ">>> sorted(clf.cv_results_.keys())\n",
        "clf.best_estimator_\n",
        "\n",
        "#clf.cv_results_"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SVC(C=1, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
              "    decision_function_shape='ovr', degree=3, gamma='scale', kernel='linear',\n",
              "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
              "    tol=0.001, verbose=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 127
        }
      ]
    }
  ]
}