Census-Embeddings-MLP.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "aula1.ipynb",
      "provenance": [],
      "mount_file_id": "19X57kipvSICNTJEO8e6ux3pLAZq1TrNQ",
      "authorship_tag": "ABX9TyPw2Pd66qhEdL3y+l+8cFGi",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/finardi/tutos/blob/master/Census-Embeddings-MLP.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ljhWsSxBCW1F",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 357
        },
        "outputId": "cf26fed2-b871-4fe7-8cfc-543b68b3e1d6"
      },
      "source": [
        "! nvidia-smi"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Sun Jul 12 12:39:21 2020       \n",
            "+-----------------------------------------------------------------------------+\n",
            "| NVIDIA-SMI 450.36.06    Driver Version: 418.67       CUDA Version: 10.1     |\n",
            "|-------------------------------+----------------------+----------------------+\n",
            "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
            "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
            "|                               |                      |               MIG M. |\n",
            "|===============================+======================+======================|\n",
            "|   0  Tesla K80           Off  | 00000000:00:04.0 Off |                    0 |\n",
            "| N/A   33C    P8    29W / 149W |      0MiB / 11441MiB |      0%      Default |\n",
            "|                               |                      |                 ERR! |\n",
            "+-------------------------------+----------------------+----------------------+\n",
            "                                                                               \n",
            "+-----------------------------------------------------------------------------+\n",
            "| Processes:                                                                  |\n",
            "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
            "|        ID   ID                                                   Usage      |\n",
            "|=============================================================================|\n",
            "|  No running processes found                                                 |\n",
            "+-----------------------------------------------------------------------------+\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DGGzrVrhDkwK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "1239c348-1d31-4531-fdcb-d918bda86755"
      },
      "source": [
        "# Basics\n",
        "import os\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "\n",
        "# PyTorch\n",
        "import torch \n",
        "from torch import nn\n",
        "import torch.nn.functional as F\n",
        "from torch.utils.data import Dataset, DataLoader\n",
        "\n",
        "# Sklearn\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import roc_auc_score\n",
        "from sklearn.metrics import roc_curve\n",
        "from sklearn.metrics import auc\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.metrics import precision_recall_curve\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "from sklearn.preprocessing import quantile_transform\n",
        "\n",
        "# Scipy\n",
        "from scipy import stats"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
            "  import pandas.util.testing as tm\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ExCYVPc1CZlH",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "1e2f4884-118c-410e-a81f-5c66a0bff96a"
      },
      "source": [
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "manual_seed = 2357\n",
        "\n",
        "def deterministic(rep=True):\n",
        "    if rep:\n",
        "        np.random.seed(manual_seed)\n",
        "        torch.manual_seed(manual_seed)\n",
        "        if torch.cuda.is_available():\n",
        "            torch.cuda.manual_seed(manual_seed)\n",
        "            torch.cuda.manual_seed_all(manual_seed)\n",
        "        torch.backends.cudnn.enabled = False \n",
        "        torch.backends.cudnn.benchmark = False\n",
        "        torch.backends.cudnn.deterministic = True\n",
        "        print(f'Deterministic experiment, seed: {manual_seed}')\n",
        "    else:\n",
        "        print('Random experiment')\n",
        "\n",
        "deterministic()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Deterministic experiment, seed: 2357\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gwVVI3JADtbh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "try:\n",
        "    df = pd.read_csv('/content/CensusData.csv');\n",
        "except:\n",
        "    !unzip '/content/drive/My Drive/Colab Notebooks/Tabular/36093_54253_bundle_archive.zip'\n",
        "    df = pd.read_csv('/content/CensusData.csv');"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Cc_N4JIrD1UN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 419
        },
        "outputId": "961c0a3b-9ed7-4090-c3e0-acdd845b2ae9"
      },
      "source": [
        "df"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Age</th>\n",
              "      <th>workclass</th>\n",
              "      <th>fnlwgt</th>\n",
              "      <th>education</th>\n",
              "      <th>education-num</th>\n",
              "      <th>Marital-status</th>\n",
              "      <th>Occupation</th>\n",
              "      <th>Relationship</th>\n",
              "      <th>Race</th>\n",
              "      <th>Sex</th>\n",
              "      <th>capital-gain</th>\n",
              "      <th>capital-loss</th>\n",
              "      <th>hours-per-week</th>\n",
              "      <th>native-country</th>\n",
              "      <th>Income</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>39</td>\n",
              "      <td>State-gov</td>\n",
              "      <td>77516</td>\n",
              "      <td>Bachelors</td>\n",
              "      <td>13</td>\n",
              "      <td>Never-married</td>\n",
              "      <td>Adm-clerical</td>\n",
              "      <td>Not-in-family</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>2174</td>\n",
              "      <td>0</td>\n",
              "      <td>40</td>\n",
              "      <td>United-States</td>\n",
              "      <td>&lt;=50K</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>50</td>\n",
              "      <td>Self-emp-not-inc</td>\n",
              "      <td>83311</td>\n",
              "      <td>Bachelors</td>\n",
              "      <td>13</td>\n",
              "      <td>Married-civ-spouse</td>\n",
              "      <td>Exec-managerial</td>\n",
              "      <td>Husband</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>13</td>\n",
              "      <td>United-States</td>\n",
              "      <td>&lt;=50K</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>38</td>\n",
              "      <td>Private</td>\n",
              "      <td>215646</td>\n",
              "      <td>HS-grad</td>\n",
              "      <td>9</td>\n",
              "      <td>Divorced</td>\n",
              "      <td>Handlers-cleaners</td>\n",
              "      <td>Not-in-family</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>40</td>\n",
              "      <td>United-States</td>\n",
              "      <td>&lt;=50K</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>53</td>\n",
              "      <td>Private</td>\n",
              "      <td>234721</td>\n",
              "      <td>11th</td>\n",
              "      <td>7</td>\n",
              "      <td>Married-civ-spouse</td>\n",
              "      <td>Handlers-cleaners</td>\n",
              "      <td>Husband</td>\n",
              "      <td>Black</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>40</td>\n",
              "      <td>United-States</td>\n",
              "      <td>&lt;=50K</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>28</td>\n",
              "      <td>Private</td>\n",
              "      <td>338409</td>\n",
              "      <td>Bachelors</td>\n",
              "      <td>13</td>\n",
              "      <td>Married-civ-spouse</td>\n",
              "      <td>Prof-specialty</td>\n",
              "      <td>Wife</td>\n",
              "      <td>Black</td>\n",
              "      <td>Female</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>40</td>\n",
              "      <td>Cuba</td>\n",
              "      <td>&lt;=50K</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",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32556</th>\n",
              "      <td>27</td>\n",
              "      <td>Private</td>\n",
              "      <td>257302</td>\n",
              "      <td>Assoc-acdm</td>\n",
              "      <td>12</td>\n",
              "      <td>Married-civ-spouse</td>\n",
              "      <td>Tech-support</td>\n",
              "      <td>Wife</td>\n",
              "      <td>White</td>\n",
              "      <td>Female</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>38</td>\n",
              "      <td>United-States</td>\n",
              "      <td>&lt;=50K</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32557</th>\n",
              "      <td>40</td>\n",
              "      <td>Private</td>\n",
              "      <td>154374</td>\n",
              "      <td>HS-grad</td>\n",
              "      <td>9</td>\n",
              "      <td>Married-civ-spouse</td>\n",
              "      <td>Machine-op-inspct</td>\n",
              "      <td>Husband</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>40</td>\n",
              "      <td>United-States</td>\n",
              "      <td>&gt;50K</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32558</th>\n",
              "      <td>58</td>\n",
              "      <td>Private</td>\n",
              "      <td>151910</td>\n",
              "      <td>HS-grad</td>\n",
              "      <td>9</td>\n",
              "      <td>Widowed</td>\n",
              "      <td>Adm-clerical</td>\n",
              "      <td>Unmarried</td>\n",
              "      <td>White</td>\n",
              "      <td>Female</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>40</td>\n",
              "      <td>United-States</td>\n",
              "      <td>&lt;=50K</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32559</th>\n",
              "      <td>22</td>\n",
              "      <td>Private</td>\n",
              "      <td>201490</td>\n",
              "      <td>HS-grad</td>\n",
              "      <td>9</td>\n",
              "      <td>Never-married</td>\n",
              "      <td>Adm-clerical</td>\n",
              "      <td>Own-child</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>20</td>\n",
              "      <td>United-States</td>\n",
              "      <td>&lt;=50K</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32560</th>\n",
              "      <td>52</td>\n",
              "      <td>Self-emp-inc</td>\n",
              "      <td>287927</td>\n",
              "      <td>HS-grad</td>\n",
              "      <td>9</td>\n",
              "      <td>Married-civ-spouse</td>\n",
              "      <td>Exec-managerial</td>\n",
              "      <td>Wife</td>\n",
              "      <td>White</td>\n",
              "      <td>Female</td>\n",
              "      <td>15024</td>\n",
              "      <td>0</td>\n",
              "      <td>40</td>\n",
              "      <td>United-States</td>\n",
              "      <td>&gt;50K</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>32561 rows × 15 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "       Age          workclass  fnlwgt  ... hours-per-week  native-country  Income\n",
              "0       39          State-gov   77516  ...             40   United-States   <=50K\n",
              "1       50   Self-emp-not-inc   83311  ...             13   United-States   <=50K\n",
              "2       38            Private  215646  ...             40   United-States   <=50K\n",
              "3       53            Private  234721  ...             40   United-States   <=50K\n",
              "4       28            Private  338409  ...             40            Cuba   <=50K\n",
              "...    ...                ...     ...  ...            ...             ...     ...\n",
              "32556   27            Private  257302  ...             38   United-States   <=50K\n",
              "32557   40            Private  154374  ...             40   United-States    >50K\n",
              "32558   58            Private  151910  ...             40   United-States   <=50K\n",
              "32559   22            Private  201490  ...             20   United-States   <=50K\n",
              "32560   52       Self-emp-inc  287927  ...             40   United-States    >50K\n",
              "\n",
              "[32561 rows x 15 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h_4qDj3w7tB5",
        "colab_type": "text"
      },
      "source": [
        "### Troca do Label para numérico"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5X2KIjyy7whu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "44ebe435-fc36-4137-c856-a5430c1fe680"
      },
      "source": [
        "# Verificar a contagem antes\n",
        "df.Income.value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              " <=50K    24720\n",
              " >50K      7841\n",
              "Name: Income, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1TZ22KPWgBC5",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "d5eeab81-37dd-4899-f12e-227fd0c12a3f"
      },
      "source": [
        "# Verificar a contagem depois\n",
        "df['Income'] = np.where(df['Income'] == ' <=50K', 0,1)\n",
        "df.Income.value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    24720\n",
              "1     7841\n",
              "Name: Income, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "W5FFvnYe79jx",
        "colab_type": "text"
      },
      "source": [
        "### Fazer split entre train e test"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gz1e4bvIZDNs",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "6f6436c7-5787-4486-9849-22ca33a12e85"
      },
      "source": [
        "df_train, df_test, y_train, y_test = train_test_split(df.iloc[:, :-1],df.Income, test_size=0.3)\n",
        "\n",
        "# volta para dataframe\n",
        "df_train = pd.DataFrame(df_train)\n",
        "df_test = pd.DataFrame(df_test)\n",
        "df_train.shape, y_train.shape, df_test.shape, y_test.shape"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((22792, 14), (22792,), (9769, 14), (9769,))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IrdPFgURgLaL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 419
        },
        "outputId": "08ab0016-cb55-4e05-e33a-907c20501aa6"
      },
      "source": [
        "df_train"
      ],
      "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>Age</th>\n",
              "      <th>workclass</th>\n",
              "      <th>fnlwgt</th>\n",
              "      <th>education</th>\n",
              "      <th>education-num</th>\n",
              "      <th>Marital-status</th>\n",
              "      <th>Occupation</th>\n",
              "      <th>Relationship</th>\n",
              "      <th>Race</th>\n",
              "      <th>Sex</th>\n",
              "      <th>capital-gain</th>\n",
              "      <th>capital-loss</th>\n",
              "      <th>hours-per-week</th>\n",
              "      <th>native-country</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>18991</th>\n",
              "      <td>24</td>\n",
              "      <td>Private</td>\n",
              "      <td>276851</td>\n",
              "      <td>HS-grad</td>\n",
              "      <td>9</td>\n",
              "      <td>Divorced</td>\n",
              "      <td>Protective-serv</td>\n",
              "      <td>Own-child</td>\n",
              "      <td>White</td>\n",
              "      <td>Female</td>\n",
              "      <td>0</td>\n",
              "      <td>1762</td>\n",
              "      <td>40</td>\n",
              "      <td>United-States</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9743</th>\n",
              "      <td>33</td>\n",
              "      <td>Self-emp-not-inc</td>\n",
              "      <td>193246</td>\n",
              "      <td>HS-grad</td>\n",
              "      <td>9</td>\n",
              "      <td>Never-married</td>\n",
              "      <td>Exec-managerial</td>\n",
              "      <td>Not-in-family</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>55</td>\n",
              "      <td>France</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>23417</th>\n",
              "      <td>50</td>\n",
              "      <td>Private</td>\n",
              "      <td>197322</td>\n",
              "      <td>11th</td>\n",
              "      <td>7</td>\n",
              "      <td>Married-civ-spouse</td>\n",
              "      <td>Exec-managerial</td>\n",
              "      <td>Husband</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>50</td>\n",
              "      <td>United-States</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25248</th>\n",
              "      <td>45</td>\n",
              "      <td>Private</td>\n",
              "      <td>75673</td>\n",
              "      <td>Assoc-voc</td>\n",
              "      <td>11</td>\n",
              "      <td>Widowed</td>\n",
              "      <td>Adm-clerical</td>\n",
              "      <td>Not-in-family</td>\n",
              "      <td>Asian-Pac-Islander</td>\n",
              "      <td>Female</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>40</td>\n",
              "      <td>United-States</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12892</th>\n",
              "      <td>27</td>\n",
              "      <td>Private</td>\n",
              "      <td>208725</td>\n",
              "      <td>Assoc-acdm</td>\n",
              "      <td>12</td>\n",
              "      <td>Never-married</td>\n",
              "      <td>Sales</td>\n",
              "      <td>Not-in-family</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>42</td>\n",
              "      <td>United-States</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>16432</th>\n",
              "      <td>18</td>\n",
              "      <td>Private</td>\n",
              "      <td>267965</td>\n",
              "      <td>11th</td>\n",
              "      <td>7</td>\n",
              "      <td>Never-married</td>\n",
              "      <td>Sales</td>\n",
              "      <td>Not-in-family</td>\n",
              "      <td>White</td>\n",
              "      <td>Female</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>15</td>\n",
              "      <td>United-States</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25934</th>\n",
              "      <td>47</td>\n",
              "      <td>Private</td>\n",
              "      <td>503454</td>\n",
              "      <td>12th</td>\n",
              "      <td>8</td>\n",
              "      <td>Never-married</td>\n",
              "      <td>Adm-clerical</td>\n",
              "      <td>Other-relative</td>\n",
              "      <td>Black</td>\n",
              "      <td>Female</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>40</td>\n",
              "      <td>United-States</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3235</th>\n",
              "      <td>36</td>\n",
              "      <td>Self-emp-not-inc</td>\n",
              "      <td>367020</td>\n",
              "      <td>7th-8th</td>\n",
              "      <td>4</td>\n",
              "      <td>Married-civ-spouse</td>\n",
              "      <td>Craft-repair</td>\n",
              "      <td>Husband</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>35</td>\n",
              "      <td>United-States</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16377</th>\n",
              "      <td>54</td>\n",
              "      <td>Self-emp-not-inc</td>\n",
              "      <td>154785</td>\n",
              "      <td>Bachelors</td>\n",
              "      <td>13</td>\n",
              "      <td>Divorced</td>\n",
              "      <td>Exec-managerial</td>\n",
              "      <td>Not-in-family</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>50</td>\n",
              "      <td>United-States</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20271</th>\n",
              "      <td>41</td>\n",
              "      <td>?</td>\n",
              "      <td>206916</td>\n",
              "      <td>Bachelors</td>\n",
              "      <td>13</td>\n",
              "      <td>Married-spouse-absent</td>\n",
              "      <td>?</td>\n",
              "      <td>Not-in-family</td>\n",
              "      <td>White</td>\n",
              "      <td>Male</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>30</td>\n",
              "      <td>United-States</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>22792 rows × 14 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "       Age          workclass  ...  hours-per-week  native-country\n",
              "18991   24            Private  ...              40   United-States\n",
              "9743    33   Self-emp-not-inc  ...              55          France\n",
              "23417   50            Private  ...              50   United-States\n",
              "25248   45            Private  ...              40   United-States\n",
              "12892   27            Private  ...              42   United-States\n",
              "...    ...                ...  ...             ...             ...\n",
              "16432   18            Private  ...              15   United-States\n",
              "25934   47            Private  ...              40   United-States\n",
              "3235    36   Self-emp-not-inc  ...              35   United-States\n",
              "16377   54   Self-emp-not-inc  ...              50   United-States\n",
              "20271   41                  ?  ...              30   United-States\n",
              "\n",
              "[22792 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cRcDi2618Rus",
        "colab_type": "text"
      },
      "source": [
        "### Tratar colunas categóricas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fhkgXstZEJpf",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 289
        },
        "outputId": "46dc940b-5a14-439a-d6d7-1424ea4b25d1"
      },
      "source": [
        "df.dtypes"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Age                int64\n",
              "workclass         object\n",
              "fnlwgt             int64\n",
              "education         object\n",
              "education-num      int64\n",
              "Marital-status    object\n",
              "Occupation        object\n",
              "Relationship      object\n",
              "Race              object\n",
              "Sex               object\n",
              "capital-gain       int64\n",
              "capital-loss       int64\n",
              "hours-per-week     int64\n",
              "native-country    object\n",
              "Income             int64\n",
              "dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OjDb3NOhJJxU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 272
        },
        "outputId": "16658286-9bd1-4b4b-c932-c96d06aa7e73"
      },
      "source": [
        "columns   = df.columns.to_list() # todas as colunas\n",
        "col_types = df.dtypes.to_list()  # todos os tipos de colunas\n",
        "col_types"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[dtype('int64'),\n",
              " dtype('O'),\n",
              " dtype('int64'),\n",
              " dtype('O'),\n",
              " dtype('int64'),\n",
              " dtype('O'),\n",
              " dtype('O'),\n",
              " dtype('O'),\n",
              " dtype('O'),\n",
              " dtype('O'),\n",
              " dtype('int64'),\n",
              " dtype('int64'),\n",
              " dtype('int64'),\n",
              " dtype('O'),\n",
              " dtype('int64')]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iI1ekW8qLXGK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "e72a3ccf-e671-4d02-829a-d33da4fad5e9"
      },
      "source": [
        "num_cols, cat_cols = [], []\n",
        "for e in zip(columns, col_types):\n",
        "    if e[1] == 'object':\n",
        "        cat_cols.append(e[0])\n",
        "    else:\n",
        "        num_cols.append(e[0])\n",
        "\n",
        "# excluit o target Income da num_cols\n",
        "num_cols.pop(-1)\n",
        "\n",
        "print(cat_cols, '\\n')\n",
        "print(num_cols)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['workclass', 'education', 'Marital-status', 'Occupation', 'Relationship', 'Race', 'Sex', 'native-country'] \n",
            "\n",
            "['Age', 'fnlwgt', 'education-num', 'capital-gain', 'capital-loss', 'hours-per-week']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iioFiC42TAG1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def map_categorical_train(df_train, cat_cols):\n",
        "    df = df_train.copy()\n",
        "    cat_dims, list_of_dicts = [], []\n",
        "    \n",
        "    for col in cat_cols:\n",
        "        j = 1\n",
        "        my_dict = {}\n",
        "        my_dict.update({' ?':0})\n",
        "        my_dict.update({'UNK':1})\n",
        "\n",
        "        for e in df[col].value_counts().index.to_list():\n",
        "            if e not in my_dict.keys():\n",
        "                j +=1\n",
        "                my_dict.update({e:j})\n",
        "        list_of_dicts.append(my_dict)\n",
        "        cat_dims.append(len(my_dict))\n",
        "\n",
        "        df[col] = df[col].apply(lambda x: my_dict[x] if x in \n",
        "                                my_dict.keys() else my_dict['UNK'])\n",
        "        \n",
        "        emb_szs = [(c, min(100, (c+10)//2)) for c in cat_dims]\n",
        "        cat_emb_dims = [e[0] for e in emb_szs]\n",
        "\n",
        "    return df, list_of_dicts, emb_szs, cat_dims\n",
        "\n",
        "#----------------------------------------------------------------------------------\n",
        "def map_categorical_test(df_test, list_of_dicts):\n",
        "    df = df_test.copy()\n",
        "    for i, col in enumerate(cat_cols):\n",
        "        df[col] = df[col].apply(lambda x: list_of_dicts[i][x] if x in \n",
        "                                list_of_dicts[i].keys() else list_of_dicts[i]['UNK'])\n",
        "    return df"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OjSvRFo6W8G_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "dc5a5f02-07b8-49d8-c56c-2ede80a5f12f"
      },
      "source": [
        "df_train_, list_of_dicts, emb_szs, cat_dims = map_categorical_train(df_train, cat_cols)   \n",
        "df_test_ = map_categorical_test(df_test, list_of_dicts)   \n",
        "print(df_train_.shape, df_test_.shape)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(22792, 14) (9769, 14)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OTVPO_c79GZi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        },
        "outputId": "cb9d8c26-1d18-434e-9c0d-1149ee835e3a"
      },
      "source": [
        "# Conferindo\n",
        "df_train_.head(5)"
      ],
      "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>Age</th>\n",
              "      <th>workclass</th>\n",
              "      <th>fnlwgt</th>\n",
              "      <th>education</th>\n",
              "      <th>education-num</th>\n",
              "      <th>Marital-status</th>\n",
              "      <th>Occupation</th>\n",
              "      <th>Relationship</th>\n",
              "      <th>Race</th>\n",
              "      <th>Sex</th>\n",
              "      <th>capital-gain</th>\n",
              "      <th>capital-loss</th>\n",
              "      <th>hours-per-week</th>\n",
              "      <th>native-country</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>18991</th>\n",
              "      <td>24</td>\n",
              "      <td>2</td>\n",
              "      <td>276851</td>\n",
              "      <td>2</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>13</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>1762</td>\n",
              "      <td>40</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9743</th>\n",
              "      <td>33</td>\n",
              "      <td>3</td>\n",
              "      <td>193246</td>\n",
              "      <td>2</td>\n",
              "      <td>9</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>55</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>23417</th>\n",
              "      <td>50</td>\n",
              "      <td>2</td>\n",
              "      <td>197322</td>\n",
              "      <td>7</td>\n",
              "      <td>7</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>50</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25248</th>\n",
              "      <td>45</td>\n",
              "      <td>2</td>\n",
              "      <td>75673</td>\n",
              "      <td>6</td>\n",
              "      <td>11</td>\n",
              "      <td>6</td>\n",
              "      <td>5</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>40</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12892</th>\n",
              "      <td>27</td>\n",
              "      <td>2</td>\n",
              "      <td>208725</td>\n",
              "      <td>8</td>\n",
              "      <td>12</td>\n",
              "      <td>3</td>\n",
              "      <td>6</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>42</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       Age  workclass  fnlwgt  ...  capital-loss  hours-per-week  native-country\n",
              "18991   24          2  276851  ...          1762              40               2\n",
              "9743    33          3  193246  ...             0              55              30\n",
              "23417   50          2  197322  ...             0              50               2\n",
              "25248   45          2   75673  ...             0              40               2\n",
              "12892   27          2  208725  ...             0              42               2\n",
              "\n",
              "[5 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 35
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "muLbz4Xa9KpR",
        "colab_type": "text"
      },
      "source": [
        "### Tratamento colunas numéricas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RKAZcUE59Nau",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def normalize(df, num_cols, cat_cols, norm='minmax'):\n",
        "    data = df.copy()\n",
        "    df_num = df[num_cols]\n",
        "    \n",
        "    if norm == 'minmax':\n",
        "        m_m_s = MinMaxScaler()\n",
        "        mim_max_scal = m_m_s.fit(df_num.values)\n",
        "        mim_max_scal = m_m_s.transform(df_num.values)\n",
        "        df_norm = pd.DataFrame(data=mim_max_scal, \n",
        "                        columns=df_num.columns, \n",
        "                        index=df_num.index)\n",
        "\n",
        "    if norm == 'quantile':\n",
        "        quantile = quantile_transform(df_num.values, \n",
        "                        random_state=2357, \n",
        "                        copy=True, \n",
        "                        output_distribution='uniform')\n",
        "        df_norm  = pd.DataFrame(data=quantile, \n",
        "                        columns=df_num.columns, \n",
        "                        index=df_num.index)\n",
        "\n",
        "    df= pd.concat((\n",
        "        df_norm, \n",
        "        df[cat_cols]), \n",
        "        axis=1)\n",
        "    return df"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BtS4IlOY-HJB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        },
        "outputId": "e0ab8ff4-e899-476b-b09a-5bc59e14178e"
      },
      "source": [
        "# df_train[num_cols]\n",
        "num_cols"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['Age',\n",
              " 'fnlwgt',\n",
              " 'education-num',\n",
              " 'capital-gain',\n",
              " 'capital-loss',\n",
              " 'hours-per-week']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-20zO9BI9p32",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df_train = normalize(df_train_, num_cols, cat_cols, norm='quantile')\n",
        "df_test  = normalize(df_test_, num_cols, cat_cols, norm='quantile') "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9rGqkpVp-Dcs",
        "colab_type": "text"
      },
      "source": [
        "# Conferindo"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gxmnnwld-nUw",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 419
        },
        "outputId": "3c513d96-eb83-4bae-ed1e-1f370a92c616"
      },
      "source": [
        "df_train"
      ],
      "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>Age</th>\n",
              "      <th>fnlwgt</th>\n",
              "      <th>education-num</th>\n",
              "      <th>capital-gain</th>\n",
              "      <th>capital-loss</th>\n",
              "      <th>hours-per-week</th>\n",
              "      <th>workclass</th>\n",
              "      <th>education</th>\n",
              "      <th>Marital-status</th>\n",
              "      <th>Occupation</th>\n",
              "      <th>Relationship</th>\n",
              "      <th>Race</th>\n",
              "      <th>Sex</th>\n",
              "      <th>native-country</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>18991</th>\n",
              "      <td>0.157157</td>\n",
              "      <td>0.828224</td>\n",
              "      <td>0.292292</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.968969</td>\n",
              "      <td>0.472973</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>13</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9743</th>\n",
              "      <td>0.391391</td>\n",
              "      <td>0.580776</td>\n",
              "      <td>0.292292</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.906907</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>23417</th>\n",
              "      <td>0.791792</td>\n",
              "      <td>0.601957</td>\n",
              "      <td>0.098098</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.845345</td>\n",
              "      <td>2</td>\n",
              "      <td>7</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25248</th>\n",
              "      <td>0.692693</td>\n",
              "      <td>0.116928</td>\n",
              "      <td>0.699199</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.472973</td>\n",
              "      <td>2</td>\n",
              "      <td>6</td>\n",
              "      <td>6</td>\n",
              "      <td>5</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12892</th>\n",
              "      <td>0.233233</td>\n",
              "      <td>0.657658</td>\n",
              "      <td>0.736737</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.711211</td>\n",
              "      <td>2</td>\n",
              "      <td>8</td>\n",
              "      <td>3</td>\n",
              "      <td>6</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</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>16432</th>\n",
              "      <td>0.020521</td>\n",
              "      <td>0.813811</td>\n",
              "      <td>0.098098</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.035536</td>\n",
              "      <td>2</td>\n",
              "      <td>7</td>\n",
              "      <td>3</td>\n",
              "      <td>6</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25934</th>\n",
              "      <td>0.737738</td>\n",
              "      <td>0.988794</td>\n",
              "      <td>0.123123</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.472973</td>\n",
              "      <td>2</td>\n",
              "      <td>13</td>\n",
              "      <td>3</td>\n",
              "      <td>5</td>\n",
              "      <td>7</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3235</th>\n",
              "      <td>0.471972</td>\n",
              "      <td>0.940512</td>\n",
              "      <td>0.026527</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.191191</td>\n",
              "      <td>3</td>\n",
              "      <td>10</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16377</th>\n",
              "      <td>0.855856</td>\n",
              "      <td>0.382639</td>\n",
              "      <td>0.835335</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.845345</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20271</th>\n",
              "      <td>0.599600</td>\n",
              "      <td>0.650760</td>\n",
              "      <td>0.835335</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.143143</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>7</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>22792 rows × 14 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "            Age    fnlwgt  education-num  ...  Race  Sex  native-country\n",
              "18991  0.157157  0.828224       0.292292  ...     2    3               2\n",
              "9743   0.391391  0.580776       0.292292  ...     2    2              30\n",
              "23417  0.791792  0.601957       0.098098  ...     2    2               2\n",
              "25248  0.692693  0.116928       0.699199  ...     4    3               2\n",
              "12892  0.233233  0.657658       0.736737  ...     2    2               2\n",
              "...         ...       ...            ...  ...   ...  ...             ...\n",
              "16432  0.020521  0.813811       0.098098  ...     2    3               2\n",
              "25934  0.737738  0.988794       0.123123  ...     3    3               2\n",
              "3235   0.471972  0.940512       0.026527  ...     2    2               2\n",
              "16377  0.855856  0.382639       0.835335  ...     2    2               2\n",
              "20271  0.599600  0.650760       0.835335  ...     2    2               2\n",
              "\n",
              "[22792 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WA2Dcphh-o5H",
        "colab_type": "text"
      },
      "source": [
        "## Criar classe do dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2JXfxrmD-sqg",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "8fa8c147-7bbb-4e5c-e38d-08f49f78076b"
      },
      "source": [
        "class CensusDataset(Dataset):\n",
        "    def __init__(self, df, num_cols, cat_cols, target):\n",
        "        super(CensusDataset, self).__init__()\n",
        "\n",
        "        self.x_cont = df[num_cols].values.copy().astype(np.float32)\n",
        "        self.x_cat  = df[cat_cols].values.copy().astype(np.int64)\n",
        "        \n",
        "        self.y      = target.values.reshape(-1).copy().astype(np.int64)\n",
        "        \n",
        "    def __len__(self):\n",
        "        return self.x_cat.shape[0]\n",
        "    \n",
        "    def __getitem__(self, index):\n",
        "        x_cont = self.x_cont[index]\n",
        "        x_cat  = self.x_cat[index]\n",
        "        \n",
        "        y      = self.y[index]\n",
        "        \n",
        "        return x_cont, x_cat, y\n",
        "\n",
        "# testando\n",
        "ds = CensusDataset(df_train, \n",
        "                     num_cols, \n",
        "                     cat_cols, \n",
        "                     y_train)\n",
        "x_cont, x_cat, y = ds[0]\n",
        "x_cont.shape, x_cat.shape, y"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((6,), (8,), 0)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 40
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XzX-03rt_Te-",
        "colab_type": "text"
      },
      "source": [
        "# Dataloaders"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t6bE6o7W3EgQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "14126b1a-8202-4414-ed14-e1f73ea14101"
      },
      "source": [
        "BATCH_SZ = 32\n",
        "\n",
        "# datasets\n",
        "ds_train = CensusDataset(df_train, \n",
        "                     num_cols, \n",
        "                     cat_cols, \n",
        "                     y_train)\n",
        "\n",
        "ds_debug = CensusDataset(df_train[:BATCH_SZ], \n",
        "                     num_cols, \n",
        "                     cat_cols, \n",
        "                     y_train[:BATCH_SZ])\n",
        "\n",
        "\n",
        "ds_test  = CensusDataset(df_test, \n",
        "                     num_cols, \n",
        "                     cat_cols, \n",
        "                     y_test)\n",
        "\n",
        "# dataloaders\n",
        "dataloaders = {\n",
        "     'train': DataLoader(ds_train,\n",
        "                         batch_size=BATCH_SZ,\n",
        "                         shuffle = True, \n",
        "                         num_workers=4,\n",
        "                         pin_memory=True),\n",
        "     'test' : DataLoader(ds_test,\n",
        "                         batch_size=BATCH_SZ,\n",
        "                         shuffle=False,\n",
        "                         num_workers=4,\n",
        "                         pin_memory=True),\n",
        "     'debug': DataLoader(ds_debug,\n",
        "                         batch_size=BATCH_SZ,\n",
        "                         shuffle=False,\n",
        "                         num_workers=4,\n",
        "                         pin_memory=True)\n",
        "     }\n",
        "\n",
        "# sanity check\n",
        "dl_sizes = {x: len(dataloaders[x]) for x in dataloaders.keys()}\n",
        "dl_sizes "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'debug': 1, 'test': 306, 'train': 713}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 41
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "G2ZAQNy2QeFR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "4080d491-1773-4961-a297-3c9b1aabdd85"
      },
      "source": [
        "x_cont, x_cat, y= next(iter(dataloaders['train']))\n",
        "torch.mean(y.type(torch.float))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "tensor(0.2500)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vz12PBQxAi_C",
        "colab_type": "text"
      },
      "source": [
        "# Modelo"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q648boUFAssI",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "fd3c412a-6cc3-4093-a252-9695ed780347"
      },
      "source": [
        "class CatModel(nn.Module):\n",
        "    def __init__(self, embedding_sizes=emb_szs, n_cont=len(num_cols), out=2):\n",
        "        super().__init__()\n",
        "        self.embeddings = nn.ModuleList([nn.Embedding(categories, size) \n",
        "                    for categories,size in embedding_sizes])\n",
        "        \n",
        "        n_emb_sum = sum(e.embedding_dim for e in self.embeddings) \n",
        "        d_model   = n_emb_sum + n_cont\n",
        "\n",
        "        self.actv = nn.ReLU()\n",
        "\n",
        "        self.fc0  = nn.Linear(d_model, n_emb_sum)\n",
        "        self.fc1  = nn.Linear(n_emb_sum, out)\n",
        "\n",
        "    def forward(self, x_cont, x_cat):\n",
        "        xcat = [e(x_cat[:,i]) for i, e in enumerate(self.embeddings)]\n",
        "        xcat = torch.cat(xcat, dim=1) \n",
        "        \n",
        "        o = torch.cat((x_cont, xcat), dim=1)\n",
        "        o = self.fc0(o)\n",
        "        o = self.actv(o)\n",
        "        return self.fc1(o)\n",
        "#--------------------------------------------------------------------------------\n",
        "model = CatModel(\n",
        "    embedding_sizes=emb_szs, \n",
        "    n_cont=len(num_cols), \n",
        "    )\n",
        "with torch.no_grad():\n",
        "    print(model(x_cont, x_cat).shape)   "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "torch.Size([32, 2])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sRLtcXLpA_Kx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 357
        },
        "outputId": "f123b271-b216-4355-af0c-d02578a8a205"
      },
      "source": [
        "def count_parameters(model):\n",
        "    return sum(p.numel() for p in model.parameters() if p.requires_grad)\n",
        "\n",
        "print('\\n','#' * 45,f'\\n # The model has {count_parameters(model):,}' \\\n",
        "       ' trainable parameters #\\n', '#' * 45,'\\n' )  \n",
        "model"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            " ############################################# \n",
            " # The model has 11,997 trainable parameters #\n",
            " ############################################# \n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "CatModel(\n",
              "  (embeddings): ModuleList(\n",
              "    (0): Embedding(10, 10)\n",
              "    (1): Embedding(18, 14)\n",
              "    (2): Embedding(9, 9)\n",
              "    (3): Embedding(16, 13)\n",
              "    (4): Embedding(8, 9)\n",
              "    (5): Embedding(7, 8)\n",
              "    (6): Embedding(4, 7)\n",
              "    (7): Embedding(43, 26)\n",
              "  )\n",
              "  (actv): ReLU()\n",
              "  (fc0): Linear(in_features=102, out_features=96, bias=True)\n",
              "  (fc1): Linear(in_features=96, out_features=2, bias=True)\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8D2pwp6fBGhx",
        "colab_type": "text"
      },
      "source": [
        "# Treino"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gye0TYAYBViY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def train(model, device, train_loader, loss_fn, optimizer):\n",
        "    model.train()\n",
        "    loss_train = []\n",
        "    for x_cont, x_cat, y in train_loader:\n",
        "        optimizer.zero_grad()\n",
        "        y_pred = model(\n",
        "            x_cont.to(device),\n",
        "            x_cat.to(device),\n",
        "            )\n",
        "        loss = loss_fn(y_pred, y.to(device))\n",
        "        loss_train.append(loss.item())\n",
        "    \n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "  \n",
        "    ave_train_loss = sum(loss_train) / len(loss_train)\n",
        "    return ave_train_loss\n",
        "\n",
        "def test(model, device, valid_loader, loss_fn):\n",
        "    model.eval()\n",
        "    preds, trues, loss_test = [],[],[]\n",
        "    for x_cont, x_cat, y in valid_loader:\n",
        "        with torch.no_grad():\n",
        "            y_pred = model(\n",
        "            x_cont.to(device),\n",
        "            x_cat.to(device),\n",
        "            )\n",
        "\n",
        "        loss = loss_fn(y_pred, y.to(device))\n",
        "        loss_test.append(loss.item())\n",
        "      \n",
        "        preds += y_pred.argmax(-1).cpu().numpy().tolist()\n",
        "        trues += y.cpu().numpy().tolist()\n",
        "\n",
        "    ave_test_loss = sum(loss_test) / len(loss_test)\n",
        "    \n",
        "    preds = np.array(preds)\n",
        "    trues = np.array(trues)\n",
        "    acc = (1.* (preds==trues)).mean() \n",
        "    return ave_test_loss, f1_score(trues, preds), acc"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MmEdJmcOraR5",
        "colab_type": "text"
      },
      "source": [
        "### LabelSmoothing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f6YOw8sxrZnp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Função do Tio Alan :)\n",
        "\n",
        "class LabelSmoothing(nn.Module):\n",
        "    def __init__(self, smoothing = 0.1):\n",
        "        assert 0.0 <= smoothing < 1.0, f\"got smoothing={smoothing}\"\n",
        "        super(LabelSmoothing, self).__init__()\n",
        "        self.smoothing = smoothing\n",
        "\n",
        "    def forward(self, input, target):\n",
        "        target = target.detach()\n",
        "        log = F.log_softmax(input, dim=1)\n",
        "        oh = torch.ones_like(input).detach_()*(self.smoothing/input.size(1))\n",
        "        oh = oh.scatter_(1, target.unsqueeze(1),\n",
        "                        (1.0 - self.smoothing + self.smoothing/input.size(1)))\n",
        "        return ((-oh*log).sum(dim=1).mean())"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8-ZfIleCESXH",
        "colab_type": "text"
      },
      "source": [
        "# <span style=\"color:orange\"> Treino completo"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IJr14D5p_H74",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 595
        },
        "outputId": "64d6c6f7-8263-45fa-c14e-dcd1ec7e8c1c"
      },
      "source": [
        "from sklearn.metrics import f1_score\n",
        "\n",
        "start = torch.cuda.Event(enable_timing=True)\n",
        "end = torch.cuda.Event(enable_timing=True)\n",
        "#---------------------------------------------------------------------------------\n",
        "deterministic()\n",
        "N_EPOCHS = 30\n",
        "SMTH     = 0.08\n",
        "\n",
        "model = CatModel(\n",
        "    embedding_sizes=emb_szs, \n",
        "    n_cont=len(num_cols), \n",
        "    ).to(device)\n",
        "\n",
        "criterion = LabelSmoothing(SMTH)\n",
        "optimizer =  torch.optim.Adam(model.parameters())\n",
        "#---------------------------------------------------------------------------------\n",
        "start.record()\n",
        "for step in range(1, N_EPOCHS+1):\n",
        "    l_train    = train(model, device, dataloaders['train'], criterion, optimizer)\n",
        "    l_test, f1,_ = test( model, device, dataloaders['test'],  criterion)\n",
        "    torch.save(model.state_dict(), os.path.join('', 'Epoca-{}.pth'.format(step)))\n",
        "    print(f'Epoch [{step}/{N_EPOCHS}] |', end=' ')\n",
        "    print(f'Train Loss: {l_train:.3f} -- Test Loss: {l_test:.3f} -- F1: {f1:.3}')\n",
        "end.record()\n",
        "torch.cuda.synchronize()    \n",
        "#---------------------------------------------------------------------------------\n",
        "print(f'\\nEnd of training')\n",
        "print(f'Tempo: {start.elapsed_time(end)/1000/60 :.3f} min.')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Deterministic experiment, seed: 2357\n",
            "Epoch [1/30] | Train Loss: 0.429 -- Test Loss: 0.418 -- F1: 0.631\n",
            "Epoch [2/30] | Train Loss: 0.410 -- Test Loss: 0.414 -- F1: 0.65\n",
            "Epoch [3/30] | Train Loss: 0.405 -- Test Loss: 0.413 -- F1: 0.65\n",
            "Epoch [4/30] | Train Loss: 0.402 -- Test Loss: 0.413 -- F1: 0.656\n",
            "Epoch [5/30] | Train Loss: 0.401 -- Test Loss: 0.414 -- F1: 0.635\n",
            "Epoch [6/30] | Train Loss: 0.399 -- Test Loss: 0.413 -- F1: 0.635\n",
            "Epoch [7/30] | Train Loss: 0.398 -- Test Loss: 0.416 -- F1: 0.624\n",
            "Epoch [8/30] | Train Loss: 0.396 -- Test Loss: 0.413 -- F1: 0.639\n",
            "Epoch [9/30] | Train Loss: 0.394 -- Test Loss: 0.415 -- F1: 0.654\n",
            "Epoch [10/30] | Train Loss: 0.392 -- Test Loss: 0.417 -- F1: 0.659\n",
            "Epoch [11/30] | Train Loss: 0.391 -- Test Loss: 0.417 -- F1: 0.633\n",
            "Epoch [12/30] | Train Loss: 0.390 -- Test Loss: 0.419 -- F1: 0.674\n",
            "Epoch [13/30] | Train Loss: 0.388 -- Test Loss: 0.415 -- F1: 0.648\n",
            "Epoch [14/30] | Train Loss: 0.387 -- Test Loss: 0.417 -- F1: 0.664\n",
            "Epoch [15/30] | Train Loss: 0.385 -- Test Loss: 0.418 -- F1: 0.647\n",
            "Epoch [16/30] | Train Loss: 0.384 -- Test Loss: 0.425 -- F1: 0.623\n",
            "Epoch [17/30] | Train Loss: 0.383 -- Test Loss: 0.421 -- F1: 0.653\n",
            "Epoch [18/30] | Train Loss: 0.382 -- Test Loss: 0.423 -- F1: 0.628\n",
            "Epoch [19/30] | Train Loss: 0.381 -- Test Loss: 0.422 -- F1: 0.625\n",
            "Epoch [20/30] | Train Loss: 0.380 -- Test Loss: 0.425 -- F1: 0.661\n",
            "Epoch [21/30] | Train Loss: 0.379 -- Test Loss: 0.426 -- F1: 0.649\n",
            "Epoch [22/30] | Train Loss: 0.378 -- Test Loss: 0.424 -- F1: 0.634\n",
            "Epoch [23/30] | Train Loss: 0.377 -- Test Loss: 0.428 -- F1: 0.64\n",
            "Epoch [24/30] | Train Loss: 0.376 -- Test Loss: 0.433 -- F1: 0.62\n",
            "Epoch [25/30] | Train Loss: 0.375 -- Test Loss: 0.431 -- F1: 0.601\n",
            "Epoch [26/30] | Train Loss: 0.374 -- Test Loss: 0.433 -- F1: 0.634\n",
            "Epoch [27/30] | Train Loss: 0.372 -- Test Loss: 0.431 -- F1: 0.637\n",
            "Epoch [28/30] | Train Loss: 0.372 -- Test Loss: 0.434 -- F1: 0.626\n",
            "Epoch [29/30] | Train Loss: 0.372 -- Test Loss: 0.435 -- F1: 0.642\n",
            "Epoch [30/30] | Train Loss: 0.371 -- Test Loss: 0.437 -- F1: 0.63\n",
            "\n",
            "End of training\n",
            "Tempo: 3.627 min.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "039Qjhr6LUG_",
        "colab_type": "text"
      },
      "source": [
        "#  <span style=\"color:orange\"> Validação no conj. de teste\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HlIw4vY5drmf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def logits_and_trues(model, dloader, device=device):\n",
        "    model.eval()\n",
        "    logits, trues = [],[]\n",
        "    for x_cont, x_cat, y in dloader:\n",
        "        with torch.no_grad():\n",
        "            y_pred = model(\n",
        "            x_cont.to(device),\n",
        "            x_cat.to(device),\n",
        "            )\n",
        "\n",
        "        logits.extend(y_pred.to('cpu').numpy().tolist())\n",
        "        trues.extend(y.to('cpu').numpy().tolist())\n",
        "    \n",
        "    prob  = F.softmax(torch.tensor(logits), dim=1).detach().numpy()\n",
        "    probs = prob[:,1]\n",
        "    probs = np.array(probs)\n",
        "    trues = np.array(trues)\n",
        "    err_ =  far_vs_frr(trues, probs, fig=False)\n",
        "  \n",
        "    return probs, trues, err_"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2LETRF7MpApO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def far_vs_frr(labels, outs, steps=1000, fig=False):\n",
        "    lives = outs[labels == 1]\n",
        "    spoofs = outs[labels == 0]\n",
        "    max_live = max(lives)\n",
        "    min_spoof = min(spoofs)\n",
        "    num_lives = len(lives)\n",
        "    num_spoofs = len(spoofs)\n",
        "    histogram = np.linspace(0,1,num=steps)\n",
        "    FRR = np.array([100*np.mean(spoofs >= hist) for hist in histogram])\n",
        "    FAR = np.array([100*np.mean(lives < hist) for hist in histogram])\n",
        "    a = FRR[FAR - FRR<= 0]\n",
        "    s_a = len(a)\n",
        "    if (FAR[s_a] - FRR[s_a]) <= (FRR[s_a+1]-FAR[s_a+1]):\n",
        "        ERR = (FAR[s_a] + FRR[s_a])/2\n",
        "        ERR_index = s_a\n",
        "    else:\n",
        "        ERR = (FAR[s_a+1] + FRR[s_a+1])/2\n",
        "        ERR_index = s_a+1\n",
        "    \n",
        "    if fig:\n",
        "        print(f'Negados: {num_lives} Aprovados: {num_spoofs}')\n",
        "        figure = plt.figure(num=1, figsize=(10, 8), dpi=80, facecolor='w', edgecolor='k')\n",
        "        ax = figure.add_subplot(111)\n",
        "        plt.plot(histogram, FRR, color='r', label='FRR', lw=2)\n",
        "        plt.plot(histogram, FAR, color='navy', label='FAR', lw=2)\n",
        "        plt.plot([histogram[ERR_index]], [ERR], marker='o', markersize=12, color=\"lime\", label=f'ERR: {ERR :.3}')\n",
        "        axes = plt.gca()\n",
        "        axes.set_ylim([0, 4*ERR])\n",
        "        plt.title('FAR (erro tipo I) vs FRR (erro tipo II)')\n",
        "        plt.legend()\n",
        "        plt.show()\n",
        "        return ERR, histogram[ERR_index]\n",
        "    else:\n",
        "        return ERR"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gGKPONUrnTge",
        "colab_type": "text"
      },
      "source": [
        "# Carrega todas as épocas salvas e verifica o melhor F$_1$ \n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M-YjtTMcih9y",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "6e85c938-0080-40a5-f4c7-ecee47d2047f"
      },
      "source": [
        "step0,step1,step2,best_recall, best,low_err = 0,0,0,0,0,100\n",
        "for i in range(1,N_EPOCHS+1):\n",
        "    model = CatModel(\n",
        "    embedding_sizes=emb_szs, \n",
        "    n_cont=len(num_cols), \n",
        "    ).to(device)\n",
        "    \n",
        "    string = '/content/Epoca-'+str(i)+'.pth'\n",
        "    model.load_state_dict(torch.load(string, map_location=device))\n",
        "    \n",
        "    probs, trues, _err = logits_and_trues(\n",
        "        model, \n",
        "        dataloaders['test'], \n",
        "        device=device\n",
        "        )\n",
        "    \n",
        "    fpr, tpr, thresholds = roc_curve(trues, probs)\n",
        "    precision, recall, thresh = precision_recall_curve(trues, probs)\n",
        "    f1_score = 2*precision*recall/(precision + recall)\n",
        "    f1_score = f1_score[:-1]\n",
        "    precision = precision[:-1]\n",
        "    recall = recall[:-1]\n",
        "    f1_score[np.isnan(f1_score)] = -1.0\n",
        "    bestF1 = np.argmax(f1_score)\n",
        "    \n",
        "    if f1_score[bestF1] > best:\n",
        "        best = f1_score[bestF1]\n",
        "        step0 = i\n",
        "\n",
        "    if recall[bestF1] > best_recall:\n",
        "        best_recall = recall[bestF1]\n",
        "        step2 = i\n",
        "\n",
        "\n",
        "    if _err < low_err:\n",
        "        low_err = _err\n",
        "        step1 = i\n",
        "    \n",
        "    print(f'Epoca: {i}')\n",
        "    print(f'   O threshold do melhor F1 score: >>> {thresh[bestF1] :.3} <<<')\n",
        "    print(f'   F1: {f1_score[bestF1] :.3} -- ', end='')\n",
        "    print(f'Precision {thresh[bestF1] :.3}: {precision[bestF1] :.3} -- ', end='')\n",
        "    print(f'Recall {thresh[bestF1] :.3}: {recall[bestF1] :.3}')\n",
        "    print(f'   ERR: {_err :.3}\\n')        "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoca: 1\n",
            "   O threshold do melhor F1 score: >>> 0.337 <<<\n",
            "   F1: 0.67 -- Precision 0.337: 0.613 -- Recall 0.337: 0.738\n",
            "   ERR: 20.0\n",
            "\n",
            "Epoca: 2\n",
            "   O threshold do melhor F1 score: >>> 0.376 <<<\n",
            "   F1: 0.68 -- Precision 0.376: 0.615 -- Recall 0.376: 0.759\n",
            "   ERR: 19.2\n",
            "\n",
            "Epoca: 3\n",
            "   O threshold do melhor F1 score: >>> 0.351 <<<\n",
            "   F1: 0.679 -- Precision 0.351: 0.6 -- Recall 0.351: 0.782\n",
            "   ERR: 19.0\n",
            "\n",
            "Epoca: 4\n",
            "   O threshold do melhor F1 score: >>> 0.341 <<<\n",
            "   F1: 0.683 -- Precision 0.341: 0.609 -- Recall 0.341: 0.777\n",
            "   ERR: 18.7\n",
            "\n",
            "Epoca: 5\n",
            "   O threshold do melhor F1 score: >>> 0.315 <<<\n",
            "   F1: 0.678 -- Precision 0.315: 0.618 -- Recall 0.315: 0.752\n",
            "   ERR: 19.2\n",
            "\n",
            "Epoca: 6\n",
            "   O threshold do melhor F1 score: >>> 0.33 <<<\n",
            "   F1: 0.684 -- Precision 0.33: 0.633 -- Recall 0.33: 0.743\n",
            "   ERR: 19.0\n",
            "\n",
            "Epoca: 7\n",
            "   O threshold do melhor F1 score: >>> 0.262 <<<\n",
            "   F1: 0.683 -- Precision 0.262: 0.606 -- Recall 0.262: 0.782\n",
            "   ERR: 18.8\n",
            "\n",
            "Epoca: 8\n",
            "   O threshold do melhor F1 score: >>> 0.316 <<<\n",
            "   F1: 0.682 -- Precision 0.316: 0.612 -- Recall 0.316: 0.772\n",
            "   ERR: 18.8\n",
            "\n",
            "Epoca: 9\n",
            "   O threshold do melhor F1 score: >>> 0.347 <<<\n",
            "   F1: 0.683 -- Precision 0.347: 0.593 -- Recall 0.347: 0.807\n",
            "   ERR: 18.7\n",
            "\n",
            "Epoca: 10\n",
            "   O threshold do melhor F1 score: >>> 0.403 <<<\n",
            "   F1: 0.681 -- Precision 0.403: 0.626 -- Recall 0.403: 0.746\n",
            "   ERR: 19.0\n",
            "\n",
            "Epoca: 11\n",
            "   O threshold do melhor F1 score: >>> 0.302 <<<\n",
            "   F1: 0.68 -- Precision 0.302: 0.621 -- Recall 0.302: 0.751\n",
            "   ERR: 18.9\n",
            "\n",
            "Epoca: 12\n",
            "   O threshold do melhor F1 score: >>> 0.378 <<<\n",
            "   F1: 0.683 -- Precision 0.378: 0.597 -- Recall 0.378: 0.799\n",
            "   ERR: 18.7\n",
            "\n",
            "Epoca: 13\n",
            "   O threshold do melhor F1 score: >>> 0.386 <<<\n",
            "   F1: 0.684 -- Precision 0.386: 0.647 -- Recall 0.386: 0.725\n",
            "   ERR: 19.0\n",
            "\n",
            "Epoca: 14\n",
            "   O threshold do melhor F1 score: >>> 0.368 <<<\n",
            "   F1: 0.68 -- Precision 0.368: 0.616 -- Recall 0.368: 0.758\n",
            "   ERR: 19.1\n",
            "\n",
            "Epoca: 15\n",
            "   O threshold do melhor F1 score: >>> 0.358 <<<\n",
            "   F1: 0.68 -- Precision 0.358: 0.613 -- Recall 0.358: 0.764\n",
            "   ERR: 19.0\n",
            "\n",
            "Epoca: 16\n",
            "   O threshold do melhor F1 score: >>> 0.272 <<<\n",
            "   F1: 0.676 -- Precision 0.272: 0.595 -- Recall 0.272: 0.782\n",
            "   ERR: 19.2\n",
            "\n",
            "Epoca: 17\n",
            "   O threshold do melhor F1 score: >>> 0.319 <<<\n",
            "   F1: 0.68 -- Precision 0.319: 0.584 -- Recall 0.319: 0.813\n",
            "   ERR: 18.9\n",
            "\n",
            "Epoca: 18\n",
            "   O threshold do melhor F1 score: >>> 0.32 <<<\n",
            "   F1: 0.68 -- Precision 0.32: 0.608 -- Recall 0.32: 0.771\n",
            "   ERR: 19.3\n",
            "\n",
            "Epoca: 19\n",
            "   O threshold do melhor F1 score: >>> 0.319 <<<\n",
            "   F1: 0.682 -- Precision 0.319: 0.6 -- Recall 0.319: 0.791\n",
            "   ERR: 18.9\n",
            "\n",
            "Epoca: 20\n",
            "   O threshold do melhor F1 score: >>> 0.436 <<<\n",
            "   F1: 0.679 -- Precision 0.436: 0.648 -- Recall 0.436: 0.714\n",
            "   ERR: 19.0\n",
            "\n",
            "Epoca: 21\n",
            "   O threshold do melhor F1 score: >>> 0.32 <<<\n",
            "   F1: 0.673 -- Precision 0.32: 0.592 -- Recall 0.32: 0.779\n",
            "   ERR: 19.6\n",
            "\n",
            "Epoca: 22\n",
            "   O threshold do melhor F1 score: >>> 0.323 <<<\n",
            "   F1: 0.676 -- Precision 0.323: 0.609 -- Recall 0.323: 0.76\n",
            "   ERR: 19.3\n",
            "\n",
            "Epoca: 23\n",
            "   O threshold do melhor F1 score: >>> 0.326 <<<\n",
            "   F1: 0.679 -- Precision 0.326: 0.622 -- Recall 0.326: 0.747\n",
            "   ERR: 19.3\n",
            "\n",
            "Epoca: 24\n",
            "   O threshold do melhor F1 score: >>> 0.301 <<<\n",
            "   F1: 0.678 -- Precision 0.301: 0.633 -- Recall 0.301: 0.73\n",
            "   ERR: 19.3\n",
            "\n",
            "Epoca: 25\n",
            "   O threshold do melhor F1 score: >>> 0.263 <<<\n",
            "   F1: 0.678 -- Precision 0.263: 0.588 -- Recall 0.263: 0.801\n",
            "   ERR: 19.1\n",
            "\n",
            "Epoca: 26\n",
            "   O threshold do melhor F1 score: >>> 0.284 <<<\n",
            "   F1: 0.671 -- Precision 0.284: 0.59 -- Recall 0.284: 0.777\n",
            "   ERR: 19.5\n",
            "\n",
            "Epoca: 27\n",
            "   O threshold do melhor F1 score: >>> 0.336 <<<\n",
            "   F1: 0.676 -- Precision 0.336: 0.606 -- Recall 0.336: 0.764\n",
            "   ERR: 19.5\n",
            "\n",
            "Epoca: 28\n",
            "   O threshold do melhor F1 score: >>> 0.301 <<<\n",
            "   F1: 0.674 -- Precision 0.301: 0.595 -- Recall 0.301: 0.779\n",
            "   ERR: 19.5\n",
            "\n",
            "Epoca: 29\n",
            "   O threshold do melhor F1 score: >>> 0.327 <<<\n",
            "   F1: 0.675 -- Precision 0.327: 0.6 -- Recall 0.327: 0.771\n",
            "   ERR: 19.5\n",
            "\n",
            "Epoca: 30\n",
            "   O threshold do melhor F1 score: >>> 0.317 <<<\n",
            "   F1: 0.674 -- Precision 0.317: 0.616 -- Recall 0.317: 0.745\n",
            "   ERR: 19.5\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U-R4KlZJfIhO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        },
        "outputId": "bc2796c0-17cc-45ff-db27-bef3f65c3831"
      },
      "source": [
        "print(f'Melhor F1: {best:.3} na época: {step0}') \n",
        "print(f'Melhor Err: {low_err:.3} na época: {step1}') \n",
        "print(f'Melhor Recall: {best_recall:.3} na época: {step2}') "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Melhor F1: 0.684 na época: 13\n",
            "Melhor Err: 18.7 na época: 4\n",
            "Melhor Recall: 0.813 na época: 17\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "fY_wyc-UiC4x",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 359
        },
        "outputId": "90460fa2-02d7-4195-e4f6-bad5ea18661f"
      },
      "source": [
        "model = CatModel(\n",
        "    embedding_sizes=emb_szs, \n",
        "    n_cont=len(num_cols), \n",
        "    ).to(device)\n",
        "\n",
        "string = '/content/Epoca-'+str(step1)+'.pth'\n",
        "model.load_state_dict(torch.load(string, map_location=device))\n",
        "\n",
        "probs, trues, err_ = logits_and_trues(model, dataloaders['test'], device=device)\n",
        "\n",
        "data = pd.DataFrame({'Target':trues, 'Prob':probs})\n",
        "data.head(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Target</th>\n",
              "      <th>Prob</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>0.834887</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>0.126702</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>0.274510</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0</td>\n",
              "      <td>0.053237</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>0.950657</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>1</td>\n",
              "      <td>0.833605</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>1</td>\n",
              "      <td>0.886142</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>1</td>\n",
              "      <td>0.726522</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>0</td>\n",
              "      <td>0.063340</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>1</td>\n",
              "      <td>0.133281</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Target      Prob\n",
              "0       1  0.834887\n",
              "1       0  0.126702\n",
              "2       1  0.274510\n",
              "3       0  0.053237\n",
              "4       1  0.950657\n",
              "5       1  0.833605\n",
              "6       1  0.886142\n",
              "7       1  0.726522\n",
              "8       0  0.063340\n",
              "9       1  0.133281"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kk5CcCL_dAeZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "FP = data[data.Target==1].sort_values(by='Prob')\n",
        "FN = data[data.Target==0].sort_values(by='Prob', ascending=False)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2aoeypaPw5MO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 419
        },
        "outputId": "a5db39ec-3ae3-4412-d5a4-3df16034c943"
      },
      "source": [
        "FP"
      ],
      "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>Target</th>\n",
              "      <th>Prob</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1947</th>\n",
              "      <td>1</td>\n",
              "      <td>0.048304</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3733</th>\n",
              "      <td>1</td>\n",
              "      <td>0.048337</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2970</th>\n",
              "      <td>1</td>\n",
              "      <td>0.048895</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8163</th>\n",
              "      <td>1</td>\n",
              "      <td>0.052123</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2481</th>\n",
              "      <td>1</td>\n",
              "      <td>0.055100</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7612</th>\n",
              "      <td>1</td>\n",
              "      <td>0.973820</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6370</th>\n",
              "      <td>1</td>\n",
              "      <td>0.975040</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5501</th>\n",
              "      <td>1</td>\n",
              "      <td>0.975818</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>314</th>\n",
              "      <td>1</td>\n",
              "      <td>0.976650</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9643</th>\n",
              "      <td>1</td>\n",
              "      <td>0.980398</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2391 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      Target      Prob\n",
              "1947       1  0.048304\n",
              "3733       1  0.048337\n",
              "2970       1  0.048895\n",
              "8163       1  0.052123\n",
              "2481       1  0.055100\n",
              "...      ...       ...\n",
              "7612       1  0.973820\n",
              "6370       1  0.975040\n",
              "5501       1  0.975818\n",
              "314        1  0.976650\n",
              "9643       1  0.980398\n",
              "\n",
              "[2391 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 60
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z747rR01Z7Gy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 419
        },
        "outputId": "d6c76d37-0193-48d1-b391-7cc65bf3453e"
      },
      "source": [
        "FN"
      ],
      "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>Target</th>\n",
              "      <th>Prob</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>8425</th>\n",
              "      <td>0</td>\n",
              "      <td>0.970708</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9670</th>\n",
              "      <td>0</td>\n",
              "      <td>0.957191</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2809</th>\n",
              "      <td>0</td>\n",
              "      <td>0.952864</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7904</th>\n",
              "      <td>0</td>\n",
              "      <td>0.944218</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4847</th>\n",
              "      <td>0</td>\n",
              "      <td>0.943765</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1128</th>\n",
              "      <td>0</td>\n",
              "      <td>0.010536</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1576</th>\n",
              "      <td>0</td>\n",
              "      <td>0.009803</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>753</th>\n",
              "      <td>0</td>\n",
              "      <td>0.008708</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6531</th>\n",
              "      <td>0</td>\n",
              "      <td>0.008492</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8505</th>\n",
              "      <td>0</td>\n",
              "      <td>0.008376</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>7378 rows × 2 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      Target      Prob\n",
              "8425       0  0.970708\n",
              "9670       0  0.957191\n",
              "2809       0  0.952864\n",
              "7904       0  0.944218\n",
              "4847       0  0.943765\n",
              "...      ...       ...\n",
              "1128       0  0.010536\n",
              "1576       0  0.009803\n",
              "753        0  0.008708\n",
              "6531       0  0.008492\n",
              "8505       0  0.008376\n",
              "\n",
              "[7378 rows x 2 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 61
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "81gX6QFOiC5M",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 530
        },
        "outputId": "d06745f7-0465-46b6-a0d4-90b252be1ef0"
      },
      "source": [
        "def plot_roc_curve(y, y_pred, gini, ks):\n",
        "    fpr, tpr, thresholds = roc_curve(y, y_pred)\n",
        "    roc_auc = auc(fpr, tpr)\n",
        "    fig = plt.figure()\n",
        "    plt.rcParams['figure.figsize'] = (15, 8)\n",
        "\n",
        "    plt.plot(fpr, tpr, 'b', lw=3, \n",
        "             label='%s KS = %0.3f,  AUC = %0.3f, GINI = %0.3f' \n",
        "             % ('Model: ',ks, roc_auc, gini))\n",
        "\n",
        "    plt.plot([0, 1], [0, 1], 'k--')\n",
        "    plt.xlim([0.0, 1.0])\n",
        "    plt.xticks(fontsize=16)\n",
        "    plt.yticks(fontsize=16)\n",
        "    plt.ylim([0.0, 1.0])\n",
        "    plt.xlabel('False Positive Rate', size=16)\n",
        "    plt.ylabel('True Positive Rate', size=16)\n",
        "    plt.legend(loc=0, fontsize=16)\n",
        "    plt.show()\n",
        "\n",
        "fpr, tpr, thresholds = roc_curve(data['Target'], data['Prob'])\n",
        "roc_auc = auc(fpr, tpr)\n",
        "gini_ = 2 * roc_auc_score(data.Target.values, data.Prob.values) - 1\n",
        "\n",
        "ks = stats.ks_2samp(data[data['Target'] == 0]['Prob'], \n",
        "                    data[data['Target'] == 1]['Prob']).statistic\n",
        "\n",
        "print(f'KS = {ks :.3} -- AUC = {roc_auc :.3} -- GINI = {gini_ :.3}')\n",
        "\n",
        "plot_roc_curve(data.Target.values, data.Prob.values, gini_, ks)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "KS = 0.634 -- AUC = 0.898 -- GINI = 0.796\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B5_wt2KBphXv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "b42f2ed9-b208-43ad-ecd8-1a003a20a566"
      },
      "source": [
        "fpr, tpr, thresholds = roc_curve(data.Target.values, data.Prob.values)\n",
        "\n",
        "precision,recall,pr_thresh = precision_recall_curve(data.Target.values, \n",
        "                                                    data.Prob.values)\n",
        "\n",
        "f1_score = 2*precision*recall/(precision+recall)\n",
        "f1_score = f1_score[:-1]\n",
        "precision = precision[:-1]\n",
        "recall = recall[:-1]\n",
        "f1_score[np.isnan(f1_score)] = -1.0\n",
        "bestF1 = np.argmax(f1_score)\n",
        "\n",
        "print(f'O threshold do melhor F1: >>> {pr_thresh[bestF1] :.3} <<<')\n",
        "print(f'   F1: {f1_score[bestF1] :.3} -- ', end='')\n",
        "print(f'Precision: {precision[bestF1] :.3} -- ', end='')\n",
        "print(f'Recall: {recall[bestF1] :.3}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "O threshold do melhor F1: >>> 0.341 <<<\n",
            "   F1: 0.683 -- Precision: 0.609 -- Recall: 0.777\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ce3gNHYWq3od",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        },
        "outputId": "872b70ce-e1af-45ac-cb4e-f0d8a9c7fa13"
      },
      "source": [
        "opt_idx = np.argmin(np.sqrt(np.square(1.0-tpr) + np.square(fpr)))\n",
        "opt_threshold = thresholds[opt_idx]\n",
        "print(f'O threshold do melhor F1: {pr_thresh[bestF1] :.3}')\n",
        "\n",
        "y_hat_best_thres = (data.Prob.values >= pr_thresh[bestF1]).astype(int)\n",
        "# y_hat_best_thres = (data.Prob.values >= 0.5).astype(int)\n",
        "\n",
        "print(f'\\nMatriz de confusão no thrs {pr_thresh[bestF1] :.3}:')\n",
        "confusion_matrix(trues, y_hat_best_thres)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "O threshold do melhor F1: 0.341\n",
            "\n",
            "Matriz de confusão no thrs 0.341:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[6184, 1194],\n",
              "       [ 533, 1858]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 68
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vpWUGgmcpqfU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 577
        },
        "outputId": "961d106f-4b1e-4821-854f-66ac15e7fb88"
      },
      "source": [
        "plt.rcParams.update({'font.size': 16})\n",
        "plt.plot(pr_thresh,f1_score, lw=2)\n",
        "plt.plot(pr_thresh,precision,  lw=2)\n",
        "plt.plot(pr_thresh,recall, lw=2)\n",
        "plt.xlabel('Threshold')\n",
        "plt.ylabel('Score')\n",
        "plt.xticks(fontsize=16)\n",
        "plt.xlim(-0.01, 1.01)\n",
        "plt.ylim(-0.02, 1.02)\n",
        "plt.yticks(fontsize=16)\n",
        "plt.grid()\n",
        "plt.legend(['F$_1$-score', 'Precision', 'Recall'],prop={'size': 16})\n",
        "plt.tight_layout(2)\n",
        "plt.savefig('precision-recall-thres', bbox_inches='tight')\n",
        "print(f'O threshold do melhor F1: >>> {pr_thresh[bestF1] :.3} <<<')\n",
        "print(f'   F1: {f1_score[bestF1] :.3} -- ', end='')\n",
        "print(f'Precision: {precision[bestF1] :.3} -- ', end='')\n",
        "print(f'Recall: {recall[bestF1] :.3}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "O threshold do melhor F1: >>> 0.341 <<<\n",
            "   F1: 0.683 -- Precision: 0.609 -- Recall: 0.777\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "avk39os_fC2l",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 598
        },
        "outputId": "6a701568-6850-420f-9f60-f8d424a40865"
      },
      "source": [
        "_,Err_point = far_vs_frr(data.Target.values, data.Prob.values, steps=2000, fig=True)\n",
        "Err_point"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Negados: 2391 Aprovados: 7378\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 800x640 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.3001500750375187"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MUmiK6RNYXuc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "0559b330-fdaa-4f9f-c69a-a2042206dd61"
      },
      "source": [
        "y_err = (data.Prob.values >= Err_point).astype(int)\n",
        "\n",
        "print(f'\\nMatriz de confusão no ponto do ERR {Err_point :.3}:')\n",
        "confusion_matrix(trues, y_err)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Matriz de confusão no ponto do ERR 0.3:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[6004, 1374],\n",
              "       [ 448, 1943]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 71
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-ODZHUY1ZHYm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "6fb7c15b-639b-4eba-d9b8-b8ab4e4b4acf"
      },
      "source": [
        "y_50 = (data.Prob.values >= 0.5).astype(int)\n",
        "\n",
        "print(f'\\nMatriz de confusão no threshold {0.50}:')\n",
        "confusion_matrix(trues, y_50)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Matriz de confusão no threshold 0.5:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[6698,  680],\n",
              "       [ 893, 1498]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 72
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y9ixWS0q85th",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 368
        },
        "outputId": "22ae2af6-9109-4ff1-9c02-14baf110f636"
      },
      "source": [
        "plt.rcParams['figure.figsize'] = (9, 5)\n",
        "\n",
        "sns.set_context(\"notebook\", font_scale=1.3)\n",
        "\n",
        "confusao = confusion_matrix(trues, y_hat_best_thres)\n",
        "group_names = ['TN\\nAprovados corretos',\n",
        "               'FN - erro tipo II (FRR)\\nNegados errados', \n",
        "               'FP - erro tipo I  (FAR)\\nAprovados errados',\n",
        "               'TP\\nNegados corretos']\n",
        "\n",
        "group_counts = ['({0:0.0f})'.format(value) for value in\n",
        "                confusao.flatten()]\n",
        "\n",
        "group_percentages = ['{0:.1%}'.format(value) for value in\n",
        "                     confusao.flatten()/np.sum(confusao)]\n",
        "\n",
        "labels = [f'{v1}\\n{v2}' for v1, v2 in\n",
        "          zip(group_names,group_counts)]\n",
        "\n",
        "labels = np.asarray(labels).reshape(2,2)\n",
        "\n",
        "cmap = 'Reds'#'binary'#'PuRd'#'jet'#'YlOrRd'#\n",
        "\n",
        "ax = sns.heatmap(confusao, annot=labels, fmt='', square=False,\n",
        "                 cmap=cmap, cbar=True, linewidths=0.7,\n",
        "                )\n",
        "\n",
        "ax.set_ylabel('Target - Real')\n",
        "ax.set_xlabel('Predição - Modelo')\n",
        "ax.set_xticklabels(['Aprovados','Negados'], va='center', \n",
        "                   rotation = 0, position=(0,-0.03), \n",
        "                   fontsize=18)\n",
        "ax.set_yticklabels(['Aprovados','Negados'], \n",
        "                   va='center', rotation = 90, position=(0,0.28),\n",
        "                   fontsize=18)\n",
        "ax.set_title(f'limiar = {pr_thresh[bestF1] :.3}', fontsize=18)\n",
        "\n",
        "#######\n",
        "# Save\n",
        "#######\n",
        "plt.savefig('confusao.png', bbox_inches='tight')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 648x360 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Q61ej_qQKRcl",
        "colab_type": "text"
      },
      "source": [
        "# FIM\n"
      ]
    }
  ]
}