Chapter_4_Regression_and_Prediction.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.2"
    },
    "colab": {
      "name": "Chapter 4 - Regression and Prediction.ipynb",
      "provenance": [],
      "include_colab_link": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/roiei/ml_study/blob/master/Chapter_4_Regression_and_Prediction.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "itVNXez5mrND"
      },
      "source": [
        "# Practical Statistics for Data Scientists (Python)\n",
        "# Chapter 4. Regression and Prediction\n",
        "> (c) 2019 Peter C. Bruce, Andrew Bruce, Peter Gedeck"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yjHRFCPbmrNI"
      },
      "source": [
        "Import required Python packages."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:23.437343Z",
          "iopub.status.busy": "2021-02-23T15:56:23.436638Z",
          "iopub.status.idle": "2021-02-23T15:56:24.686931Z",
          "shell.execute_reply": "2021-02-23T15:56:24.687251Z"
        },
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 388
        },
        "id": "Fi3_mmUemrNJ",
        "outputId": "4b2fd767-46c1-4bbe-e718-159ff6c504a0"
      },
      "source": [
        "from pathlib import Path\n",
        "\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.metrics import r2_score, mean_squared_error\n",
        "from sklearn.linear_model import LinearRegression\n",
        "\n",
        "import statsmodels.api as sm\n",
        "import statsmodels.formula.api as smf\n",
        "from statsmodels.stats.outliers_influence import OLSInfluence\n",
        "\n",
        "from pygam import LinearGAM, s, l\n",
        "from pygam.datasets import wage\n",
        "\n",
        "\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from dmba import stepwise_selection\n",
        "from dmba import AIC_score"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "ModuleNotFoundError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mModuleNotFoundError\u001b[0m                       Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-5-a112e435281a>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     10\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mstatsmodels\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstats\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moutliers_influence\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mOLSInfluence\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     11\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 12\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mpygam\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mLinearGAM\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ml\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     13\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mpygam\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdatasets\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mwage\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     14\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mModuleNotFoundError\u001b[0m: No module named 'pygam'",
            "",
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0;32m\nNOTE: If your import is failing due to a missing package, you can\nmanually install dependencies using either !pip or !apt.\n\nTo view examples of installing some common dependencies, click the\n\"Open Examples\" button below.\n\u001b[0;31m---------------------------------------------------------------------------\u001b[0m\n"
          ],
          "errorDetails": {
            "actions": [
              {
                "action": "open_url",
                "actionText": "Open Examples",
                "url": "/notebooks/snippets/importing_libraries.ipynb"
              }
            ]
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:24.692659Z",
          "iopub.status.busy": "2021-02-23T15:56:24.692022Z",
          "iopub.status.idle": "2021-02-23T15:56:24.694637Z",
          "shell.execute_reply": "2021-02-23T15:56:24.694002Z"
        },
        "id": "jMM6hVC3mrNM"
      },
      "source": [
        "%matplotlib inline"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:24.699436Z",
          "iopub.status.busy": "2021-02-23T15:56:24.698519Z",
          "iopub.status.idle": "2021-02-23T15:56:24.982072Z",
          "shell.execute_reply": "2021-02-23T15:56:24.981413Z"
        },
        "id": "POxnNrWGmrNM"
      },
      "source": [
        "try:\n",
        "    import common\n",
        "    DATA = common.dataDirectory()\n",
        "except ImportError:\n",
        "    DATA = Path().resolve() / 'data'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jtg41VeUmrNN"
      },
      "source": [
        "Define paths to data sets. If you don't keep your data in the same directory as the code, adapt the path names."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:24.985549Z",
          "iopub.status.busy": "2021-02-23T15:56:24.984969Z",
          "iopub.status.idle": "2021-02-23T15:56:24.988143Z",
          "shell.execute_reply": "2021-02-23T15:56:24.987227Z"
        },
        "id": "MnTPHJQamrNO"
      },
      "source": [
        "LUNG_CSV = DATA / 'LungDisease.csv'\n",
        "HOUSE_CSV = DATA / 'house_sales.csv'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DhPib047mrNP"
      },
      "source": [
        "# Simple Linear Regression\n",
        "## The Regression Equation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:24.993943Z",
          "iopub.status.busy": "2021-02-23T15:56:24.993298Z",
          "iopub.status.idle": "2021-02-23T15:56:25.195902Z",
          "shell.execute_reply": "2021-02-23T15:56:25.195434Z"
        },
        "id": "3O9sSYlFmrNR",
        "outputId": "409e6839-25e2-4520-c10d-31ed99b0e366"
      },
      "source": [
        "lung = pd.read_csv(LUNG_CSV)\n",
        "\n",
        "lung.plot.scatter(x='Exposure', y='PEFR')\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "96DzsXrCmrNS"
      },
      "source": [
        "We can use the `LinearRegression` model from _scikit-learn_."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:25.204834Z",
          "iopub.status.busy": "2021-02-23T15:56:25.204170Z",
          "iopub.status.idle": "2021-02-23T15:56:25.207031Z",
          "shell.execute_reply": "2021-02-23T15:56:25.207407Z"
        },
        "id": "hqjbpNWHmrNT",
        "outputId": "ea633a8e-c653-46a1-9b86-cb1b12c75b3c"
      },
      "source": [
        "predictors = ['Exposure']\n",
        "outcome = 'PEFR'\n",
        "\n",
        "model = LinearRegression()\n",
        "model.fit(lung[predictors], lung[outcome])\n",
        "\n",
        "print(f'Intercept: {model.intercept_:.3f}')\n",
        "print(f'Coefficient Exposure: {model.coef_[0]:.3f}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Intercept: 424.583\n",
            "Coefficient Exposure: -4.185\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:25.239844Z",
          "iopub.status.busy": "2021-02-23T15:56:25.232977Z",
          "iopub.status.idle": "2021-02-23T15:56:25.457699Z",
          "shell.execute_reply": "2021-02-23T15:56:25.458014Z"
        },
        "id": "zlaRWl1FmrNT",
        "outputId": "7108351e-ba38-472a-fd47-f0a354d93ba2"
      },
      "source": [
        "fig, ax = plt.subplots(figsize=(4, 4))\n",
        "ax.set_xlim(0, 23)\n",
        "ax.set_ylim(295, 450)\n",
        "ax.set_xlabel('Exposure')\n",
        "ax.set_ylabel('PEFR')\n",
        "ax.plot((0, 23), model.predict([[0], [23]]))\n",
        "ax.text(0.4, model.intercept_, r'$b_0$', size='larger')\n",
        "\n",
        "x = [[7.5], [17.5]]\n",
        "y = model.predict(x)\n",
        "ax.plot((7.5, 7.5, 17.5), (y[0], y[1], y[1]), '--')\n",
        "ax.text(5, np.mean(y), r'$\\Delta Y$', size='larger')\n",
        "ax.text(12, y[1] - 10, r'$\\Delta X$', size='larger')\n",
        "ax.text(12, 390, r'$b_1 = \\frac{\\Delta Y}{\\Delta X}$', size='larger')\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 288x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yR4Fj806mrNU"
      },
      "source": [
        "## Fitted Values and Residuals\n",
        "The method `predict` of a fitted _scikit-learn_ model can be used to predict new data points."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:25.463673Z",
          "iopub.status.busy": "2021-02-23T15:56:25.462925Z",
          "iopub.status.idle": "2021-02-23T15:56:25.466735Z",
          "shell.execute_reply": "2021-02-23T15:56:25.466228Z"
        },
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 190
        },
        "id": "ZAxLYsCXmrNV",
        "outputId": "92c41174-44cb-461a-b64d-c5bd61004bf0"
      },
      "source": [
        "fitted = model.predict(lung[predictors])\n",
        "residuals = lung[outcome] - fitted"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "NameError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-4-793ce5b28b14>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mfitted\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlung\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mpredictors\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0mresiduals\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlung\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0moutcome\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m-\u001b[0m \u001b[0mfitted\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mNameError\u001b[0m: name 'model' is not defined"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:25.480897Z",
          "iopub.status.busy": "2021-02-23T15:56:25.480377Z",
          "iopub.status.idle": "2021-02-23T15:56:25.718636Z",
          "shell.execute_reply": "2021-02-23T15:56:25.718959Z"
        },
        "id": "xzFWBqbumrNV",
        "outputId": "becf2b7e-014c-4790-c7e7-78a5510cd5c2"
      },
      "source": [
        "ax = lung.plot.scatter(x='Exposure', y='PEFR', figsize=(4, 4))\n",
        "ax.plot(lung.Exposure, fitted)\n",
        "for x, yactual, yfitted in zip(lung.Exposure, lung.PEFR, fitted): \n",
        "    ax.plot((x, x), (yactual, yfitted), '--', color='C1')\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 288x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R6csUNVnmrNW"
      },
      "source": [
        "# Multiple linear regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:25.722954Z",
          "iopub.status.busy": "2021-02-23T15:56:25.722440Z",
          "iopub.status.idle": "2021-02-23T15:56:25.791129Z",
          "shell.execute_reply": "2021-02-23T15:56:25.791562Z"
        },
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nymZxT1RmrNW",
        "outputId": "7584482d-a70c-4b02-9689-f41b73573a9a"
      },
      "source": [
        "subset = ['AdjSalePrice', 'SqFtTotLiving', 'SqFtLot', 'Bathrooms', \n",
        "          'Bedrooms', 'BldgGrade']\n",
        "\n",
        "house = pd.read_csv('house_sales.csv', sep='\\t')\n",
        "print(house[subset].head())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "   AdjSalePrice  SqFtTotLiving  SqFtLot  Bathrooms  Bedrooms  BldgGrade\n",
            "1      300805.0           2400     9373       3.00         6          7\n",
            "2     1076162.0           3764    20156       3.75         4         10\n",
            "3      761805.0           2060    26036       1.75         4          8\n",
            "4      442065.0           3200     8618       3.75         5          7\n",
            "5      297065.0           1720     8620       1.75         4          7\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 213
        },
        "id": "oVg_wpyNnhfU",
        "outputId": "a5c51030-3ad6-454e-896b-37b5eead3e34"
      },
      "source": [
        "house.head(3)"
      ],
      "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>DocumentDate</th>\n",
              "      <th>SalePrice</th>\n",
              "      <th>PropertyID</th>\n",
              "      <th>PropertyType</th>\n",
              "      <th>ym</th>\n",
              "      <th>zhvi_px</th>\n",
              "      <th>zhvi_idx</th>\n",
              "      <th>AdjSalePrice</th>\n",
              "      <th>NbrLivingUnits</th>\n",
              "      <th>SqFtLot</th>\n",
              "      <th>SqFtTotLiving</th>\n",
              "      <th>SqFtFinBasement</th>\n",
              "      <th>Bathrooms</th>\n",
              "      <th>Bedrooms</th>\n",
              "      <th>BldgGrade</th>\n",
              "      <th>YrBuilt</th>\n",
              "      <th>YrRenovated</th>\n",
              "      <th>TrafficNoise</th>\n",
              "      <th>LandVal</th>\n",
              "      <th>ImpsVal</th>\n",
              "      <th>ZipCode</th>\n",
              "      <th>NewConstruction</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2014-09-16</td>\n",
              "      <td>280000</td>\n",
              "      <td>1000102</td>\n",
              "      <td>Multiplex</td>\n",
              "      <td>2014-09-01</td>\n",
              "      <td>405100</td>\n",
              "      <td>0.930836</td>\n",
              "      <td>300805.0</td>\n",
              "      <td>2</td>\n",
              "      <td>9373</td>\n",
              "      <td>2400</td>\n",
              "      <td>0</td>\n",
              "      <td>3.00</td>\n",
              "      <td>6</td>\n",
              "      <td>7</td>\n",
              "      <td>1991</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>70000</td>\n",
              "      <td>229000</td>\n",
              "      <td>98002</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2006-06-16</td>\n",
              "      <td>1000000</td>\n",
              "      <td>1200013</td>\n",
              "      <td>Single Family</td>\n",
              "      <td>2006-06-01</td>\n",
              "      <td>404400</td>\n",
              "      <td>0.929228</td>\n",
              "      <td>1076162.0</td>\n",
              "      <td>1</td>\n",
              "      <td>20156</td>\n",
              "      <td>3764</td>\n",
              "      <td>1452</td>\n",
              "      <td>3.75</td>\n",
              "      <td>4</td>\n",
              "      <td>10</td>\n",
              "      <td>2005</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>203000</td>\n",
              "      <td>590000</td>\n",
              "      <td>98166</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2007-01-29</td>\n",
              "      <td>745000</td>\n",
              "      <td>1200019</td>\n",
              "      <td>Single Family</td>\n",
              "      <td>2007-01-01</td>\n",
              "      <td>425600</td>\n",
              "      <td>0.977941</td>\n",
              "      <td>761805.0</td>\n",
              "      <td>1</td>\n",
              "      <td>26036</td>\n",
              "      <td>2060</td>\n",
              "      <td>900</td>\n",
              "      <td>1.75</td>\n",
              "      <td>4</td>\n",
              "      <td>8</td>\n",
              "      <td>1947</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>183000</td>\n",
              "      <td>275000</td>\n",
              "      <td>98166</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  DocumentDate  SalePrice  PropertyID  ... ImpsVal ZipCode  NewConstruction\n",
              "1   2014-09-16     280000     1000102  ...  229000   98002            False\n",
              "2   2006-06-16    1000000     1200013  ...  590000   98166             True\n",
              "3   2007-01-29     745000     1200019  ...  275000   98166            False\n",
              "\n",
              "[3 rows x 22 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:25.797251Z",
          "iopub.status.busy": "2021-02-23T15:56:25.796800Z",
          "iopub.status.idle": "2021-02-23T15:56:25.802102Z",
          "shell.execute_reply": "2021-02-23T15:56:25.801681Z"
        },
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lSM1t2NlmrNW",
        "outputId": "d78a1f63-953c-4695-8e62-8f03a8aac587"
      },
      "source": [
        "predictors = ['SqFtTotLiving', 'SqFtLot', 'Bathrooms', \n",
        "              'Bedrooms', 'BldgGrade']\n",
        "outcome = 'AdjSalePrice'\n",
        "\n",
        "house_lm = LinearRegression()\n",
        "house_lm.fit(house[predictors], house[outcome])\n",
        "\n",
        "print(f'Intercept: {house_lm.intercept_:.3f}')\n",
        "print('Coefficients:')\n",
        "for name, coef in zip(predictors, house_lm.coef_):\n",
        "    print(f' {name}: {coef}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Intercept: -521871.368\n",
            "Coefficients:\n",
            " SqFtTotLiving: 228.83060360240796\n",
            " SqFtLot: -0.06046682065307607\n",
            " Bathrooms: -19442.840398321056\n",
            " Bedrooms: -47769.95518521438\n",
            " BldgGrade: 106106.96307898083\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UzUo9P5-nzoF",
        "outputId": "9e801a7b-5f27-4b58-dc09-eae87da61577"
      },
      "source": [
        "house[outcome]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1         300805.0\n",
              "2        1076162.0\n",
              "3         761805.0\n",
              "4         442065.0\n",
              "5         297065.0\n",
              "           ...    \n",
              "27057     443803.0\n",
              "27058    1586196.0\n",
              "27061     220744.0\n",
              "27062     342207.0\n",
              "27063     475489.0\n",
              "Name: AdjSalePrice, Length: 22687, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4_FJ15q3mrNX"
      },
      "source": [
        "## Assessing the Model\n",
        "_Scikit-learn_ provides a number of metrics to determine the quality of a model. Here we use the `r2_score`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:25.807280Z",
          "iopub.status.busy": "2021-02-23T15:56:25.806848Z",
          "iopub.status.idle": "2021-02-23T15:56:25.816006Z",
          "shell.execute_reply": "2021-02-23T15:56:25.815487Z"
        },
        "id": "Rd6lWVDfmrNX",
        "outputId": "4dfa2982-1c42-445b-bc3b-127cf671d2b6"
      },
      "source": [
        "fitted = house_lm.predict(house[predictors])\n",
        "RMSE = np.sqrt(mean_squared_error(house[outcome], fitted))\n",
        "r2 = r2_score(house[outcome], fitted)\n",
        "print(f'RMSE: {RMSE:.0f}')\n",
        "print(f'r2: {r2:.4f}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "RMSE: 261220\n",
            "r2: 0.5406\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TBZNssxkmrNX"
      },
      "source": [
        "While _scikit-learn_ provides a variety of different metrics, _statsmodels_ provides a more in-depth analysis of the linear regression model. This package has two different ways of specifying the model, one that is similar to _scikit-learn_ and one that allows specifying _R_-style formulas. Here we use the first approach. As _statsmodels_ doesn't add an intercept automaticaly, we need to add a constant column with value 1 to the predictors. We can use the _pandas_ method assign for this."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:25.821880Z",
          "iopub.status.busy": "2021-02-23T15:56:25.821282Z",
          "iopub.status.idle": "2021-02-23T15:56:25.842522Z",
          "shell.execute_reply": "2021-02-23T15:56:25.843266Z"
        },
        "id": "pmXjzxp3mrNX",
        "outputId": "0e64d79b-fc18-487c-e928-560e200b7acf"
      },
      "source": [
        "model = sm.OLS(house[outcome], house[predictors].assign(const=1))\n",
        "results = model.fit()\n",
        "print(results.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "                            OLS Regression Results                            \n",
            "==============================================================================\n",
            "Dep. Variable:           AdjSalePrice   R-squared:                       0.541\n",
            "Model:                            OLS   Adj. R-squared:                  0.540\n",
            "Method:                 Least Squares   F-statistic:                     5338.\n",
            "Date:                Tue, 23 Feb 2021   Prob (F-statistic):               0.00\n",
            "Time:                        15:56:25   Log-Likelihood:            -3.1517e+05\n",
            "No. Observations:               22687   AIC:                         6.304e+05\n",
            "Df Residuals:                   22681   BIC:                         6.304e+05\n",
            "Df Model:                           5                                         \n",
            "Covariance Type:            nonrobust                                         \n",
            "=================================================================================\n",
            "                    coef    std err          t      P>|t|      [0.025      0.975]\n",
            "---------------------------------------------------------------------------------\n",
            "SqFtTotLiving   228.8306      3.899     58.694      0.000     221.189     236.472\n",
            "SqFtLot          -0.0605      0.061     -0.988      0.323      -0.180       0.059\n",
            "Bathrooms     -1.944e+04   3625.388     -5.363      0.000   -2.65e+04   -1.23e+04\n",
            "Bedrooms      -4.777e+04   2489.732    -19.187      0.000   -5.27e+04   -4.29e+04\n",
            "BldgGrade      1.061e+05   2396.445     44.277      0.000    1.01e+05    1.11e+05\n",
            "const         -5.219e+05   1.57e+04    -33.342      0.000   -5.53e+05   -4.91e+05\n",
            "==============================================================================\n",
            "Omnibus:                    29676.557   Durbin-Watson:                   1.247\n",
            "Prob(Omnibus):                  0.000   Jarque-Bera (JB):         19390738.346\n",
            "Skew:                           6.889   Prob(JB):                         0.00\n",
            "Kurtosis:                     145.559   Cond. No.                     2.86e+05\n",
            "==============================================================================\n",
            "\n",
            "Notes:\n",
            "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
            "[2] The condition number is large, 2.86e+05. This might indicate that there are\n",
            "strong multicollinearity or other numerical problems.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Geo7oaVSmrNY"
      },
      "source": [
        "## Model Selection and Stepwise Regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:25.853500Z",
          "iopub.status.busy": "2021-02-23T15:56:25.848218Z",
          "iopub.status.idle": "2021-02-23T15:56:25.889391Z",
          "shell.execute_reply": "2021-02-23T15:56:25.890176Z"
        },
        "id": "iYPtKzOTmrNZ",
        "outputId": "9085654f-5e18-4c82-bf40-3331cc523c1c"
      },
      "source": [
        "predictors = ['SqFtTotLiving', 'SqFtLot', 'Bathrooms', 'Bedrooms',\n",
        "              'BldgGrade', 'PropertyType', 'NbrLivingUnits',\n",
        "              'SqFtFinBasement', 'YrBuilt', 'YrRenovated', \n",
        "              'NewConstruction']\n",
        "\n",
        "X = pd.get_dummies(house[predictors], drop_first=True)\n",
        "X['NewConstruction'] = [1 if nc else 0 for nc in X['NewConstruction']]\n",
        "\n",
        "house_full = sm.OLS(house[outcome], X.assign(const=1))\n",
        "results = house_full.fit()\n",
        "print(results.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "                            OLS Regression Results                            \n",
            "==============================================================================\n",
            "Dep. Variable:           AdjSalePrice   R-squared:                       0.595\n",
            "Model:                            OLS   Adj. R-squared:                  0.594\n",
            "Method:                 Least Squares   F-statistic:                     2771.\n",
            "Date:                Tue, 23 Feb 2021   Prob (F-statistic):               0.00\n",
            "Time:                        15:56:25   Log-Likelihood:            -3.1375e+05\n",
            "No. Observations:               22687   AIC:                         6.275e+05\n",
            "Df Residuals:                   22674   BIC:                         6.276e+05\n",
            "Df Model:                          12                                         \n",
            "Covariance Type:            nonrobust                                         \n",
            "==============================================================================================\n",
            "                                 coef    std err          t      P>|t|      [0.025      0.975]\n",
            "----------------------------------------------------------------------------------------------\n",
            "SqFtTotLiving                198.6364      4.234     46.920      0.000     190.338     206.934\n",
            "SqFtLot                        0.0771      0.058      1.330      0.184      -0.037       0.191\n",
            "Bathrooms                   4.286e+04   3808.114     11.255      0.000    3.54e+04    5.03e+04\n",
            "Bedrooms                   -5.187e+04   2396.904    -21.638      0.000   -5.66e+04   -4.72e+04\n",
            "BldgGrade                   1.373e+05   2441.242     56.228      0.000    1.32e+05    1.42e+05\n",
            "NbrLivingUnits              5723.8438   1.76e+04      0.326      0.744   -2.87e+04    4.01e+04\n",
            "SqFtFinBasement                7.0611      4.627      1.526      0.127      -2.009      16.131\n",
            "YrBuilt                    -3574.2210     77.228    -46.282      0.000   -3725.593   -3422.849\n",
            "YrRenovated                   -2.5311      3.924     -0.645      0.519     -10.222       5.160\n",
            "NewConstruction            -2489.1122   5936.692     -0.419      0.675   -1.41e+04    9147.211\n",
            "PropertyType_Single Family  2.997e+04   2.61e+04      1.149      0.251   -2.12e+04    8.11e+04\n",
            "PropertyType_Townhouse      9.286e+04    2.7e+04      3.438      0.001    3.99e+04    1.46e+05\n",
            "const                       6.182e+06   1.55e+05     39.902      0.000    5.88e+06    6.49e+06\n",
            "==============================================================================\n",
            "Omnibus:                    31006.128   Durbin-Watson:                   1.393\n",
            "Prob(Omnibus):                  0.000   Jarque-Bera (JB):         26251977.078\n",
            "Skew:                           7.427   Prob(JB):                         0.00\n",
            "Kurtosis:                     168.984   Cond. No.                     2.98e+06\n",
            "==============================================================================\n",
            "\n",
            "Notes:\n",
            "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
            "[2] The condition number is large, 2.98e+06. This might indicate that there are\n",
            "strong multicollinearity or other numerical problems.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a78x1coXmrNZ"
      },
      "source": [
        "We can use the `stepwise_selection` method from the _dmba_ package."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:25.896023Z",
          "iopub.status.busy": "2021-02-23T15:56:25.895069Z",
          "iopub.status.idle": "2021-02-23T15:56:26.501169Z",
          "shell.execute_reply": "2021-02-23T15:56:26.501922Z"
        },
        "id": "eRCyPtZxmrNZ",
        "outputId": "3d932c54-3abf-4220-c941-22c78466ad00"
      },
      "source": [
        "y = house[outcome]\n",
        "\n",
        "def train_model(variables):\n",
        "    if len(variables) == 0:\n",
        "        return None\n",
        "    model = LinearRegression()\n",
        "    model.fit(X[variables], y)\n",
        "    return model\n",
        "\n",
        "def score_model(model, variables):\n",
        "    if len(variables) == 0:\n",
        "        return AIC_score(y, [y.mean()] * len(y), model, df=1)\n",
        "    return AIC_score(y, model.predict(X[variables]), model)\n",
        "\n",
        "best_model, best_variables = stepwise_selection(X.columns, train_model, score_model, \n",
        "                                                verbose=True)\n",
        "\n",
        "print()\n",
        "print(f'Intercept: {best_model.intercept_:.3f}')\n",
        "print('Coefficients:')\n",
        "for name, coef in zip(best_variables, best_model.coef_):\n",
        "    print(f' {name}: {coef}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Variables: SqFtTotLiving, SqFtLot, Bathrooms, Bedrooms, BldgGrade, NbrLivingUnits, SqFtFinBasement, YrBuilt, YrRenovated, NewConstruction, PropertyType_Single Family, PropertyType_Townhouse\n",
            "Start: score=647988.32, constant\n",
            "Step: score=633013.35, add SqFtTotLiving\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: score=630793.74, add BldgGrade\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: score=628230.29, add YrBuilt\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: score=627784.16, add Bedrooms\n",
            "Step: score=627602.21, add Bathrooms\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: score=627525.65, add PropertyType_Townhouse\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: score=627525.08, add SqFtFinBasement\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: score=627524.98, add PropertyType_Single Family\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: score=627524.98, unchanged None\n",
            "\n",
            "Intercept: 6178645.017\n",
            "Coefficients:\n",
            " SqFtTotLiving: 199.27755304201781\n",
            " BldgGrade: 137159.56022619802\n",
            " YrBuilt: -3565.4249392492793\n",
            " Bedrooms: -51947.383673613425\n",
            " Bathrooms: 42396.164527718094\n",
            " PropertyType_Townhouse: 84479.16203300333\n",
            " SqFtFinBasement: 7.046974967583083\n",
            " PropertyType_Single Family: 22912.055187017646\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oND5Md9nmrNa"
      },
      "source": [
        "## Weighted regression\n",
        "We can calculate the Year from the date column using either a list comprehension or the data frame's `apply` method."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.515703Z",
          "iopub.status.busy": "2021-02-23T15:56:26.510171Z",
          "iopub.status.idle": "2021-02-23T15:56:26.539501Z",
          "shell.execute_reply": "2021-02-23T15:56:26.540490Z"
        },
        "id": "sKuyPK3_mrNa"
      },
      "source": [
        "house['Year'] = [int(date.split('-')[0]) for date in house.DocumentDate]\n",
        "house['Year'] = house.DocumentDate.apply(lambda d: int(d.split('-')[0]))\n",
        "house['Weight'] = house.Year - 2005"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.546070Z",
          "iopub.status.busy": "2021-02-23T15:56:26.545497Z",
          "iopub.status.idle": "2021-02-23T15:56:26.567338Z",
          "shell.execute_reply": "2021-02-23T15:56:26.568037Z"
        },
        "id": "YIZ25ktfmrNa",
        "outputId": "aeba3f3b-f299-4142-954b-b07e0e427df7"
      },
      "source": [
        "predictors = ['SqFtTotLiving', 'SqFtLot', 'Bathrooms', \n",
        "              'Bedrooms', 'BldgGrade']\n",
        "outcome = 'AdjSalePrice'\n",
        "\n",
        "house_wt = LinearRegression()\n",
        "house_wt.fit(house[predictors], house[outcome], sample_weight=house.Weight)\n",
        "pd.DataFrame({\n",
        "    'predictor': predictors,\n",
        "    'house_lm': house_lm.coef_,\n",
        "    'house_wt': house_wt.coef_,\n",
        "}).append({\n",
        "    'predictor': 'intercept', \n",
        "    'house_lm': house_lm.intercept_,\n",
        "    'house_wt': house_wt.intercept_,\n",
        "}, ignore_index=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "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>predictor</th>\n",
              "      <th>house_lm</th>\n",
              "      <th>house_wt</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>SqFtTotLiving</td>\n",
              "      <td>228.830604</td>\n",
              "      <td>245.024089</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>SqFtLot</td>\n",
              "      <td>-0.060467</td>\n",
              "      <td>-0.292415</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Bathrooms</td>\n",
              "      <td>-19442.840398</td>\n",
              "      <td>-26085.970109</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Bedrooms</td>\n",
              "      <td>-47769.955185</td>\n",
              "      <td>-53608.876436</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>BldgGrade</td>\n",
              "      <td>106106.963079</td>\n",
              "      <td>115242.434726</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>intercept</td>\n",
              "      <td>-521871.368188</td>\n",
              "      <td>-584189.329446</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       predictor       house_lm       house_wt\n",
              "0  SqFtTotLiving     228.830604     245.024089\n",
              "1        SqFtLot      -0.060467      -0.292415\n",
              "2      Bathrooms  -19442.840398  -26085.970109\n",
              "3       Bedrooms  -47769.955185  -53608.876436\n",
              "4      BldgGrade  106106.963079  115242.434726\n",
              "5      intercept -521871.368188 -584189.329446"
            ]
          },
          "execution_count": 1,
          "metadata": {},
          "output_type": "execute_result"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.576014Z",
          "iopub.status.busy": "2021-02-23T15:56:26.575427Z",
          "iopub.status.idle": "2021-02-23T15:56:26.598776Z",
          "shell.execute_reply": "2021-02-23T15:56:26.597664Z"
        },
        "scrolled": false,
        "id": "iVvcEm5nmrNb",
        "outputId": "63e40a1f-50f3-4ca6-f416-5e99fcd71957"
      },
      "source": [
        "residuals = pd.DataFrame({\n",
        "    'abs_residual_lm': np.abs(house_lm.predict(house[predictors]) - house[outcome]),\n",
        "    'abs_residual_wt': np.abs(house_wt.predict(house[predictors]) - house[outcome]),\n",
        "    'Year': house['Year'],\n",
        "})\n",
        "print(residuals.head())\n",
        "# axes = residuals.boxplot(['abs_residual_lm', 'abs_residual_wt'], by='Year', figsize=(10, 4))\n",
        "# axes[0].set_ylim(0, 300000)\n",
        "\n",
        "pd.DataFrame(([year, np.mean(group['abs_residual_lm']), np.mean(group['abs_residual_wt'])] \n",
        "              for year, group in residuals.groupby('Year')),\n",
        "             columns=['Year', 'mean abs_residual_lm', 'mean abs_residual_wt'])\n",
        "# for year, group in residuals.groupby('Year'):\n",
        "#     print(year, np.mean(group['abs_residual_lm']), np.mean(group['abs_residual_wt']))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "   abs_residual_lm  abs_residual_wt  Year\n",
            "1    123750.814194    107108.553965  2014\n",
            "2     59145.413089     96191.882094  2006\n",
            "3    190108.725716    187004.492880  2007\n",
            "4    198788.774412    196132.996857  2008\n",
            "5     91774.996129     84277.577512  2013\n"
          ]
        },
        {
          "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>Year</th>\n",
              "      <th>mean abs_residual_lm</th>\n",
              "      <th>mean abs_residual_wt</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2006</td>\n",
              "      <td>140540.303585</td>\n",
              "      <td>146557.454636</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2007</td>\n",
              "      <td>147747.577959</td>\n",
              "      <td>152848.523235</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2008</td>\n",
              "      <td>142086.905943</td>\n",
              "      <td>146360.411668</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2009</td>\n",
              "      <td>147016.720883</td>\n",
              "      <td>151182.924825</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2010</td>\n",
              "      <td>163267.674885</td>\n",
              "      <td>166364.476152</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>2011</td>\n",
              "      <td>169937.385744</td>\n",
              "      <td>172950.876028</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>2012</td>\n",
              "      <td>169506.670053</td>\n",
              "      <td>171874.424266</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>2013</td>\n",
              "      <td>203659.777510</td>\n",
              "      <td>206242.199403</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>2014</td>\n",
              "      <td>184452.840665</td>\n",
              "      <td>186668.573750</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>2015</td>\n",
              "      <td>172323.435147</td>\n",
              "      <td>169842.742053</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Year  mean abs_residual_lm  mean abs_residual_wt\n",
              "0  2006         140540.303585         146557.454636\n",
              "1  2007         147747.577959         152848.523235\n",
              "2  2008         142086.905943         146360.411668\n",
              "3  2009         147016.720883         151182.924825\n",
              "4  2010         163267.674885         166364.476152\n",
              "5  2011         169937.385744         172950.876028\n",
              "6  2012         169506.670053         171874.424266\n",
              "7  2013         203659.777510         206242.199403\n",
              "8  2014         184452.840665         186668.573750\n",
              "9  2015         172323.435147         169842.742053"
            ]
          },
          "execution_count": 1,
          "metadata": {},
          "output_type": "execute_result"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uhIVrPM7mrNb"
      },
      "source": [
        "# Factor variables in regression\n",
        "## Dummy Variables Representation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.603077Z",
          "iopub.status.busy": "2021-02-23T15:56:26.602141Z",
          "iopub.status.idle": "2021-02-23T15:56:26.608258Z",
          "shell.execute_reply": "2021-02-23T15:56:26.607393Z"
        },
        "id": "0xE1PgjxmrNb",
        "outputId": "6fd08290-a5b1-4e56-a4b2-c27471e5efb3"
      },
      "source": [
        "print(house.PropertyType.head())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "1        Multiplex\n",
            "2    Single Family\n",
            "3    Single Family\n",
            "4    Single Family\n",
            "5    Single Family\n",
            "Name: PropertyType, dtype: object\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.612364Z",
          "iopub.status.busy": "2021-02-23T15:56:26.611753Z",
          "iopub.status.idle": "2021-02-23T15:56:26.618950Z",
          "shell.execute_reply": "2021-02-23T15:56:26.619653Z"
        },
        "id": "ep3nwyT3mrNc",
        "outputId": "c4e22e48-3912-4928-a1d4-a388d8b2341e"
      },
      "source": [
        "print(pd.get_dummies(house['PropertyType']).head(6))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "   Multiplex  Single Family  Townhouse\n",
            "1          1              0          0\n",
            "2          0              1          0\n",
            "3          0              1          0\n",
            "4          0              1          0\n",
            "5          0              1          0\n",
            "6          0              0          1\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.625204Z",
          "iopub.status.busy": "2021-02-23T15:56:26.623540Z",
          "iopub.status.idle": "2021-02-23T15:56:26.631121Z",
          "shell.execute_reply": "2021-02-23T15:56:26.631780Z"
        },
        "id": "cjdL8AZbmrNc",
        "outputId": "e384cf33-133a-4524-f298-2684fe71fdc5"
      },
      "source": [
        "print(pd.get_dummies(house['PropertyType'], drop_first=True).head(6))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "   Single Family  Townhouse\n",
            "1              0          0\n",
            "2              1          0\n",
            "3              1          0\n",
            "4              1          0\n",
            "5              1          0\n",
            "6              0          1\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.634305Z",
          "iopub.status.busy": "2021-02-23T15:56:26.633709Z",
          "iopub.status.idle": "2021-02-23T15:56:26.649822Z",
          "shell.execute_reply": "2021-02-23T15:56:26.650376Z"
        },
        "id": "mxQRCjcTmrNc",
        "outputId": "33ed0693-92fb-4686-f2ad-e3e4f1645116"
      },
      "source": [
        "predictors = ['SqFtTotLiving', 'SqFtLot', 'Bathrooms', 'Bedrooms',\n",
        "              'BldgGrade', 'PropertyType']\n",
        "\n",
        "X = pd.get_dummies(house[predictors], drop_first=True)\n",
        "\n",
        "house_lm_factor = LinearRegression()\n",
        "house_lm_factor.fit(X, house[outcome])\n",
        "\n",
        "print(f'Intercept: {house_lm_factor.intercept_:.3f}')\n",
        "print('Coefficients:')\n",
        "for name, coef in zip(X.columns, house_lm_factor.coef_):\n",
        "    print(f' {name}: {coef}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Intercept: -446841.366\n",
            "Coefficients:\n",
            " SqFtTotLiving: 223.37362892503862\n",
            " SqFtLot: -0.0703679813681326\n",
            " Bathrooms: -15979.01347341495\n",
            " Bedrooms: -50889.73218483036\n",
            " BldgGrade: 109416.30516146174\n",
            " PropertyType_Single Family: -84678.2162954929\n",
            " PropertyType_Townhouse: -115121.97921609206\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RjfuWZF8mrNc"
      },
      "source": [
        "## Factor Variables with many levels"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.653284Z",
          "iopub.status.busy": "2021-02-23T15:56:26.652654Z",
          "iopub.status.idle": "2021-02-23T15:56:26.662874Z",
          "shell.execute_reply": "2021-02-23T15:56:26.663407Z"
        },
        "id": "MX3opCqOmrNc",
        "outputId": "95d92f58-45d7-400a-b06f-878a34910407"
      },
      "source": [
        "print(pd.DataFrame(house['ZipCode'].value_counts()).transpose())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "         98038  98103  98042  98115  98117  98052  98034  98033  98059  98074  \\\n",
            "ZipCode    788    671    641    620    619    614    575    517    513    502   \n",
            "\n",
            "         ...  98051  98024  98354  98050  98288  98057  98224  98113  98043  \\\n",
            "ZipCode  ...     32     31      9      7      4      4      3      1      1   \n",
            "\n",
            "         98068  \n",
            "ZipCode      1  \n",
            "\n",
            "[1 rows x 80 columns]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.666173Z",
          "iopub.status.busy": "2021-02-23T15:56:26.665511Z",
          "iopub.status.idle": "2021-02-23T15:56:26.765941Z",
          "shell.execute_reply": "2021-02-23T15:56:26.767035Z"
        },
        "id": "tdyQvfBEmrNf",
        "outputId": "ee147110-05fc-461a-e271-3ce5637f42d6"
      },
      "source": [
        "house = pd.read_csv(HOUSE_CSV, sep='\\t')\n",
        "\n",
        "predictors = ['SqFtTotLiving', 'SqFtLot', 'Bathrooms', \n",
        "              'Bedrooms', 'BldgGrade']\n",
        "outcome = 'AdjSalePrice'\n",
        "\n",
        "house_lm = LinearRegression()\n",
        "house_lm.fit(house[predictors], house[outcome])\n",
        "\n",
        "\n",
        "zip_groups = pd.DataFrame([\n",
        "    *pd.DataFrame({\n",
        "        'ZipCode': house['ZipCode'],\n",
        "        'residual' : house[outcome] - house_lm.predict(house[predictors]),\n",
        "    })\n",
        "    .groupby(['ZipCode'])\n",
        "    .apply(lambda x: {\n",
        "        'ZipCode': x.iloc[0,0],\n",
        "        'count': len(x),\n",
        "        'median_residual': x.residual.median()\n",
        "    })\n",
        "]).sort_values('median_residual')\n",
        "zip_groups['cum_count'] = np.cumsum(zip_groups['count'])\n",
        "zip_groups['ZipGroup'] = pd.qcut(zip_groups['cum_count'], 5, labels=False, retbins=False)\n",
        "zip_groups.head()\n",
        "print(zip_groups.ZipGroup.value_counts().sort_index())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0    16\n",
            "1    16\n",
            "2    16\n",
            "3    16\n",
            "4    16\n",
            "Name: ZipGroup, dtype: int64\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.770016Z",
          "iopub.status.busy": "2021-02-23T15:56:26.769214Z",
          "iopub.status.idle": "2021-02-23T15:56:26.787245Z",
          "shell.execute_reply": "2021-02-23T15:56:26.787917Z"
        },
        "id": "8QDcsMeSmrNg"
      },
      "source": [
        "to_join = zip_groups[['ZipCode', 'ZipGroup']].set_index('ZipCode')\n",
        "house = house.join(to_join, on='ZipCode')\n",
        "house['ZipGroup'] = house['ZipGroup'].astype('category')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uMOkz3N3mrNg"
      },
      "source": [
        "# Interpreting the Regression Equation\n",
        "## Correlated predictors"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xcKmmKA6mrNg"
      },
      "source": [
        "The results from the stepwise regression are."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.791082Z",
          "iopub.status.busy": "2021-02-23T15:56:26.790442Z",
          "iopub.status.idle": "2021-02-23T15:56:26.795839Z",
          "shell.execute_reply": "2021-02-23T15:56:26.796362Z"
        },
        "id": "-7YHafDnmrNg",
        "outputId": "4450132a-66f8-4c01-e5c5-602008fdf8d2"
      },
      "source": [
        "print(f'Intercept: {best_model.intercept_:.3f}')\n",
        "print('Coefficients:')\n",
        "for name, coef in zip(best_variables, best_model.coef_):\n",
        "    print(f' {name}: {coef}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Intercept: 6178645.017\n",
            "Coefficients:\n",
            " SqFtTotLiving: 199.27755304201781\n",
            " BldgGrade: 137159.56022619802\n",
            " YrBuilt: -3565.4249392492793\n",
            " Bedrooms: -51947.383673613425\n",
            " Bathrooms: 42396.164527718094\n",
            " PropertyType_Townhouse: 84479.16203300333\n",
            " SqFtFinBasement: 7.046974967583083\n",
            " PropertyType_Single Family: 22912.055187017646\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.799128Z",
          "iopub.status.busy": "2021-02-23T15:56:26.798447Z",
          "iopub.status.idle": "2021-02-23T15:56:26.814157Z",
          "shell.execute_reply": "2021-02-23T15:56:26.814801Z"
        },
        "id": "wV295idjmrNh",
        "outputId": "7c0f2b6b-7026-47a5-815d-3b02dea1d19f"
      },
      "source": [
        "predictors = ['Bedrooms', 'BldgGrade', 'PropertyType', 'YrBuilt']\n",
        "outcome = 'AdjSalePrice'\n",
        "\n",
        "X = pd.get_dummies(house[predictors], drop_first=True)\n",
        "\n",
        "reduced_lm = LinearRegression()\n",
        "reduced_lm.fit(X, house[outcome])\n",
        "\n",
        "\n",
        "print(f'Intercept: {reduced_lm.intercept_:.3f}')\n",
        "print('Coefficients:')\n",
        "for name, coef in zip(X.columns, reduced_lm.coef_):\n",
        "    print(f' {name}: {coef}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Intercept: 4913973.344\n",
            "Coefficients:\n",
            " Bedrooms: 27150.537230204864\n",
            " BldgGrade: 248997.7936621382\n",
            " YrBuilt: -3211.7448621550866\n",
            " PropertyType_Single Family: -19898.495340497393\n",
            " PropertyType_Townhouse: -47355.436873339735\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AzRlrTwRmrNh"
      },
      "source": [
        "## Confounding variables"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.818014Z",
          "iopub.status.busy": "2021-02-23T15:56:26.817345Z",
          "iopub.status.idle": "2021-02-23T15:56:26.836376Z",
          "shell.execute_reply": "2021-02-23T15:56:26.837046Z"
        },
        "scrolled": true,
        "id": "ROADYGOwmrNh",
        "outputId": "2b9e26d4-5397-446b-ab6f-7902e4b48260"
      },
      "source": [
        "predictors = ['SqFtTotLiving', 'SqFtLot', 'Bathrooms', 'Bedrooms',\n",
        "              'BldgGrade', 'PropertyType', 'ZipGroup']\n",
        "outcome = 'AdjSalePrice'\n",
        "\n",
        "X = pd.get_dummies(house[predictors], drop_first=True)\n",
        "\n",
        "confounding_lm = LinearRegression()\n",
        "confounding_lm.fit(X, house[outcome])\n",
        "\n",
        "print(f'Intercept: {confounding_lm.intercept_:.3f}')\n",
        "print('Coefficients:')\n",
        "for name, coef in zip(X.columns, confounding_lm.coef_):\n",
        "    print(f' {name}: {coef}')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Intercept: -666637.469\n",
            "Coefficients:\n",
            " SqFtTotLiving: 210.61266005580174\n",
            " SqFtLot: 0.45498713854659556\n",
            " Bathrooms: 5928.425640001798\n",
            " Bedrooms: -41682.87184074474\n",
            " BldgGrade: 98541.18352725956\n",
            " PropertyType_Single Family: 19323.625287919207\n",
            " PropertyType_Townhouse: -78198.72092762376\n",
            " ZipGroup_1: 53317.173306598124\n",
            " ZipGroup_2: 116251.58883563516\n",
            " ZipGroup_3: 178360.53178793335\n",
            " ZipGroup_4: 338408.60185651964\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "be3gnAEjmrNi"
      },
      "source": [
        "## Interactions and Main Effects"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.840138Z",
          "iopub.status.busy": "2021-02-23T15:56:26.839304Z",
          "iopub.status.idle": "2021-02-23T15:56:26.933913Z",
          "shell.execute_reply": "2021-02-23T15:56:26.933270Z"
        },
        "scrolled": false,
        "id": "7X4XEe9MmrNi",
        "outputId": "da6cb4dd-df23-42d1-d43c-ad28e7a33e23"
      },
      "source": [
        "model = smf.ols(formula='AdjSalePrice ~  SqFtTotLiving*ZipGroup + SqFtLot + ' +\n",
        "     'Bathrooms + Bedrooms + BldgGrade + PropertyType', data=house)\n",
        "results = model.fit()\n",
        "print(results.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "                            OLS Regression Results                            \n",
            "==============================================================================\n",
            "Dep. Variable:           AdjSalePrice   R-squared:                       0.682\n",
            "Model:                            OLS   Adj. R-squared:                  0.682\n",
            "Method:                 Least Squares   F-statistic:                     3247.\n",
            "Date:                Tue, 23 Feb 2021   Prob (F-statistic):               0.00\n",
            "Time:                        15:56:26   Log-Likelihood:            -3.1098e+05\n",
            "No. Observations:               22687   AIC:                         6.220e+05\n",
            "Df Residuals:                   22671   BIC:                         6.221e+05\n",
            "Df Model:                          15                                         \n",
            "Covariance Type:            nonrobust                                         \n",
            "=================================================================================================\n",
            "                                    coef    std err          t      P>|t|      [0.025      0.975]\n",
            "-------------------------------------------------------------------------------------------------\n",
            "Intercept                     -4.853e+05   2.05e+04    -23.701      0.000   -5.25e+05   -4.45e+05\n",
            "ZipGroup[T.1]                 -1.113e+04   1.34e+04     -0.830      0.407   -3.74e+04    1.52e+04\n",
            "ZipGroup[T.2]                  2.032e+04   1.18e+04      1.717      0.086   -2877.441    4.35e+04\n",
            "ZipGroup[T.3]                   2.05e+04   1.21e+04      1.697      0.090   -3180.870    4.42e+04\n",
            "ZipGroup[T.4]                 -1.499e+05   1.13e+04    -13.285      0.000   -1.72e+05   -1.28e+05\n",
            "PropertyType[T.Single Family]  1.357e+04   1.39e+04      0.975      0.330   -1.37e+04    4.09e+04\n",
            "PropertyType[T.Townhouse]     -5.884e+04   1.51e+04     -3.888      0.000   -8.85e+04   -2.92e+04\n",
            "SqFtTotLiving                   114.7650      4.863     23.600      0.000     105.233     124.297\n",
            "SqFtTotLiving:ZipGroup[T.1]      32.6043      5.712      5.708      0.000      21.409      43.799\n",
            "SqFtTotLiving:ZipGroup[T.2]      41.7822      5.187      8.056      0.000      31.616      51.948\n",
            "SqFtTotLiving:ZipGroup[T.3]      69.3415      5.619     12.341      0.000      58.329      80.354\n",
            "SqFtTotLiving:ZipGroup[T.4]     226.6836      4.820     47.032      0.000     217.237     236.131\n",
            "SqFtLot                           0.6869      0.052     13.296      0.000       0.586       0.788\n",
            "Bathrooms                     -3619.4533   3202.296     -1.130      0.258   -9896.174    2657.267\n",
            "Bedrooms                       -4.18e+04   2120.279    -19.715      0.000    -4.6e+04   -3.76e+04\n",
            "BldgGrade                      1.047e+05   2069.472     50.592      0.000    1.01e+05    1.09e+05\n",
            "==============================================================================\n",
            "Omnibus:                    30927.394   Durbin-Watson:                   1.581\n",
            "Prob(Omnibus):                  0.000   Jarque-Bera (JB):         34361794.502\n",
            "Skew:                           7.279   Prob(JB):                         0.00\n",
            "Kurtosis:                     193.101   Cond. No.                     5.80e+05\n",
            "==============================================================================\n",
            "\n",
            "Notes:\n",
            "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
            "[2] The condition number is large, 5.8e+05. This might indicate that there are\n",
            "strong multicollinearity or other numerical problems.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pbp8Iwz6mrNi"
      },
      "source": [
        "> Results differ from R due to different binning. Enforcing the same binning gives identical results"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OZRVCiJ-mrNi"
      },
      "source": [
        "# Testing the Assumptions: Regression Diagnostics\n",
        "## Outliers"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZFTMf2jTmrNi"
      },
      "source": [
        "The _statsmodels_ package has the most developed support for outlier analysis. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.949771Z",
          "iopub.status.busy": "2021-02-23T15:56:26.943922Z",
          "iopub.status.idle": "2021-02-23T15:56:26.952555Z",
          "shell.execute_reply": "2021-02-23T15:56:26.953176Z"
        },
        "id": "9qnq70ogmrNi",
        "outputId": "16ff9e8c-c5d0-4519-e0ea-48aa06f88284"
      },
      "source": [
        "house_98105 = house.loc[house['ZipCode'] == 98105, ]\n",
        "\n",
        "predictors = ['SqFtTotLiving', 'SqFtLot', 'Bathrooms', 'Bedrooms',\n",
        "              'BldgGrade']\n",
        "outcome = 'AdjSalePrice'\n",
        "\n",
        "house_outlier = sm.OLS(house_98105[outcome], house_98105[predictors].assign(const=1))\n",
        "result_98105 = house_outlier.fit()\n",
        "print(result_98105.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "                            OLS Regression Results                            \n",
            "==============================================================================\n",
            "Dep. Variable:           AdjSalePrice   R-squared:                       0.795\n",
            "Model:                            OLS   Adj. R-squared:                  0.792\n",
            "Method:                 Least Squares   F-statistic:                     238.7\n",
            "Date:                Tue, 23 Feb 2021   Prob (F-statistic):          1.69e-103\n",
            "Time:                        15:56:26   Log-Likelihood:                -4226.0\n",
            "No. Observations:                 313   AIC:                             8464.\n",
            "Df Residuals:                     307   BIC:                             8486.\n",
            "Df Model:                           5                                         \n",
            "Covariance Type:            nonrobust                                         \n",
            "=================================================================================\n",
            "                    coef    std err          t      P>|t|      [0.025      0.975]\n",
            "---------------------------------------------------------------------------------\n",
            "SqFtTotLiving   209.6023     24.408      8.587      0.000     161.574     257.631\n",
            "SqFtLot          38.9333      5.330      7.305      0.000      28.445      49.421\n",
            "Bathrooms      2282.2641      2e+04      0.114      0.909    -3.7e+04    4.16e+04\n",
            "Bedrooms      -2.632e+04   1.29e+04     -2.043      0.042   -5.17e+04    -973.867\n",
            "BldgGrade        1.3e+05   1.52e+04      8.533      0.000       1e+05     1.6e+05\n",
            "const         -7.725e+05   9.83e+04     -7.861      0.000   -9.66e+05   -5.79e+05\n",
            "==============================================================================\n",
            "Omnibus:                       82.127   Durbin-Watson:                   1.508\n",
            "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              586.561\n",
            "Skew:                           0.859   Prob(JB):                    4.26e-128\n",
            "Kurtosis:                       9.483   Cond. No.                     5.63e+04\n",
            "==============================================================================\n",
            "\n",
            "Notes:\n",
            "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
            "[2] The condition number is large, 5.63e+04. This might indicate that there are\n",
            "strong multicollinearity or other numerical problems.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "urjW82dymrNj"
      },
      "source": [
        "The `OLSInfluence` class is initialized with the OLS regression results and gives access to a number of usefule properties. Here we use the studentized residuals. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.958979Z",
          "iopub.status.busy": "2021-02-23T15:56:26.958097Z",
          "iopub.status.idle": "2021-02-23T15:56:26.963288Z",
          "shell.execute_reply": "2021-02-23T15:56:26.963803Z"
        },
        "id": "xwEYXF9_mrNj",
        "outputId": "13bb0c21-c53e-46e2-faaf-e61cb137f337"
      },
      "source": [
        "influence = OLSInfluence(result_98105)\n",
        "sresiduals = influence.resid_studentized_internal\n",
        "\n",
        "print(sresiduals.idxmin(), sresiduals.min())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "24333 -4.3267318040785625\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.967792Z",
          "iopub.status.busy": "2021-02-23T15:56:26.967183Z",
          "iopub.status.idle": "2021-02-23T15:56:26.971390Z",
          "shell.execute_reply": "2021-02-23T15:56:26.970994Z"
        },
        "id": "oTB_OZS_mrNj",
        "outputId": "030ca8c9-103e-466e-e7f9-1422d7adb736"
      },
      "source": [
        "print(result_98105.resid.loc[sresiduals.idxmin()])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "-757753.6192115827\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.976449Z",
          "iopub.status.busy": "2021-02-23T15:56:26.975944Z",
          "iopub.status.idle": "2021-02-23T15:56:26.980813Z",
          "shell.execute_reply": "2021-02-23T15:56:26.981163Z"
        },
        "id": "8-1UWf9RmrNj",
        "outputId": "0289e249-6a89-4567-fef6-58cd9c77162b"
      },
      "source": [
        "outlier = house_98105.loc[sresiduals.idxmin(), :]\n",
        "print('AdjSalePrice', outlier[outcome])\n",
        "print(outlier[predictors])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "AdjSalePrice 119748.0\n",
            "SqFtTotLiving    2900\n",
            "SqFtLot          7276\n",
            "Bathrooms         3.0\n",
            "Bedrooms            6\n",
            "BldgGrade           7\n",
            "Name: 24333, dtype: object\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EhDAuOjVmrNj"
      },
      "source": [
        "## Influential values"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:26.990236Z",
          "iopub.status.busy": "2021-02-23T15:56:26.989703Z",
          "iopub.status.idle": "2021-02-23T15:56:27.103197Z",
          "shell.execute_reply": "2021-02-23T15:56:27.102631Z"
        },
        "id": "hvPl86iJmrNk",
        "outputId": "6eafa3e1-1fb9-4ece-8bee-de7ae1cd2ae9"
      },
      "source": [
        "%matplotlib inline\n",
        "from scipy.stats import linregress\n",
        "\n",
        "np.random.seed(5)\n",
        "x = np.random.normal(size=25)\n",
        "y = -x / 5 + np.random.normal(size=25)\n",
        "x[0] = 8\n",
        "y[0] = 8\n",
        "\n",
        "def abline(slope, intercept, ax):\n",
        "    \"\"\"Calculate coordinates of a line based on slope and intercept\"\"\"\n",
        "    x_vals = np.array(ax.get_xlim())\n",
        "    return (x_vals, intercept + slope * x_vals)\n",
        "\n",
        "fig, ax = plt.subplots(figsize=(4, 4))\n",
        "ax.scatter(x, y)\n",
        "slope, intercept, _, _, _ = linregress(x, y)\n",
        "ax.plot(*abline(slope, intercept, ax))\n",
        "slope, intercept, _, _, _ = linregress(x[1:], y[1:])\n",
        "ax.plot(*abline(slope, intercept, ax), '--')\n",
        "ax.set_xlim(-2.5, 8.5)\n",
        "ax.set_ylim(-2.5, 8.5)\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 288x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PVyBuz-WmrNk"
      },
      "source": [
        "The package _statsmodel_ provides a number of plots to analyze the data point influence"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:27.110432Z",
          "iopub.status.busy": "2021-02-23T15:56:27.109970Z",
          "iopub.status.idle": "2021-02-23T15:56:27.212159Z",
          "shell.execute_reply": "2021-02-23T15:56:27.211825Z"
        },
        "id": "qrwppi_kmrNk",
        "outputId": "722266c5-ae34-4ba0-c008-c00defa910d4"
      },
      "source": [
        "influence = OLSInfluence(result_98105)\n",
        "fig, ax = plt.subplots(figsize=(5, 5))\n",
        "ax.axhline(-2.5, linestyle='--', color='C1')\n",
        "ax.axhline(2.5, linestyle='--', color='C1')\n",
        "ax.scatter(influence.hat_matrix_diag, influence.resid_studentized_internal, \n",
        "           s=1000 * np.sqrt(influence.cooks_distance[0]),\n",
        "           alpha=0.5)\n",
        "\n",
        "ax.set_xlabel('hat values')\n",
        "ax.set_ylabel('studentized residuals')\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:27.222952Z",
          "iopub.status.busy": "2021-02-23T15:56:27.222361Z",
          "iopub.status.idle": "2021-02-23T15:56:27.226467Z",
          "shell.execute_reply": "2021-02-23T15:56:27.226137Z"
        },
        "id": "sa-i3l-NmrNk",
        "outputId": "e51bb160-265a-47a6-9d12-181fc327acdf"
      },
      "source": [
        "mask = [dist < .08 for dist in influence.cooks_distance[0]]\n",
        "house_infl = house_98105.loc[mask]\n",
        "\n",
        "ols_infl = sm.OLS(house_infl[outcome], house_infl[predictors].assign(const=1))\n",
        "result_infl = ols_infl.fit()\n",
        "\n",
        "pd.DataFrame({\n",
        "    'Original': result_98105.params,\n",
        "    'Influential removed': result_infl.params,\n",
        "})"
      ],
      "execution_count": null,
      "outputs": [
        {
          "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>Original</th>\n",
              "      <th>Influential removed</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>SqFtTotLiving</th>\n",
              "      <td>209.602346</td>\n",
              "      <td>230.052569</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>SqFtLot</th>\n",
              "      <td>38.933315</td>\n",
              "      <td>33.141600</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Bathrooms</th>\n",
              "      <td>2282.264145</td>\n",
              "      <td>-16131.879785</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Bedrooms</th>\n",
              "      <td>-26320.268796</td>\n",
              "      <td>-22887.865318</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>BldgGrade</th>\n",
              "      <td>130000.099737</td>\n",
              "      <td>114870.559737</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>const</th>\n",
              "      <td>-772549.862447</td>\n",
              "      <td>-647137.096716</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                    Original  Influential removed\n",
              "SqFtTotLiving     209.602346           230.052569\n",
              "SqFtLot            38.933315            33.141600\n",
              "Bathrooms        2282.264145        -16131.879785\n",
              "Bedrooms       -26320.268796        -22887.865318\n",
              "BldgGrade      130000.099737        114870.559737\n",
              "const         -772549.862447       -647137.096716"
            ]
          },
          "execution_count": 1,
          "metadata": {},
          "output_type": "execute_result"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DG2TSr82mrNk"
      },
      "source": [
        "## Heteroskedasticity, Non-Normality and Correlated Errors"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4LlKBqmJmrNl"
      },
      "source": [
        "The `regplot` in _seaborn_ allows adding a lowess smoothing line to the scatterplot."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:27.245890Z",
          "iopub.status.busy": "2021-02-23T15:56:27.243835Z",
          "iopub.status.idle": "2021-02-23T15:56:27.404224Z",
          "shell.execute_reply": "2021-02-23T15:56:27.403713Z"
        },
        "id": "vm7YmBcpmrNl",
        "outputId": "3d81e25d-bc3a-4ae5-b4f0-3e15881a8d5a"
      },
      "source": [
        "fig, ax = plt.subplots(figsize=(5, 5))\n",
        "sns.regplot(x=result_98105.fittedvalues, y=np.abs(result_98105.resid), \n",
        "            scatter_kws={'alpha': 0.25},\n",
        "            line_kws={'color': 'C1'},\n",
        "            lowess=True, ax=ax)\n",
        "ax.set_xlabel('predicted')\n",
        "ax.set_ylabel('abs(residual)')\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:27.408200Z",
          "iopub.status.busy": "2021-02-23T15:56:27.407745Z",
          "iopub.status.idle": "2021-02-23T15:56:27.527820Z",
          "shell.execute_reply": "2021-02-23T15:56:27.527050Z"
        },
        "id": "YWk--VemmrNl",
        "outputId": "fc8ac6ba-2422-499b-b13b-6f3e3436cc50"
      },
      "source": [
        "fig, ax = plt.subplots(figsize=(4, 4))\n",
        "pd.Series(influence.resid_studentized_internal).hist(ax=ax)\n",
        "ax.set_xlabel('std. residual')\n",
        "ax.set_ylabel('Frequency')\n",
        "\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 288x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XxNsCK2YmrNl"
      },
      "source": [
        "## Partial Residual Plots and Nonlinearity"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:27.590455Z",
          "iopub.status.busy": "2021-02-23T15:56:27.532234Z",
          "iopub.status.idle": "2021-02-23T15:56:27.690812Z",
          "shell.execute_reply": "2021-02-23T15:56:27.690361Z"
        },
        "id": "igwD3xr0mrNl",
        "outputId": "5641f3fd-f083-4842-a1d9-354076f1b542"
      },
      "source": [
        "fig, ax = plt.subplots(figsize=(5, 5))\n",
        "fig = sm.graphics.plot_ccpr(result_98105, 'SqFtTotLiving', ax=ax)\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:27.704919Z",
          "iopub.status.busy": "2021-02-23T15:56:27.704440Z",
          "iopub.status.idle": "2021-02-23T15:56:28.197463Z",
          "shell.execute_reply": "2021-02-23T15:56:28.196888Z"
        },
        "id": "wbW3WWgrmrNl",
        "outputId": "f0dd8ee9-f802-403d-ce38-da62811e0ef0"
      },
      "source": [
        "fig = plt.figure(figsize=(8, 12))\n",
        "fig = sm.graphics.plot_ccpr_grid(result_98105, fig=fig)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x864 with 5 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fAbWrmPMmrNl"
      },
      "source": [
        "## Polynomial and Spline Regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:28.211182Z",
          "iopub.status.busy": "2021-02-23T15:56:28.210179Z",
          "iopub.status.idle": "2021-02-23T15:56:28.221277Z",
          "shell.execute_reply": "2021-02-23T15:56:28.220726Z"
        },
        "id": "5GVm9WhZmrNm",
        "outputId": "832fbb92-e96a-412b-c46a-f4851572f97d"
      },
      "source": [
        "model_poly = smf.ols(formula='AdjSalePrice ~  SqFtTotLiving + np.power(SqFtTotLiving, 2) + ' + \n",
        "                'SqFtLot + Bathrooms + Bedrooms + BldgGrade', data=house_98105)\n",
        "result_poly = model_poly.fit()\n",
        "print(result_poly.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "                            OLS Regression Results                            \n",
            "==============================================================================\n",
            "Dep. Variable:           AdjSalePrice   R-squared:                       0.806\n",
            "Model:                            OLS   Adj. R-squared:                  0.802\n",
            "Method:                 Least Squares   F-statistic:                     211.6\n",
            "Date:                Tue, 23 Feb 2021   Prob (F-statistic):          9.95e-106\n",
            "Time:                        15:56:28   Log-Likelihood:                -4217.9\n",
            "No. Observations:                 313   AIC:                             8450.\n",
            "Df Residuals:                     306   BIC:                             8476.\n",
            "Df Model:                           6                                         \n",
            "Covariance Type:            nonrobust                                         \n",
            "==============================================================================================\n",
            "                                 coef    std err          t      P>|t|      [0.025      0.975]\n",
            "----------------------------------------------------------------------------------------------\n",
            "Intercept                  -6.159e+05   1.03e+05     -5.953      0.000   -8.19e+05   -4.12e+05\n",
            "SqFtTotLiving                  7.4521     55.418      0.134      0.893    -101.597     116.501\n",
            "np.power(SqFtTotLiving, 2)     0.0388      0.010      4.040      0.000       0.020       0.058\n",
            "SqFtLot                       32.5594      5.436      5.990      0.000      21.863      43.256\n",
            "Bathrooms                  -1435.1231   1.95e+04     -0.074      0.941   -3.99e+04     3.7e+04\n",
            "Bedrooms                   -9191.9441   1.33e+04     -0.693      0.489   -3.53e+04    1.69e+04\n",
            "BldgGrade                   1.357e+05   1.49e+04      9.087      0.000    1.06e+05    1.65e+05\n",
            "==============================================================================\n",
            "Omnibus:                       75.161   Durbin-Watson:                   1.625\n",
            "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              637.978\n",
            "Skew:                           0.699   Prob(JB):                    2.92e-139\n",
            "Kurtosis:                       9.853   Cond. No.                     7.37e+07\n",
            "==============================================================================\n",
            "\n",
            "Notes:\n",
            "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
            "[2] The condition number is large, 7.37e+07. This might indicate that there are\n",
            "strong multicollinearity or other numerical problems.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "895NJZofmrNm"
      },
      "source": [
        "The statsmodels implementation of a partial residual plot works only for linear term. Here is an implementation of a partial residual plot that, while inefficient, works for the polynomial regression."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:28.235963Z",
          "iopub.status.busy": "2021-02-23T15:56:28.235471Z",
          "iopub.status.idle": "2021-02-23T15:56:28.396107Z",
          "shell.execute_reply": "2021-02-23T15:56:28.396468Z"
        },
        "scrolled": false,
        "id": "JfYzDfaVmrNm",
        "outputId": "0216b7d5-acf6-45f4-e083-0f29838e5444"
      },
      "source": [
        "def partialResidualPlot(model, df, outcome, feature, ax):\n",
        "    y_pred = model.predict(df)\n",
        "    copy_df = df.copy()\n",
        "    for c in copy_df.columns:\n",
        "        if c == feature:\n",
        "            continue\n",
        "        copy_df[c] = 0.0\n",
        "    feature_prediction = model.predict(copy_df)\n",
        "    results = pd.DataFrame({\n",
        "        'feature': df[feature],\n",
        "        'residual': df[outcome] - y_pred,\n",
        "        'ypartial': feature_prediction - model.params[0],\n",
        "    })\n",
        "    results = results.sort_values(by=['feature'])\n",
        "    smoothed = sm.nonparametric.lowess(results.ypartial, results.feature, frac=1/3)\n",
        "    \n",
        "    ax.scatter(results.feature, results.ypartial + results.residual)\n",
        "    ax.plot(smoothed[:, 0], smoothed[:, 1], color='gray')\n",
        "    ax.plot(results.feature, results.ypartial, color='black')\n",
        "    ax.set_xlabel(feature)\n",
        "    ax.set_ylabel(f'Residual + {feature} contribution')\n",
        "    return ax\n",
        "\n",
        "fig, ax = plt.subplots(figsize=(5, 5))\n",
        "partialResidualPlot(result_poly, house_98105, 'AdjSalePrice', 'SqFtTotLiving', ax)\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()\n",
        "print(result_poly.params[2])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0.03879128168236527\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nJJ8lAVmmrNm"
      },
      "source": [
        "## Splines"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:28.407649Z",
          "iopub.status.busy": "2021-02-23T15:56:28.406803Z",
          "iopub.status.idle": "2021-02-23T15:56:28.428989Z",
          "shell.execute_reply": "2021-02-23T15:56:28.429432Z"
        },
        "id": "BsGALAqQmrNm",
        "outputId": "78360a8f-a084-44b8-d3f2-85d199121d3b"
      },
      "source": [
        "formula = ('AdjSalePrice ~ bs(SqFtTotLiving, df=6, degree=3) + ' + \n",
        "           'SqFtLot + Bathrooms + Bedrooms + BldgGrade')\n",
        "model_spline = smf.ols(formula=formula, data=house_98105)\n",
        "result_spline = model_spline.fit()\n",
        "print(result_spline.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "                            OLS Regression Results                            \n",
            "==============================================================================\n",
            "Dep. Variable:           AdjSalePrice   R-squared:                       0.814\n",
            "Model:                            OLS   Adj. R-squared:                  0.807\n",
            "Method:                 Least Squares   F-statistic:                     131.8\n",
            "Date:                Tue, 23 Feb 2021   Prob (F-statistic):          7.10e-104\n",
            "Time:                        15:56:28   Log-Likelihood:                -4211.4\n",
            "No. Observations:                 313   AIC:                             8445.\n",
            "Df Residuals:                     302   BIC:                             8486.\n",
            "Df Model:                          10                                         \n",
            "Covariance Type:            nonrobust                                         \n",
            "========================================================================================================\n",
            "                                           coef    std err          t      P>|t|      [0.025      0.975]\n",
            "--------------------------------------------------------------------------------------------------------\n",
            "Intercept                            -4.142e+05   1.43e+05     -2.899      0.004   -6.95e+05   -1.33e+05\n",
            "bs(SqFtTotLiving, df=6, degree=3)[0] -1.995e+05   1.86e+05     -1.076      0.283   -5.65e+05    1.66e+05\n",
            "bs(SqFtTotLiving, df=6, degree=3)[1] -1.206e+05   1.23e+05     -0.983      0.326   -3.62e+05    1.21e+05\n",
            "bs(SqFtTotLiving, df=6, degree=3)[2] -7.164e+04   1.36e+05     -0.525      0.600    -3.4e+05    1.97e+05\n",
            "bs(SqFtTotLiving, df=6, degree=3)[3]  1.957e+05   1.62e+05      1.212      0.227   -1.22e+05    5.14e+05\n",
            "bs(SqFtTotLiving, df=6, degree=3)[4]  8.452e+05   2.18e+05      3.878      0.000    4.16e+05    1.27e+06\n",
            "bs(SqFtTotLiving, df=6, degree=3)[5]  6.955e+05   2.14e+05      3.255      0.001    2.75e+05    1.12e+06\n",
            "SqFtLot                                 33.3258      5.454      6.110      0.000      22.592      44.059\n",
            "Bathrooms                            -4778.2080   1.94e+04     -0.246      0.806    -4.3e+04    3.34e+04\n",
            "Bedrooms                             -5778.7045   1.32e+04     -0.437      0.663   -3.18e+04    2.03e+04\n",
            "BldgGrade                             1.345e+05   1.52e+04      8.842      0.000    1.05e+05    1.64e+05\n",
            "==============================================================================\n",
            "Omnibus:                       58.816   Durbin-Watson:                   1.633\n",
            "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              622.021\n",
            "Skew:                           0.330   Prob(JB):                    8.51e-136\n",
            "Kurtosis:                       9.874   Cond. No.                     1.97e+05\n",
            "==============================================================================\n",
            "\n",
            "Notes:\n",
            "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
            "[2] The condition number is large, 1.97e+05. This might indicate that there are\n",
            "strong multicollinearity or other numerical problems.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:28.445814Z",
          "iopub.status.busy": "2021-02-23T15:56:28.445227Z",
          "iopub.status.idle": "2021-02-23T15:56:28.611508Z",
          "shell.execute_reply": "2021-02-23T15:56:28.610935Z"
        },
        "id": "8HABZWePmrNm",
        "outputId": "d1aa521e-60d2-4c1c-ec65-a0af9b070b5b"
      },
      "source": [
        "fig, ax = plt.subplots(figsize=(5, 5))\n",
        "partialResidualPlot(result_spline, house_98105, 'AdjSalePrice', 'SqFtTotLiving', ax)\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1tD_0WbQmrNn"
      },
      "source": [
        "## Generalized Additive Models"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:28.621500Z",
          "iopub.status.busy": "2021-02-23T15:56:28.620964Z",
          "iopub.status.idle": "2021-02-23T15:56:28.783393Z",
          "shell.execute_reply": "2021-02-23T15:56:28.783932Z"
        },
        "id": "cFzzRGcOmrNn",
        "outputId": "ab7b2c2c-b218-4009-de51-d72c242f68f5"
      },
      "source": [
        "predictors = ['SqFtTotLiving', 'SqFtLot', 'Bathrooms', \n",
        "              'Bedrooms', 'BldgGrade']\n",
        "outcome = 'AdjSalePrice'\n",
        "X = house_98105[predictors].values\n",
        "y = house_98105[outcome]\n",
        "\n",
        "## model\n",
        "gam = LinearGAM(s(0, n_splines=12) + l(1) + l(2) + l(3) + l(4))\n",
        "gam.gridsearch(X, y)\n",
        "print(gam.summary())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\r",
            "N/A% (0 of 11) |                         | Elapsed Time: 0:00:00 ETA:  --:--:--"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\r",
            " 36% (4 of 11) |#########                | Elapsed Time: 0:00:00 ETA:  00:00:00"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\r",
            " 72% (8 of 11) |##################       | Elapsed Time: 0:00:00 ETA:   0:00:00"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\r",
            "100% (11 of 11) |########################| Elapsed Time: 0:00:00 Time:  0:00:00"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "LinearGAM                                                                                                 \n",
            "=============================================== ==========================================================\n",
            "Distribution:                        NormalDist Effective DoF:                                      7.6772\n",
            "Link Function:                     IdentityLink Log Likelihood:                                 -7833.1159\n",
            "Number of Samples:                          313 AIC:                                            15683.5863\n",
            "                                                AICc:                                             15684.14\n",
            "                                                GCV:                                       30838885095.172\n",
            "                                                Scale:                                    29480381715.8334\n",
            "                                                Pseudo R-Squared:                                   0.8117\n",
            "==========================================================================================================\n",
            "Feature Function                  Lambda               Rank         EDoF         P > x        Sig. Code   \n",
            "================================= ==================== ============ ============ ============ ============\n",
            "s(0)                              [15.8489]            12           4.3          1.11e-16     ***         \n",
            "l(1)                              [15.8489]            1            0.9          2.35e-10     ***         \n",
            "l(2)                              [15.8489]            1            0.8          8.45e-01                 \n",
            "l(3)                              [15.8489]            1            0.9          3.79e-01                 \n",
            "l(4)                              [15.8489]            1            0.8          1.11e-16     ***         \n",
            "intercept                                              1            0.0          9.14e-01                 \n",
            "==========================================================================================================\n",
            "Significance codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n",
            "\n",
            "WARNING: Fitting splines and a linear function to a feature introduces a model identifiability problem\n",
            "         which can cause p-values to appear significant when they are not.\n",
            "\n",
            "WARNING: p-values calculated in this manner behave correctly for un-penalized models or models with\n",
            "         known smoothing parameters, but when smoothing parameters have been estimated, the p-values\n",
            "         are typically lower than they should be, meaning that the tests reject the null too readily.\n",
            "None\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-1-2454e06428fa>:10: UserWarning: KNOWN BUG: p-values computed in this summary are likely much smaller than they should be. \n",
            " \n",
            "Please do not make inferences based on these values! \n",
            "\n",
            "Collaborate on a solution, and stay up to date at: \n",
            "github.com/dswah/pyGAM/issues/163 \n",
            "\n",
            "  print(gam.summary())\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:28.801985Z",
          "iopub.status.busy": "2021-02-23T15:56:28.801360Z",
          "iopub.status.idle": "2021-02-23T15:56:29.242609Z",
          "shell.execute_reply": "2021-02-23T15:56:29.241958Z"
        },
        "id": "c3lC1zNPmrNn",
        "outputId": "6b75eb9d-932a-4664-c6fa-26a4c18a2099"
      },
      "source": [
        "fig, axes = plt.subplots(figsize=(8, 8), ncols=2, nrows=3)\n",
        "\n",
        "titles = ['SqFtTotLiving', 'SqFtLot', 'Bathrooms', 'Bedrooms', 'BldgGrade']\n",
        "for i, title in enumerate(titles):\n",
        "    ax = axes[i // 2, i % 2]\n",
        "    XX = gam.generate_X_grid(term=i)\n",
        "    ax.plot(XX[:, i], gam.partial_dependence(term=i, X=XX))\n",
        "    ax.plot(XX[:, i], gam.partial_dependence(term=i, X=XX, width=.95)[1], c='r', ls='--')\n",
        "    ax.set_title(titles[i]);\n",
        "    \n",
        "axes[2][1].set_visible(False)\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x576 with 6 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZFxmveuemrNn"
      },
      "source": [
        "# Additional material - not in book\n",
        "# Regularization\n",
        "## Lasso"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:29.245755Z",
          "iopub.status.busy": "2021-02-23T15:56:29.245211Z",
          "iopub.status.idle": "2021-02-23T15:56:29.248309Z",
          "shell.execute_reply": "2021-02-23T15:56:29.247811Z"
        },
        "id": "VEJoDXiMmrNn"
      },
      "source": [
        "from sklearn.linear_model import Lasso, LassoLars, LassoCV, LassoLarsCV\n",
        "from sklearn.preprocessing import StandardScaler"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:29.252866Z",
          "iopub.status.busy": "2021-02-23T15:56:29.252246Z",
          "iopub.status.idle": "2021-02-23T15:56:29.293637Z",
          "shell.execute_reply": "2021-02-23T15:56:29.293248Z"
        },
        "id": "rfCdmaJrmrNn",
        "outputId": "774b6a60-db6b-498b-f2ca-5161d2107156"
      },
      "source": [
        "subset = ['AdjSalePrice', 'SqFtTotLiving', 'SqFtLot', 'Bathrooms', \n",
        "          'Bedrooms', 'BldgGrade']\n",
        "\n",
        "house = pd.read_csv(HOUSE_CSV, sep='\\t')\n",
        "print(house[subset].head())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "   AdjSalePrice  SqFtTotLiving  SqFtLot  Bathrooms  Bedrooms  BldgGrade\n",
            "1      300805.0           2400     9373       3.00         6          7\n",
            "2     1076162.0           3764    20156       3.75         4         10\n",
            "3      761805.0           2060    26036       1.75         4          8\n",
            "4      442065.0           3200     8618       3.75         5          7\n",
            "5      297065.0           1720     8620       1.75         4          7\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:29.299761Z",
          "iopub.status.busy": "2021-02-23T15:56:29.299227Z",
          "iopub.status.idle": "2021-02-23T15:56:29.319612Z",
          "shell.execute_reply": "2021-02-23T15:56:29.319104Z"
        },
        "id": "6GE5dcPEmrNo",
        "outputId": "ea885099-8010-44f0-b5eb-fe5ce274b5cf"
      },
      "source": [
        "predictors = ['SqFtTotLiving', 'SqFtLot', 'Bathrooms', 'Bedrooms',\n",
        "              'BldgGrade', 'PropertyType', 'NbrLivingUnits',\n",
        "              'SqFtFinBasement', 'YrBuilt', 'YrRenovated', \n",
        "              'NewConstruction']\n",
        "outcome = 'AdjSalePrice'\n",
        "\n",
        "X = pd.get_dummies(house[predictors], drop_first=True)\n",
        "X['NewConstruction'] = [1 if nc else 0 for nc in X['NewConstruction']]\n",
        "columns = X.columns\n",
        "# X = StandardScaler().fit_transform(X * 1.0)\n",
        "y = house[outcome]\n",
        "\n",
        "house_lm = LinearRegression()\n",
        "print(house_lm.fit(X, y))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "LinearRegression()\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:29.323849Z",
          "iopub.status.busy": "2021-02-23T15:56:29.323397Z",
          "iopub.status.idle": "2021-02-23T15:56:29.374657Z",
          "shell.execute_reply": "2021-02-23T15:56:29.373327Z"
        },
        "id": "NxopOCpqmrNo",
        "outputId": "cdde5964-3845-49b0-e1dc-eb5d943e82c3"
      },
      "source": [
        "house_lasso = Lasso(alpha=10)\n",
        "print(house_lasso.fit(X, y))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Lasso(alpha=10)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:29.383613Z",
          "iopub.status.busy": "2021-02-23T15:56:29.382782Z",
          "iopub.status.idle": "2021-02-23T15:56:30.038177Z",
          "shell.execute_reply": "2021-02-23T15:56:30.038557Z"
        },
        "id": "QULQFPOnmrNo",
        "outputId": "7b8c1fdc-d216-4d80-9633-e81858395343"
      },
      "source": [
        "Method = LassoLars\n",
        "MethodCV = LassoLarsCV\n",
        "Method = Lasso\n",
        "MethodCV = LassoCV\n",
        "\n",
        "alpha_values = []\n",
        "results = []\n",
        "for alpha in [0.001, 0.01, 0.1, 1, 10, 100, 1000, 10000, 100000, 1000000, 10000000]:\n",
        "    model = Method(alpha=alpha)\n",
        "    model.fit(X, y)\n",
        "    alpha_values.append(alpha)\n",
        "    results.append(model.coef_)\n",
        "modelCV = MethodCV(cv=5)\n",
        "modelCV.fit(X, y)\n",
        "ax = pd.DataFrame(results, index=alpha_values, columns=columns).plot(logx=True, legend=False)\n",
        "ax.axvline(modelCV.alpha_)\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:30.046882Z",
          "iopub.status.busy": "2021-02-23T15:56:30.046305Z",
          "iopub.status.idle": "2021-02-23T15:56:30.049376Z",
          "shell.execute_reply": "2021-02-23T15:56:30.049771Z"
        },
        "id": "jy5QNmMPmrNo",
        "outputId": "1da7a756-abc5-429f-8468-7bc4516af701"
      },
      "source": [
        "pd.DataFrame({\n",
        "    'name': columns,\n",
        "    'coef': modelCV.coef_, \n",
        "})"
      ],
      "execution_count": null,
      "outputs": [
        {
          "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>name</th>\n",
              "      <th>coef</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>SqFtTotLiving</td>\n",
              "      <td>289.048846</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>SqFtLot</td>\n",
              "      <td>0.029471</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Bathrooms</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Bedrooms</td>\n",
              "      <td>-0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>BldgGrade</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>NbrLivingUnits</td>\n",
              "      <td>-0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>SqFtFinBasement</td>\n",
              "      <td>3.316479</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>YrBuilt</td>\n",
              "      <td>-0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>YrRenovated</td>\n",
              "      <td>45.727472</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>NewConstruction</td>\n",
              "      <td>-0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>PropertyType_Single Family</td>\n",
              "      <td>-0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>PropertyType_Townhouse</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                          name        coef\n",
              "0                SqFtTotLiving  289.048846\n",
              "1                      SqFtLot    0.029471\n",
              "2                    Bathrooms    0.000000\n",
              "3                     Bedrooms   -0.000000\n",
              "4                    BldgGrade    0.000000\n",
              "5               NbrLivingUnits   -0.000000\n",
              "6              SqFtFinBasement    3.316479\n",
              "7                      YrBuilt   -0.000000\n",
              "8                  YrRenovated   45.727472\n",
              "9              NewConstruction   -0.000000\n",
              "10  PropertyType_Single Family   -0.000000\n",
              "11      PropertyType_Townhouse    0.000000"
            ]
          },
          "execution_count": 1,
          "metadata": {},
          "output_type": "execute_result"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "execution": {
          "iopub.execute_input": "2021-02-23T15:56:30.053603Z",
          "iopub.status.busy": "2021-02-23T15:56:30.053056Z",
          "iopub.status.idle": "2021-02-23T15:56:30.055694Z",
          "shell.execute_reply": "2021-02-23T15:56:30.055279Z"
        },
        "id": "2U57GxW6mrNo"
      },
      "source": [
        "# Intercept: 6177658.144\n",
        "# Coefficients:\n",
        "#  SqFtTotLiving: 199.27474217544048\n",
        "#  BldgGrade: 137181.13724627026\n",
        "#  YrBuilt: -3564.934870415041\n",
        "#  Bedrooms: -51974.76845567939\n",
        "#  Bathrooms: 42403.059999677665\n",
        "#  PropertyType_Townhouse: 84378.9333363999\n",
        "#  SqFtFinBasement: 7.032178917565108\n",
        "#  PropertyType_Single Family: 22854.87954019308"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}