02_regressao_linear.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "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/luasampaio/CienciaDados/blob/main/02_regressao_linear.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ODUU2N6ALSR8"
      },
      "source": [
        "#regressão utilizando o otimizador\n",
        "#Definindo as bibliotecas\n",
        "import numpy as np  #biblioteca necessária para trabalhar com os vetores e matrizes\n",
        "import scipy   #biblioteca necessária para obter as funções de treinamento\n",
        "import matplotlib.pyplot as plt #biblioteca utilizada para construir os gráficos\n",
        "from scipy.optimize import curve_fit # biblioteca necessária para realiza a otimização dos MSE"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Transformando em Array\n"
      ],
      "metadata": {
        "id": "PnOt4aU74HN4"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xq-z1IceLfFl"
      },
      "source": [
        "#definindo as variáveis\n",
        "idade=[18,25,57,45,26,64,37,40,24,33]  # variável independente\n",
        "salarioAnual=[15000,29000,68000,52000,32000,80000,41000,45000,26000,33000] #variável dependente\n",
        "\n",
        "xData = np.array(idade)    #transformando a lista em array\n",
        "yData = np.array(salarioAnual) #transformando a lista em array"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "23qIMoYDL3Ii"
      },
      "source": [
        "#define a função a ser otimizada (regressão simples)\n",
        "def equacaoLinear(x, a, b):\n",
        "    return a * x + b"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o9KGtuywL9rT"
      },
      "source": [
        "#gera os parâmetros iniciais para o otimizador\n",
        "parametrosIniciais = np.array([1.0, 1.0])\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iN5NP65jMEM5"
      },
      "source": [
        "#realiza a otimização através do erro médio quadrado (MSE)\n",
        "parametrosOtimizados, pcov = curve_fit(equacaoLinear, xData, yData, parametrosIniciais)\n",
        "#parametrosOtimizados - contém os parâmetros de ajuste da curva\n",
        "#pcov - contém a covariância dos parâmteros encontrados\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WEQYYk1yMMer"
      },
      "source": [
        "#realiza a previsão dos dados através do modelo (constroi a equação linear)\n",
        "pervisaoModelo = equacaoLinear(xData, *parametrosOtimizados) #utiliza a função linear com os parâmetros otimizados\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gRCipFZcMyFX"
      },
      "source": [
        "#encontra o erro obsoluto (linhas verticais)\n",
        "erroAbsoluto = pervisaoModelo - yData #(valor previsto - valor real)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "klbnW7zCND0f",
        "outputId": "3602418f-7b70-47bf-9728-638d0475a10d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "source": [
        "#calcula o erro quadrado entre cada medida\n",
        "SE = np.square(erroAbsoluto)\n",
        "#calcula o MSE\n",
        "MSE = np.mean(SE)\n",
        "print('SE: ', SE)\n",
        "print('MSE: ', MSE)\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "SE:  [ 4587883.33064101  6834760.82563804   413069.18137435   634115.64144357\n",
            " 18436761.23538558  4467167.3659006   1517955.21070226  1424802.59030848\n",
            "   873982.13514337 15601891.57382987]\n",
            "MSE:  5479238.909036714\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vQpeeAaTNkVw",
        "outputId": "308265b4-f3e2-411d-b078-8d4b2b10ba67",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "#realiza o cálculo do coeficiente de determinação\n",
        "Rsquared = 1.0 - (np.var(erroAbsoluto) / np.var(yData)) # numpy.var - encontra a variância entre os dados do vetor\n",
        "print('Coeficiente de Determinação:', Rsquared)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Coeficiente de Determinação: 0.9846300347582353\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "icuiNsUvQxiw",
        "outputId": "e5a14073-ce74-4d38-d576-2cf2bc979817",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "#mostra os parâmetros da regressão\n",
        "print('Y = {}X {}'.format(parametrosOtimizados[0],parametrosOtimizados[1]))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Y = 1320.5325666669085X -6627.651716729711\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nyKuCk5KOAIP",
        "outputId": "27353006-8b86-47dc-afd1-924d681d3acc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 399
        }
      },
      "source": [
        "#realiza o plot da figura\n",
        "f = plt.figure(figsize=(4, 4), dpi=100) #indica o tamanho da figura\n",
        "axes = f.add_subplot(111) #cria os objetos para o subplot\n",
        "\n",
        "# frealiza o plot dos dados (pontos no gráfico)\n",
        "axes.plot(xData, yData,'ro')\n",
        "\n",
        "# cria os dados para serem utilizados na construção da linha (equação)\n",
        "xModel = np.linspace(min(xData), max(xData)) #encontra os valores maximos e mínimos da \"linha\"\n",
        "yModel = equacaoLinear(xModel, *parametrosOtimizados) # acplica a função com os parâmetros obtidos\n",
        "\n",
        "# realiza o plot da \"linha\"\n",
        "axes.plot(xModel, yModel)\n",
        "plt.xlabel(\"Idade\")\n",
        "plt.ylabel(\"Salário Anual (R$)\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'Salário Anual (R$)')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 400x400 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kcAXKpE3UB51"
      },
      "source": [
        "**Agora vamos mostrar como o mesmo código ficaria com a utilização do Sklearn**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N6kaanJ-P3JM"
      },
      "source": [
        "#utiliza as funções do sklearn para construir a regressão linear\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.metrics import mean_squared_error"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0ZrYHOdjQHDe"
      },
      "source": [
        "reg= LinearRegression() #objeto para a regressão linear\n",
        "x_ModeloSklern=xData.reshape((-1, 1)) # na regressão linear é necessário que o X seja 2D\n",
        "regressao= reg.fit (x_ModeloSklern,yData) # realiza a regressão"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uVDHTZAwSKO2"
      },
      "source": [
        "previsao=reg.predict(x_ModeloSklern)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZsZFiD0URzrs",
        "outputId": "ef254b6d-a109-48b8-df6d-207b814f48aa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "MSE= mean_squared_error(yData,previsao) # encontra o MSE através do sklearn\n",
        "print('MSE: ', MSE)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "MSE:  5479238.909036706\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KwHDu49MSqrS",
        "outputId": "4c6b7131-37ff-423d-a86a-416e3c900222",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "#parâmetros encontrados\n",
        "print('Y = {}X {}'.format(reg.coef_,reg.intercept_))\n",
        "\n",
        "from sklearn.metrics import r2_score #método para o cálculo do R2\n",
        "\n",
        "R_2 = r2_score(yData, previsao)  #realiza o cálculo do R2\n",
        "\n",
        "print(\"Coeficiente de Determinação (R2):\", R_2)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Y = [1320.53256496]X -6627.651646877624\n",
            "Coeficiente de Determinação (R2): 0.9846300347582353\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cpuvt2kTT4Yn"
      },
      "source": [
        "**Agora vamos realizar a construção de uma regressão para um banco de dados real (mais complexo)**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hBaG8ElKTTGy"
      },
      "source": [
        "#regressão utilizando um BD mais complexo\n",
        "import pandas as pd\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e5GCyid2TcCE",
        "outputId": "12628b92-d01c-439f-ad51-2e06e8371f47",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 74
        }
      },
      "source": [
        "#função utilizada apenas para realizar o upload dos dados para o google colab\n",
        "from google.colab import files\n",
        "uploaded = files.upload()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-7131ead9-36cd-4d63-afe5-9e566db06ab2\" name=\"files[]\" multiple disabled />\n",
              "     <output id=\"result-7131ead9-36cd-4d63-afe5-9e566db06ab2\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving headbrain.csv to headbrain.csv\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rIRhiDjXTkL-",
        "outputId": "1f7bcc1d-d3e7-4164-ec98-3037c355b45a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "#realiza a leitura do banco de dados\n",
        "data=pd.read_csv(\"headbrain.csv\")\n",
        "data.head()  #realiza visualização das 5 primeiras linhas do BD"
      ],
      "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>Gender</th>\n",
              "      <th>Age Range</th>\n",
              "      <th>Head Size(cm^3)</th>\n",
              "      <th>Brain Weight(grams)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>4512</td>\n",
              "      <td>1530</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3738</td>\n",
              "      <td>1297</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>4261</td>\n",
              "      <td>1335</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3777</td>\n",
              "      <td>1282</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>4177</td>\n",
              "      <td>1590</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Gender  Age Range  Head Size(cm^3)  Brain Weight(grams)\n",
              "0       1          1             4512                 1530\n",
              "1       1          1             3738                 1297\n",
              "2       1          1             4261                 1335\n",
              "3       1          1             3777                 1282\n",
              "4       1          1             4177                 1590"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F1sEYTloUPOE"
      },
      "source": [
        "#como o formato dos dados em cada uma das colunas do dataframe são séries é necessário converter array\n",
        "x=data[\"Head Size(cm^3)\"].values\n",
        "y=data[\"Brain Weight(grams)\"].values"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dgMiVz78UpUy"
      },
      "source": [
        "#Realiza a construção do modelo de regressão\n",
        "reg= LinearRegression()\n",
        "x_Reshaped=x.reshape((-1, 1)) #coloca os dados no formato 2D\n",
        "regressao= reg.fit (x_Reshaped,y) # encontra os coeficientes (realiza a regressão)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1YmoYMelUwdu"
      },
      "source": [
        "#realiza a previsão\n",
        "previsao=reg.predict(x_Reshaped)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e1z1RPNWWx7i",
        "outputId": "eb826623-8535-4fb3-f5b0-4649a5387b3d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "#parâmetros encontrados\n",
        "print('Y = {}X {}'.format(reg.coef_,reg.intercept_))\n",
        "\n",
        "R_2 = r2_score(y, previsao)  #realiza o cálculo do R2\n",
        "\n",
        "print(\"Coeficiente de Determinação (R2):\", R_2)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Y = [0.26342934]X 325.5734210494426\n",
            "Coeficiente de Determinação (R2): 0.639311719957\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yHZZM5sQV7fV",
        "outputId": "f4df7708-f303-49df-a176-bf17e2d00c17",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 382
        }
      },
      "source": [
        "#realiza o plot dos dados\n",
        "plt.figure(figsize=(4, 4), dpi=100)\n",
        "plt.scatter(x, y,  color='gray') #realiza o plot do gráfico de dispersão\n",
        "plt.plot(x, previsao, color='red', linewidth=2) # realiza o plto da \"linha\"\n",
        "plt.xlabel(\"Head Size(cm^3)\")\n",
        "plt.ylabel(\"Brain Weight(grams)\")\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 400x400 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}