Modelos_regressivos_bootcamp_WoMakersCode_(aula_2).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Modelos regressivos bootcamp WoMakersCode (aula 2).ipynb",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/mariacmartins/datascience-bootcamp-womakerscode/blob/master/Modelos_regressivos_bootcamp_WoMakersCode_(aula_2).ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wSQ-d1TyHTht",
        "colab_type": "text"
      },
      "source": [
        "#Modelos regressivos\n",
        "\n",
        "![](https://media.giphy.com/media/4JVTF9zR9BicshFAb7/giphy.gif)\n",
        "\n",
        "[sklearn.linear_model.LinearRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LinearRegression.html)\n",
        "\n",
        "[sklearn.preprocessing.PolynomialFeatures](https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.PolynomialFeatures.html)\n",
        "\n",
        "\n",
        "###Execícios de Regressão linear\n",
        "\n",
        "**1. Treine uma regressão linear usando a biblioteca sklearn utilizando um dataset de aluguéis de bicicletas**\n",
        "\n",
        "* Calcule os erros utilizando MSE, MAE e R²\n",
        "\n",
        "* Com base nesse resultado, responda: O modelo está bem acurado?\n",
        "\n",
        "Responda: \n",
        "\n",
        "* Qual variável contribui mais para esse modelo? E qual contribui menos?\n",
        "\n",
        "* Retire a variável que contribui menos e treine novamente a regressão. O modelo ficou melhor? "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UbgyAzixAQ16",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.metrics import r2_score, mean_squared_error, mean_absolute_error"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DE--CGKlHqBt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "url = 'https://raw.githubusercontent.com/WoMakersCode/data-science-bootcamp/master/3.%20Modelos%20regressivos/Exercicio_Regressao/day.csv'"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BvbO4J2iHvFu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df = pd.read_csv(url)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uVKsNARXH9DG",
        "colab_type": "code",
        "outputId": "b2871bff-dfff-42f0-907b-d73f6e1eb07e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 0,
      "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>season</th>\n",
              "      <th>yr</th>\n",
              "      <th>mnth</th>\n",
              "      <th>holiday</th>\n",
              "      <th>weekday</th>\n",
              "      <th>workingday</th>\n",
              "      <th>cnt</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>6</td>\n",
              "      <td>0</td>\n",
              "      <td>985</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>801</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1349</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1562</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1600</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   season  yr  mnth  holiday  weekday  workingday   cnt\n",
              "0       1   0     1        0        6           0   985\n",
              "1       1   0     1        0        0           0   801\n",
              "2       1   0     1        0        1           1  1349\n",
              "3       1   0     1        0        2           1  1562\n",
              "4       1   0     1        0        3           1  1600"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0aMHUmOlIDi_",
        "colab_type": "code",
        "outputId": "89f43bd7-c079-43f8-e982-70eccad600cf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 221
        }
      },
      "source": [
        "df.info()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 731 entries, 0 to 730\n",
            "Data columns (total 7 columns):\n",
            "season        731 non-null int64\n",
            "yr            731 non-null int64\n",
            "mnth          731 non-null int64\n",
            "holiday       731 non-null int64\n",
            "weekday       731 non-null int64\n",
            "workingday    731 non-null int64\n",
            "cnt           731 non-null int64\n",
            "dtypes: int64(7)\n",
            "memory usage: 40.1 KB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TPL8bzl8IGBV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#separando os dados em 4 dataframes diferentes(X: variáveis independentes, y: variável resposta)\n",
        "#com 30% dos dados sendo usados para teste.\n",
        "X_treino, X_teste, y_treino, y_teste = train_test_split(df.drop(columns='cnt'), df['cnt'], test_size=0.3)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JXPLm5KtKIlX",
        "colab_type": "code",
        "outputId": "adb57ff8-8615-4acf-dce4-207cbf670bbf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "lr = LinearRegression() #Criando a regressão\n",
        "lr.fit(X_treino, y_treino) #Ajustando/treinando o modelo"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZO2LHEpVKaCx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predito = lr.predict(X_teste) #predito é um vetor com todos os valores preditos pelo modelo"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hCzDt8HKKoas",
        "colab_type": "code",
        "outputId": "4145bd7b-faf3-4046-d64b-6187ecc66810",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "predito[:5]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([4452.728594  , 6053.99916384, 4694.67333777, 4798.33011986,\n",
              "       2671.75048193])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fuppdR6pLhUy",
        "colab_type": "code",
        "outputId": "088eb5f9-e1e0-4f1d-8383-bd2e89af4ede",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print('R²: ', lr.score(X_teste, y_teste))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "R²:  0.4782729039809177\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lD4wyIoSMCel",
        "colab_type": "code",
        "outputId": "4f1e12cb-b930-4198-c2e1-815f685b4600",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print('MSE: ', mean_squared_error(y_teste, predito))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "MSE:  1983344.6705639032\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yA3cxGz7MLbd",
        "colab_type": "code",
        "outputId": "021859c5-d567-404b-a682-e077cb4bd3e6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print('MAE: ', mean_absolute_error(y_teste, predito))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "MAE:  1196.1982546736492\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QXJddoiIMTQY",
        "colab_type": "code",
        "outputId": "ab4bef8b-75be-4a05-9003-48c797f141fb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "lr.coef_ #coeficientes de cada variável independente"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1023.39710628, 2131.97057138, -123.52735169, -331.12188971,\n",
              "         98.54682247,  228.25629865])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U7N3IFQvMZAW",
        "colab_type": "code",
        "outputId": "3eb92831-7551-42c5-d04c-3d2859f8bc64",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "for i,coef in enumerate(lr.coef_):\n",
        "    print(df.columns[i],' : ', coef)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "season  :  1023.3971062836274\n",
            "yr  :  2131.970571380867\n",
            "mnth  :  -123.52735169173434\n",
            "holiday  :  -331.12188970631894\n",
            "weekday  :  98.54682247317078\n",
            "workingday  :  228.25629864549197\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "plMeUHuMNVM5",
        "colab_type": "text"
      },
      "source": [
        "Variável que contribui menos: valor mais próximo de 0\n",
        "\n",
        "Variável que contribui mais: maior valor"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WZtXQWWbMpmO",
        "colab_type": "code",
        "outputId": "7998f0cc-0beb-4111-d670-5e378b71c317",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "lr.intercept_\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1242.5930278216942"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NhomWfudNmeI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_treino, X_teste, y_treino, y_teste = train_test_split(df.drop(columns=['cnt','weekday']), df['cnt'], test_size=0.3) \n",
        "#Retire a variável que contribui menos e treine novamente a regressão. "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0LKI5K3NNyrA",
        "colab_type": "code",
        "outputId": "4bbd3eec-9ed3-42f0-8ee7-fa07e817e7d0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "lr = LinearRegression()\n",
        "lr.fit(X_treino, y_treino)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_mu174gxN4HZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predito = lr.predict(X_teste)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eFTf7onvN7Xa",
        "colab_type": "code",
        "outputId": "4228896a-d34c-479b-e000-804a7a916b71",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print('R²: ', lr.score(X_teste, y_teste))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "R²:  0.5141437165368424\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H1ID_rnHOBRT",
        "colab_type": "code",
        "outputId": "d039cb2d-8110-4053-f781-5f2090a43e15",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print('MSE: ', mean_squared_error(y_teste, predito))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "MSE:  1907971.5842023394\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ju3dGiwQOEW8",
        "colab_type": "code",
        "outputId": "c35da003-a82c-476b-8591-af70b35b304d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "\n",
        "print('MAE: ', mean_absolute_error(y_teste, predito))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "MAE:  1167.0632818292847\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aB4jUKBSOHb3",
        "colab_type": "code",
        "outputId": "1db6de4d-8b9e-4028-e88d-f0631bef88a5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        }
      },
      "source": [
        "for i,coef in enumerate(lr.coef_):\n",
        "    print(df.columns[i],' : ', coef)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "season  :  988.5226731693197\n",
            "yr  :  2172.3784735679833\n",
            "mnth  :  -120.31108842875301\n",
            "holiday  :  -402.75747916043\n",
            "weekday  :  256.9167585924692\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "l_1XNUE0YfRU",
        "colab_type": "text"
      },
      "source": [
        "### Utilizando regressão linear para prever a expectativa de vida usando o Índice de Massa Corporal"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dK5NWaMIYrfr",
        "colab_type": "text"
      },
      "source": [
        "Neste exercício, você trabalhará com dados sobre a expectativa média de vida ao nascer e o IMC médio. Os dados são do Gapminder.\n",
        "\n",
        "O arquivo possui três colunas, contendo os seguintes dados:\n",
        "\n",
        "País: \n",
        "* O país em que a pessoa nasceu. \n",
        "\n",
        "Expectativa de vida: \n",
        "* A expectativa média de vida no nascimento de uma pessoa naquele país. \n",
        "\n",
        "IMC:\n",
        "* O IMC médio naquele país.\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "Preveja usando um IMC de 21.07931 e atribua-o à variável laos_life_exp."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a2BMZLmCZqX9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "from sklearn.linear_model import LinearRegression"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6PprTGLWZw5n",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "url = 'https://raw.githubusercontent.com/WoMakersCode/data-science-bootcamp/master/3.%20Modelos%20regressivos/Exercicio_1/bmi_and_life_expectancy.csv'"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rvf-UkgoZ5oI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "bmi_life_data = pd.read_csv(url)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RZpmSlVdaBL-",
        "colab_type": "code",
        "outputId": "eca2d8dd-badc-4557-c83a-9c24bb21ec06",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "bmi_life_data.head()"
      ],
      "execution_count": 0,
      "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>Country</th>\n",
              "      <th>Life expectancy</th>\n",
              "      <th>BMI</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Afghanistan</td>\n",
              "      <td>52.8</td>\n",
              "      <td>20.62058</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Albania</td>\n",
              "      <td>76.8</td>\n",
              "      <td>26.44657</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Algeria</td>\n",
              "      <td>75.5</td>\n",
              "      <td>24.59620</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Andorra</td>\n",
              "      <td>84.6</td>\n",
              "      <td>27.63048</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Angola</td>\n",
              "      <td>56.7</td>\n",
              "      <td>22.25083</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       Country  Life expectancy       BMI\n",
              "0  Afghanistan             52.8  20.62058\n",
              "1      Albania             76.8  26.44657\n",
              "2      Algeria             75.5  24.59620\n",
              "3      Andorra             84.6  27.63048\n",
              "4       Angola             56.7  22.25083"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 33
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yFsz4bW7cqYS",
        "colab_type": "code",
        "outputId": "836a4259-8f0a-4e07-b2d6-a6a4f719d443",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "bmi_life_model = LinearRegression()\n",
        "bmi_life_model.fit(bmi_life_data[['BMI']], bmi_life_data[['Life expectancy']])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 34
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EWZ1AllFctUV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "laos_life_exp = bmi_life_model.predict([[21.07931]])\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8UU8W2-Hcvn7",
        "colab_type": "code",
        "outputId": "85c86dc2-26f5-4a33-c621-0334e39dd804",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print(laos_life_exp)\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[60.31564716]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FBo0yNmedvMb",
        "colab_type": "text"
      },
      "source": [
        "# Regressão Polinomial "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SpSHsAjgd922",
        "colab_type": "text"
      },
      "source": [
        "Agora vamos tentar praticar um pouco sobre como implementar regressão polinomial. No arquivo data.csv você poderá encontrar dados gerados para um feature preditora ('Var_X') e a feature de saída ('Var_Y'), seguindo um padrão não-linear.\n",
        "\n",
        "Use a classe PolynomialFeatures do sklearn's para estender a coluna de feature de previsão em várias colunas com recursos polinomiais. Brinque com diferentes graus de polinômio e use o botão Test Run para ver o que melhor se encaixa.\n",
        "\n",
        "Para concluir, siga as etapas abaixo:\n",
        "\n",
        "1. **Carregue os dados**\n",
        "\n",
        "  Os dados estão no arquivo chamado 'data.csv'. Observe que esses dados têm uma linha de cabeçalho. Certifique-se de dividir os dados na feature de previsão em X e na feature de resultado em y. Para X, verifique se ele está em uma matriz bidimensional de 20 linhas por 1 coluna. Pode ser necessário usar a função de remodelação do NumPy para fazer isso.\n",
        "\n",
        "2. **Crie recursos polinomiais**\n",
        "\n",
        "  Crie uma instância da classe PolynomialFeatures do sklearn e atribua-a à variável poly_feat. Preste atenção em como definir o grau de recursos, pois será assim que o exercício será avaliado. Crie os recursos polinomiais usando o método .fit_transform () do objeto PolynomialFeatures. O lado \"apto\" do método considera quantos recursos são necessários na saída e o lado \"transformação\" aplica essas considerações aos dados fornecidos ao método como argumento. Atribua a nova matriz de recurso à variável X_poly.\n",
        "\n",
        "3. **Construa um modelo de regressão polinomial**\n",
        "\n",
        "  Crie um modelo de regressão polinomial combinando a classe LinearRegression do sklearn com os recursos polinomiais. Atribua o modelo de ajuste ao poly_model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MmpJen77eIaE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.preprocessing import PolynomialFeatures"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HR9vA-5JfK8k",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_data = pd.read_csv('https://raw.githubusercontent.com/WoMakersCode/data-science-bootcamp/master/3.%20Modelos%20regressivos/Exercicio_3/data.csv')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ilB5WeBwfR9u",
        "colab_type": "code",
        "outputId": "04a94e79-5072-4cc5-afbe-29403a550881",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "train_data.head()"
      ],
      "execution_count": 129,
      "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>Var_X</th>\n",
              "      <th>Var_Y</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>-0.33532</td>\n",
              "      <td>6.66854</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.02160</td>\n",
              "      <td>3.86398</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>-1.19438</td>\n",
              "      <td>5.16161</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>-0.65046</td>\n",
              "      <td>8.43823</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>-0.28001</td>\n",
              "      <td>5.57201</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     Var_X    Var_Y\n",
              "0 -0.33532  6.66854\n",
              "1  0.02160  3.86398\n",
              "2 -1.19438  5.16161\n",
              "3 -0.65046  8.43823\n",
              "4 -0.28001  5.57201"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 129
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CgVVV-1afWGL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Crie a variável do preditor e de saída. \n",
        "X = train_data['Var_X'].values.reshape(-1, 1)\n",
        "y = train_data['Var_Y'].values"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QAsAA9HRj-Zs",
        "colab_type": "code",
        "outputId": "4d829c87-33a2-46ee-968c-148d2a85b688",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 357
        }
      },
      "source": [
        "X"
      ],
      "execution_count": 131,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[-0.33532],\n",
              "       [ 0.0216 ],\n",
              "       [-1.19438],\n",
              "       [-0.65046],\n",
              "       [-0.28001],\n",
              "       [ 1.93258],\n",
              "       [ 1.2262 ],\n",
              "       [ 0.74727],\n",
              "       [ 3.32853],\n",
              "       [ 2.87457],\n",
              "       [-1.48662],\n",
              "       [ 0.37629],\n",
              "       [ 1.43918],\n",
              "       [ 0.24183],\n",
              "       [-2.7914 ],\n",
              "       [ 1.08176],\n",
              "       [ 2.81555],\n",
              "       [ 0.54924],\n",
              "       [ 2.36449],\n",
              "       [-1.01925]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 131
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cZlHYqJPkBlD",
        "colab_type": "code",
        "outputId": "37155a3e-282a-4bec-8f16-62690cb732fc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "source": [
        "y"
      ],
      "execution_count": 132,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([  6.66854,   3.86398,   5.16161,   8.43823,   5.57201, -11.1327 ,\n",
              "        -5.31226,  -4.63725,   3.8065 ,  -6.06084,   7.22328,   2.38887,\n",
              "        -7.13415,   2.00412,   4.29794,  -5.86553,  -5.20711,  -3.52863,\n",
              "       -10.16202,   5.31123])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 132
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MOSufgMIgLYP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Crie as features polinomiais e depois faça um fit e transform na feature preditora\n",
        "poly_feat = PolynomialFeatures(degree = 4)\n",
        "X_poly = poly_feat.fit_transform(X)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-Zg2qMqQhsgc",
        "colab_type": "code",
        "outputId": "06c323e8-1fdc-45f4-8726-ae393a3a4982",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 697
        }
      },
      "source": [
        "X_poly"
      ],
      "execution_count": 134,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 1.00000000e+00, -3.35320000e-01,  1.12439502e-01,\n",
              "        -3.77032139e-02,  1.26426417e-02],\n",
              "       [ 1.00000000e+00,  2.16000000e-02,  4.66560000e-04,\n",
              "         1.00776960e-05,  2.17678234e-07],\n",
              "       [ 1.00000000e+00, -1.19438000e+00,  1.42654358e+00,\n",
              "        -1.70383513e+00,  2.03502660e+00],\n",
              "       [ 1.00000000e+00, -6.50460000e-01,  4.23098212e-01,\n",
              "        -2.75208463e-01,  1.79012097e-01],\n",
              "       [ 1.00000000e+00, -2.80010000e-01,  7.84056001e-02,\n",
              "        -2.19543521e-02,  6.14743813e-03],\n",
              "       [ 1.00000000e+00,  1.93258000e+00,  3.73486546e+00,\n",
              "         7.21792628e+00,  1.39492200e+01],\n",
              "       [ 1.00000000e+00,  1.22620000e+00,  1.50356644e+00,\n",
              "         1.84367317e+00,  2.26071204e+00],\n",
              "       [ 1.00000000e+00,  7.47270000e-01,  5.58412453e-01,\n",
              "         4.17284874e-01,  3.11824468e-01],\n",
              "       [ 1.00000000e+00,  3.32853000e+00,  1.10791120e+01,\n",
              "         3.68771565e+01,  1.22746722e+02],\n",
              "       [ 1.00000000e+00,  2.87457000e+00,  8.26315268e+00,\n",
              "         2.37530108e+01,  6.82796923e+01],\n",
              "       [ 1.00000000e+00, -1.48662000e+00,  2.21003902e+00,\n",
              "        -3.28548821e+00,  4.88427249e+00],\n",
              "       [ 1.00000000e+00,  3.76290000e-01,  1.41594164e-01,\n",
              "         5.32804680e-02,  2.00489073e-02],\n",
              "       [ 1.00000000e+00,  1.43918000e+00,  2.07123907e+00,\n",
              "         2.98088585e+00,  4.29003130e+00],\n",
              "       [ 1.00000000e+00,  2.41830000e-01,  5.84817489e-02,\n",
              "         1.41426413e-02,  3.42011495e-03],\n",
              "       [ 1.00000000e+00, -2.79140000e+00,  7.79191396e+00,\n",
              "        -2.17503486e+01,  6.07139232e+01],\n",
              "       [ 1.00000000e+00,  1.08176000e+00,  1.17020470e+00,\n",
              "         1.26588063e+00,  1.36937903e+00],\n",
              "       [ 1.00000000e+00,  2.81555000e+00,  7.92732180e+00,\n",
              "         2.23197709e+01,  6.28424310e+01],\n",
              "       [ 1.00000000e+00,  5.49240000e-01,  3.01664578e-01,\n",
              "         1.65686253e-01,  9.10015174e-02],\n",
              "       [ 1.00000000e+00,  2.36449000e+00,  5.59081296e+00,\n",
              "         1.32194213e+01,  3.12571896e+01],\n",
              "       [ 1.00000000e+00, -1.01925000e+00,  1.03887056e+00,\n",
              "        -1.05886882e+00,  1.07925205e+00]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 134
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NVJqLb0WghAS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Crie e dê fit no modelo de regressão polinomial\n",
        "poly_model = LinearRegression(fit_intercept = False).fit(X_poly, y)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jNkI4mWdq4YY",
        "colab_type": "code",
        "outputId": "88351a31-04d2-464c-cb99-d44f37a0fded",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print('R²:', poly_model.score(X_poly, y))"
      ],
      "execution_count": 136,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "R²: 0.9532581924980009\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "goU6crJ1KEhS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.preprocessing import PolynomialFeatures\n",
        "from sklearn.model_selection import train_test_split\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jgVJuufOKOed",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_data = pd.read_csv('https://raw.githubusercontent.com/WoMakersCode/data-science-bootcamp/master/3.%20Modelos%20regressivos/Exercicio_3/data.csv')\n",
        "X = train_data['Var_X'].values.reshape(-1, 1)\n",
        "y = train_data['Var_Y'].values"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n2mOLQTPKVGP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "xtrain, xtest, ytrain, ytest = train_test_split(X, y, test_size = 0.2)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4rnIWaI6NsPx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "1c641d2d-ebd5-45c7-cf3d-004d66b8f422"
      },
      "source": [
        "X_novo = train_data['Var_X'].values\n",
        "lin_reg = LinearRegression()\n",
        "lin_reg.fit(X, y)"
      ],
      "execution_count": 162,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 162
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aa9wqVuHKbTI",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "outputId": "09ae13ce-2e5c-42c5-a304-544e2ae55e35"
      },
      "source": [
        "def viz_linear():\n",
        "    plt.scatter(X, y, color='red')\n",
        "    plt.plot(X, lin_reg.predict(X), color='blue')\n",
        "    plt.title('Regressao Linear')\n",
        "    plt.xlabel('X')\n",
        "    plt.ylabel('Y')\n",
        "    plt.show()\n",
        "\n",
        "viz_linear()"
      ],
      "execution_count": 170,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZAAAAEWCAYAAABIVsEJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n2SUEAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CEGdoyyPKW-U",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "poly_feat = PolynomialFeatures(degree = 4)\n",
        "X_poly = poly_feat.fit_transform(X)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B7WkthM8KZZ5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "ploy_reg = LinearRegression()\n",
        "poly_model = ploy_reg.fit(X_poly, y)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "usRiHGtpO_r_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "outputId": "8373dd9e-5fcb-4dcb-a083-a148f6f797b8"
      },
      "source": [
        "def viz_poly():\n",
        "    plt.scatter(X, y, color='red')\n",
        "    plt.plot(X, ploy_reg.predict(X_poly), color='blue')\n",
        "    plt.title('Regressão Polinomial')\n",
        "    plt.xlabel('X')\n",
        "    plt.ylabel('Y')\n",
        "    plt.show()\n",
        "\n",
        "viz_poly()"
      ],
      "execution_count": 171,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZAAAAEWCAYAAABIVsEJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JZYZCTUTP9Bm",
        "colab_type": "text"
      },
      "source": [
        "# Feature Scaling\n",
        "\n",
        "Agora vamos tentar praticar um pouco sobre como normalizar nossos dados. User a classe StanderScaler do Sklearn para estandatizar antes de uma regressão linear.\n",
        "\n",
        "Para concluir, siga as etapas abaixo:\n",
        "\n",
        "1. **Carregue os dados**\n",
        "\n",
        "  Os dados estão no arquivo chamado 'data.csv'. Observe que não há linha de cabeçalho neste arquivo. Divida os dados para que os seis features preditoras (primeiras seis colunas) sejam armazenados em X e o resultado (última coluna) seja armazenado em y.\n",
        "\n",
        "2. **Execute o dimensionamento de recursos nos dados via padronização**\n",
        "\n",
        "  Crie uma instância do StandardScaler do sklearn e atribua a variavel scaler. Calcule os parâmetros de dimensionamento usando o método .fit_transform () na matriz de recursos do preditor, que também retorna as variáveis do preditor em seus valores padronizados. Armazene esses valores padronizados em X_scaled.\n",
        "\n",
        "3. **Ajustar dados usando regressão linear com regularização de Lasso**\n",
        "\n",
        "  Crie uma instância da classe Lasso do sklearn e atribua-a à variável lasso_reg. Você não precisa definir nenhum valor de parâmetro: use os valores padrão para o questionário. Use o método .fit () do objeto Lasso para ajustar o modelo de regressão aos dados. Certifique-se de aplicar o ajuste aos dados padronizados da etapa anterior (X_scaled), não aos dados originais.\n",
        "\n",
        "4. **Inspecione os coeficientes do modelo de regressão**\n",
        "\n",
        "  Obtenha os coeficientes do modelo de regressão de ajuste usando o atributo .coef_ do objeto Lasso. Armazene isso na variável reg_coef: os coeficientes serão impressos"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v2UG3p1JQP0P",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Importe as bibliotecas que irá usar\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import r2_score, mean_squared_error, mean_absolute_error"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vc6zwSNBQuLj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_data = pd.read_csv('https://raw.githubusercontent.com/WoMakersCode/data-science-bootcamp/master/3.%20Modelos%20regressivos/Exercicio_4/data.csv', header = None)\n",
        "X = train_data.iloc[:,:-1]\n",
        "y = train_data.iloc[:,-1]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oXyybwPpQ9gp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X_treino, X_teste, y_treino, y_teste = train_test_split(train_data.iloc[:,:-1], train_data.iloc[:,-1], test_size=0.3)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YMOJlxmaQ_u4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "\n",
        "#Crie o objeto de normalização\n",
        "scaler = StandardScaler()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W3UnAmI5RA3p",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Crie e de fit do objeto de normalização\n",
        "X_scaled = scaler.fit_transform(X_treino)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "anTbMPLiZGvU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "lr = LinearRegression()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Xeu5doFqRGUI",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "359205c4-f65e-4d19-9c40-25a1312298d3"
      },
      "source": [
        "#Dê um fit no seu modelo\n",
        "lr.fit(X_scaled, y_treino)"
      ],
      "execution_count": 207,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 207
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LqN8ZJ5GRIW9",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "c3ad2a7c-70be-49e0-9329-ec6c6b94c296"
      },
      "source": [
        "#Pegue e imprima os coeficientes do seu modelo de regressão.\n",
        "reg_coef = lr.coef_\n",
        "print(reg_coef)"
      ],
      "execution_count": 208,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[ -0.05406792   4.96636333   9.67874028  -0.68075741 -11.96567766\n",
            "   1.77142283]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tFb12EtvRQsV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predito = lr.predict(X_teste)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b-6088vkRU8w",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "1e9d0e0a-f1a7-4813-803e-dcb297ee7114"
      },
      "source": [
        "predito[:5]\n"
      ],
      "execution_count": 210,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 72.79077263, -68.95833579,  64.98387668, -51.14973212,\n",
              "       -60.86737071])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 210
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DD1_QB1jRV2w",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "41c29cdc-961f-412b-e514-696d80f0947c"
      },
      "source": [
        "print('R²: ', lr.score(X_teste, y_teste))"
      ],
      "execution_count": 211,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "R²:  -5.289588629487875\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UhbPEH8rRapv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "c87724b6-754d-481a-b652-9eb62c82229c"
      },
      "source": [
        "print('MSE: ', mean_squared_error(y_teste, predito))"
      ],
      "execution_count": 212,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "MSE:  2008.6290402994068\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vC_zuJVARbjr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "736ab7a2-1899-4290-a10d-4b4db47a6d2d"
      },
      "source": [
        "print('MAE: ', mean_absolute_error(y_teste, predito))"
      ],
      "execution_count": 213,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "MAE:  36.866724288903114\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "szTSKGfDRdea",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "030aa7ad-660e-4725-fd6c-826884f9ad9e"
      },
      "source": [
        "lr.coef_\n"
      ],
      "execution_count": 214,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ -0.05406792,   4.96636333,   9.67874028,  -0.68075741,\n",
              "       -11.96567766,   1.77142283])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 214
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dXyOQwnmRyGL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        },
        "outputId": "2e6a3dfc-8ba2-43e9-a0a8-5e09a2971f86"
      },
      "source": [
        "for i,coef in enumerate(lr.coef_):\n",
        "    print(train_data.columns[i],' : ', coef)"
      ],
      "execution_count": 215,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0  :  -0.05406791989265848\n",
            "1  :  4.966363326956316\n",
            "2  :  9.678740278624716\n",
            "3  :  -0.680757405701423\n",
            "4  :  -11.965677658765557\n",
            "5  :  1.7714228260832319\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}