XGBoost.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "XGBoost.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0EYCNysPEGcm",
        "outputId": "6a9f3295-ad74-4ff6-e7f9-ceedded91638"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import xgboost as xgb\n",
        "from sklearn.model_selection import RandomizedSearchCV\n",
        "from sklearn.metrics import classification_report, roc_auc_score\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.metrics import confusion_matrix\n",
        "import seaborn as sns\n",
        "\n",
        "from warnings import simplefilter\n",
        "simplefilter(action='ignore', category=FutureWarning)"
      ],
      "metadata": {
        "id": "0ad7h0QYITrE"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train = pd.read_csv(\"/content/drive/MyDrive/ES_Twitch_Dataset/Train_Final.csv\")\n",
        "test = pd.read_csv(\"/content/drive/MyDrive/ES_Twitch_Dataset/Test_Final.csv\")"
      ],
      "metadata": {
        "id": "d3-FTHWHIcHA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x_train = train.drop({'Target', 'Unnamed: 0'}, axis = 1)\n",
        "y_train = train['Target']"
      ],
      "metadata": {
        "id": "iG2Ed1YVIg0B"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x_test = test.drop({'Target', 'Unnamed: 0'}, axis = 1)\n",
        "y_test = test['Target']"
      ],
      "metadata": {
        "id": "VEJUeETuIh-M"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "xgbc = xgb.XGBClassifier(random_state=35)\n",
        "\n",
        "params = {\"objective\" :['binary:logistic', 'reg:logistic'],\n",
        "          \"max_depth\": [4, 5, 6, 10, 15, 20],\n",
        "          \"n_estimators\" : [50, 100, 150],\n",
        "          \"learning_rate\" : [0.1, 0.2, 0.3, 0.4],\n",
        "          \"colsample_bytree\" : [0.1, 0.2, 0.3, 0.4],\n",
        "          \"alpha\" : [5, 10, 15, 20]\n",
        "         }\n",
        "\n",
        "model = RandomizedSearchCV(xgbc, params, random_state=35, n_jobs = 1, n_iter = 15, cv = 5, scoring=\"roc_auc\")"
      ],
      "metadata": {
        "id": "kJjc3eNpIjxr"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Entreno el modelo"
      ],
      "metadata": {
        "id": "W4AHgKblJgBd"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model.fit(x_train, y_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "j0C_i8I1JXuj",
        "outputId": "e45ba3f8-8142-40a2-c76d-289e63ee6238"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomizedSearchCV(cv=5, estimator=XGBClassifier(random_state=35), n_iter=15,\n",
              "                   n_jobs=1,\n",
              "                   param_distributions={'alpha': [5, 10, 15, 20],\n",
              "                                        'colsample_bytree': [0.1, 0.2, 0.3,\n",
              "                                                             0.4],\n",
              "                                        'learning_rate': [0.1, 0.2, 0.3, 0.4],\n",
              "                                        'max_depth': [4, 5, 6, 10, 15, 20],\n",
              "                                        'n_estimators': [50, 100, 150],\n",
              "                                        'objective': ['binary:logistic',\n",
              "                                                      'reg:logistic']},\n",
              "                   random_state=35, scoring='roc_auc')"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Mejores hiperparametros encontrados"
      ],
      "metadata": {
        "id": "YYUTj1IKJkNv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "best_params = model.best_estimator_.get_params()\n",
        "best_params"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9Y94F2OeJiVb",
        "outputId": "9057a871-99a6-4301-d31c-3e1f250ffad3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'alpha': 5,\n",
              " 'base_score': 0.5,\n",
              " 'booster': 'gbtree',\n",
              " 'colsample_bylevel': 1,\n",
              " 'colsample_bynode': 1,\n",
              " 'colsample_bytree': 0.4,\n",
              " 'gamma': 0,\n",
              " 'learning_rate': 0.2,\n",
              " 'max_delta_step': 0,\n",
              " 'max_depth': 5,\n",
              " 'min_child_weight': 1,\n",
              " 'missing': None,\n",
              " 'n_estimators': 50,\n",
              " 'n_jobs': 1,\n",
              " 'nthread': None,\n",
              " 'objective': 'reg:logistic',\n",
              " 'random_state': 35,\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}"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "El Randomized Search ya me devuelve el mejor modelo posible, así que no es necesario volver a entrenar."
      ],
      "metadata": {
        "id": "Mxe8OX3xJpTC"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Scores"
      ],
      "metadata": {
        "id": "A7NkNUoLJqr2"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Calculo el score para el set de entrenamiento"
      ],
      "metadata": {
        "id": "SVjcNicCJsv7"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "roc_auc_score(y_train, model.predict_proba(x_train)[:,1])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Y_KDyriLJmHe",
        "outputId": "f412eae9-9a6c-43ed-ae95-36e70e016b2c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9666420980602399"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Calculo el score para el set de test:"
      ],
      "metadata": {
        "id": "wVawLeVKJuE8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "roc_auc_score(y_test, model.predict_proba(x_test)[:,1])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "u5SHdN5uJxIK",
        "outputId": "bfc76c1d-5ef8-401e-ebba-f981d89821cd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8916735406587499"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Matriz de confusion"
      ],
      "metadata": {
        "id": "dcI6UuvhJ1Ko"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "C = confusion_matrix(y_test, model.predict(x_test))  \n",
        "plt.figure(figsize=(15,8))\n",
        "labels = [0,1]\n",
        "sns.heatmap(C, annot=True, cmap=\"Blues\", fmt=\"d\", xticklabels=labels, yticklabels=labels, annot_kws={'size': 15})\n",
        "plt.xlabel('Predicted Values', fontsize=18, fontweight='bold')\n",
        "plt.ylabel('Original Values', fontsize=18, fontweight='bold')\n",
        "plt.title(\"Confusion matrix\", fontsize = 18, fontweight='bold')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 542
        },
        "id": "H2OFA1l7J05U",
        "outputId": "24ff3681-acad-4947-a2a0-8e7bec84a64f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Confusion matrix')"
            ]
          },
          "metadata": {},
          "execution_count": 13
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1080x576 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Features mas importantes"
      ],
      "metadata": {
        "id": "Wi5ZfkdBKEzG"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "xgbc = xgb.XGBClassifier(**best_params)\n",
        "xgbc.fit(x_train, y_train)\n",
        "importances = xgbc.feature_importances_\n",
        "columns_names = list(x_train.columns.values)\n",
        "dic = {}\n",
        "\n",
        "for i in range(len(importances)):\n",
        "  dic[columns_names[i]] = importances[i]\n",
        "\n",
        "df = pd.DataFrame([[key, dic[key]] for key in dic.keys()], columns=['Feature', 'Importancia'])\n",
        "df = df.sort_values('Importancia',ascending=False)\n",
        "df.set_index('Feature',inplace = True)\n",
        "\n",
        "df.iloc[0:10].plot.bar(logy=True, figsize=(15, 7), rot=60, color='lightpink', fontsize=15)\n",
        "plt.title('TOP 10: Importancia de features basado en impurezas', fontsize = 18, fontweight='bold')\n",
        "plt.xlabel('Feature', fontsize=18, fontweight='bold')\n",
        "plt.ylabel('Importancia', fontsize=18, fontweight='bold')\n",
        "plt.yscale(value='linear')\n",
        "plt.legend(('Feature',), loc='upper right', prop={'size':15})\n",
        "\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 612
        },
        "id": "nSBLKSXcJ5yM",
        "outputId": "aa7051a3-f856-468f-8926-7726e6609fb9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1080x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}