AI4H_practical2b_solutions.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "AI4H_practical2b_solutions.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "hide_input": false,
    "kernelspec": {
      "display_name": "jads-nhs-proms",
      "language": "python",
      "name": "jads-nhs-proms"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.6"
    },
    "toc": {
      "base_numbering": 1,
      "nav_menu": {},
      "number_sections": true,
      "sideBar": true,
      "skip_h1_title": true,
      "title_cell": "Table of Contents",
      "title_sidebar": "Contents",
      "toc_cell": false,
      "toc_position": {},
      "toc_section_display": true,
      "toc_window_display": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Jlokkerbol/masterclass/blob/main/2_0_data_preparation_DSBM.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ps-NB5SwlrNd"
      },
      "source": [
        "# Data Preparation\n",
        "In this notebook we will go through the data preparation phase. Steps such as cleaning up data, filling in missing values, selecting relevant features and defining outcomes are part of this phase. We will again follow the osteoarthritis case study that we also used last week. We will make use of the functionalities that Pandas offers for working with data. The data set we will work with today, is an extended version of the data set you saw last week. There are some more data features and outcomes."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C1bEDtlnF4a9"
      },
      "source": [
        "\n",
        "## Objectives Data Preparation\n",
        "\n",
        "### Making the data 'tidy'\n",
        "\n",
        "- Each variable must have its own column\n",
        "- Each observation must have its own row\n",
        "- Each value must have its own cell\n",
        "\n",
        "### Selecting relevant information \n",
        "\n",
        "- Select appropriate observations \n",
        "- Select available features\n",
        "- Select useful features\n",
        "\n",
        "### Cleaning data\n",
        "\n",
        "- Clean inconsistencies\n",
        "- Handle missing data\n",
        "- Setting appropriate data types\n",
        "\n",
        "### Constructing data\n",
        "\n",
        "- Define outcome Y (clustered classes, categories)\n",
        "- Redefine and/or construct features (feature engineering)\n",
        "- Decide if and how to handle correlated input features"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Background to osteoarthritis case study\n",
        "\n",
        "_taken from [narrative seminar Osteoarthritis by Hunter & Bierma-Zeinstra (2019) in the Lancet](https://github.com/dkapitan/jads-nhs-proms/blob/master/references/hunter2019osteaoarthritis.pdf)._\n",
        "\n",
        "Outcomes from total joint replacement can be optimised if patient selection identifies marked joint space narrowing. Most improvement will be made in patients with complete joint space loss and evident bone attrition. Up to 25% of patients presenting for total joint replacement continue to complain of pain and disability one year after (well-performed) surgery. Careful preoperative patient selection (including consideration of the poor outcomes that are more common in people who are depressed, have minimal radiographic disease, have minimal pain, and who are morbidly obese), shared decision making about surgery, and informing patients about realistic outcomes of surgery are needed to minimise the likelihood of dissatisfaction.\n",
        "\n",
        "__Recap from previous lecture__\n",
        "- Good outcome for knee replacement Y is measured using difference in Oxford Knee Score (OKS)\n",
        "- Research has shown that an improvement in OKS score of approx. 30% is relevant ([van der Wees 2017](https://github.com/dkapitan/jads-nhs-proms/blob/master/references/vanderwees2017patient-reported.pdf)). Hence an increase of +14 points is considered a 'good' outcome.\n",
        "- to account for the ceiling effect, a high final `t1_oks_score` is also considered as a good outcome (even if `delta_oks_score` is smaller than 14)."
      ],
      "metadata": {
        "id": "MpJVT6DiVRLI"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tGkaSFtoDRsm"
      },
      "source": [
        "# Importing data, libraries and style settings\n",
        "We will first import some relevant packages, the data, and set some style options to generate nice plots. Besides some of the packages you already know, we also import seaborn, a library that can create some special plots, and Scikit-learn (sklearn), which is the most important machine learning library for Python. We will see more of the Scikit-learn functionality in the next practical sessions, when we will start training machine learning models."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_95_D3ywlrNh"
      },
      "source": [
        "import warnings\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib as mpl\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "from sklearn.feature_selection import chi2, VarianceThreshold\n",
        "from sklearn.model_selection import train_test_split\n",
        "import sklearn.linear_model\n",
        "\n",
        "\n",
        "#supressing warnings for readability\n",
        "warnings.filterwarnings(\"ignore\")\n",
        "\n",
        "# To plot pretty figures directly within Jupyter\n",
        "%matplotlib inline\n",
        "\n",
        "# choose your own style: https://matplotlib.org/3.1.0/gallery/style_sheets/style_sheets_reference.html\n",
        "plt.style.use('seaborn-whitegrid')\n",
        "\n",
        "# Go to town with https://matplotlib.org/tutorials/introductory/customizing.html\n",
        "# plt.rcParams.keys()\n",
        "mpl.rc('axes', labelsize=14, titlesize=14)\n",
        "mpl.rc('figure', titlesize=20)\n",
        "mpl.rc('xtick', labelsize=12)\n",
        "mpl.rc('ytick', labelsize=12)\n",
        "\n",
        "# contants for figsize\n",
        "S = (8,8)\n",
        "M = (12,12)\n",
        "L = (14,14)\n",
        "\n",
        "# pandas options\n",
        "pd.set_option(\"display.max.columns\", None)\n",
        "pd.set_option(\"display.max.rows\", None)\n",
        "pd.set_option(\"display.precision\", 2)\n",
        "\n",
        "# import data\n",
        "df_raw = pd.read_parquet('https://github.com/jads-nl/execute-nhs-proms/blob/master/data/interim/knee-provider.parquet?raw=true')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Understanding data with Pandas\n",
        "In this step, we quickly recap what we learned about the data last week. As you have seen in the Pandas introduction, the library has some useful functions to understand your data at a glance. Below an overview of the data statistics for all features."
      ],
      "metadata": {
        "id": "_QAtFbrVY_P1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Display data information such as data types\n",
        "df_raw.info()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8OrNKZU0Zzej",
        "outputId": "d0ff86e9-abf4-4203-9d51-ea5d2e4e6190"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 139236 entries, 0 to 139235\n",
            "Data columns (total 81 columns):\n",
            " #   Column                        Non-Null Count   Dtype   \n",
            "---  ------                        --------------   -----   \n",
            " 0   provider_code                 139236 non-null  category\n",
            " 1   procedure                     139236 non-null  category\n",
            " 2   revision_flag                 139236 non-null  uint8   \n",
            " 3   year                          139236 non-null  category\n",
            " 4   age_band                      129834 non-null  category\n",
            " 5   gender                        129834 non-null  float32 \n",
            " 6   t0_assisted                   139236 non-null  uint8   \n",
            " 7   t0_assisted_by                139236 non-null  uint8   \n",
            " 8   t0_symptom_period             139236 non-null  uint8   \n",
            " 9   t0_previous_surgery           139236 non-null  uint8   \n",
            " 10  t0_living_arrangements        139236 non-null  uint8   \n",
            " 11  t0_disability                 139236 non-null  uint8   \n",
            " 12  heart_disease                 139236 non-null  uint8   \n",
            " 13  high_bp                       139236 non-null  uint8   \n",
            " 14  stroke                        139236 non-null  uint8   \n",
            " 15  circulation                   139236 non-null  uint8   \n",
            " 16  lung_disease                  139236 non-null  uint8   \n",
            " 17  diabetes                      139236 non-null  uint8   \n",
            " 18  kidney_disease                139236 non-null  uint8   \n",
            " 19  nervous_system                139236 non-null  uint8   \n",
            " 20  liver_disease                 139236 non-null  uint8   \n",
            " 21  cancer                        139236 non-null  uint8   \n",
            " 22  depression                    139236 non-null  uint8   \n",
            " 23  arthritis                     139236 non-null  uint8   \n",
            " 24  t0_mobility                   139236 non-null  uint8   \n",
            " 25  t0_self_care                  139236 non-null  uint8   \n",
            " 26  t0_activity                   139236 non-null  uint8   \n",
            " 27  t0_discomfort                 139236 non-null  uint8   \n",
            " 28  t0_anxiety                    139236 non-null  uint8   \n",
            " 29  t0_eq5d_index_profile         139236 non-null  int64   \n",
            " 30  t0_eq5d_index                 131736 non-null  float32 \n",
            " 31  t1_assisted                   139236 non-null  uint8   \n",
            " 32  t1_assisted_by                139236 non-null  uint8   \n",
            " 33  t1_living_arrangements        139236 non-null  uint8   \n",
            " 34  t1_disability                 139236 non-null  uint8   \n",
            " 35  t1_mobility                   139236 non-null  uint8   \n",
            " 36  t1_self_care                  139236 non-null  uint8   \n",
            " 37  t1_activity                   139236 non-null  uint8   \n",
            " 38  t1_discomfort                 139236 non-null  uint8   \n",
            " 39  t1_anxiety                    139236 non-null  uint8   \n",
            " 40  t1_satisfaction               139236 non-null  uint8   \n",
            " 41  t1_sucess                     139236 non-null  uint8   \n",
            " 42  t1_allergy                    139236 non-null  uint8   \n",
            " 43  t1_bleeding                   139236 non-null  uint8   \n",
            " 44  t1_wound                      139236 non-null  uint8   \n",
            " 45  t1_urine                      139236 non-null  uint8   \n",
            " 46  t1_further_surgery            139236 non-null  uint8   \n",
            " 47  t1_readmitted                 139236 non-null  uint8   \n",
            " 48  t1_eq5d_index_profile         139236 non-null  int64   \n",
            " 49  t1_eq5d_index                 132962 non-null  float32 \n",
            " 50  oks_eq_5d_index_t1_predicted  124914 non-null  float32 \n",
            " 51  t0_eq_vas                     139236 non-null  uint16  \n",
            " 52  t1_eq_vas                     139236 non-null  uint16  \n",
            " 53  oks_eq_vas_t1_predicted       119813 non-null  float32 \n",
            " 54  oks_t0_pain                   139236 non-null  uint8   \n",
            " 55  oks_t0_night_pain             139236 non-null  uint8   \n",
            " 56  oks_t0_washing                139236 non-null  uint8   \n",
            " 57  oks_t0_transport              139236 non-null  uint8   \n",
            " 58  oks_t0_walking                139236 non-null  uint8   \n",
            " 59  oks_t0_standing               139236 non-null  uint8   \n",
            " 60  oks_t0_limping                139236 non-null  uint8   \n",
            " 61  oks_t0_kneeling               139236 non-null  uint8   \n",
            " 62  oks_t0_work                   139236 non-null  uint8   \n",
            " 63  oks_t0_confidence             139236 non-null  uint8   \n",
            " 64  oks_t0_shopping               139236 non-null  uint8   \n",
            " 65  oks_t0_stairs                 139236 non-null  uint8   \n",
            " 66  oks_t0_score                  137567 non-null  float32 \n",
            " 67  oks_t1_pain                   139236 non-null  uint8   \n",
            " 68  oks_t1_night_pain             139236 non-null  uint8   \n",
            " 69  oks_t1_washing                139236 non-null  uint8   \n",
            " 70  oks_t1_transport              139236 non-null  uint8   \n",
            " 71  oks_t1_walking                139236 non-null  uint8   \n",
            " 72  oks_t1_standing               139236 non-null  uint8   \n",
            " 73  oks_t1_limping                139236 non-null  uint8   \n",
            " 74  oks_t1_kneeling               139236 non-null  uint8   \n",
            " 75  oks_t1_work                   139236 non-null  uint8   \n",
            " 76  oks_t1_confidence             139236 non-null  uint8   \n",
            " 77  oks_t1_shopping               139236 non-null  uint8   \n",
            " 78  oks_t1_stairs                 139236 non-null  uint8   \n",
            " 79  oks_t1_score                  136657 non-null  float32 \n",
            " 80  oks_oks_t1_predicted          133855 non-null  float32 \n",
            "dtypes: category(4), float32(8), int64(2), uint16(2), uint8(65)\n",
            "memory usage: 16.2 MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "As a quick reminder of last week, below an overview of the data statistics for all features."
      ],
      "metadata": {
        "id": "N7TIYrgZZ5xF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Display data statistics \n",
        "pd.set_option(\"display.precision\", 2)\n",
        "df_raw.describe().transpose()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "JF3hWC7LZWkl",
        "outputId": "ac0f1603-701f-46a1-e8c8-36783c5e98a2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-2f1ef39b-ff6e-4677-80f3-7770d17f883e\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>revision_flag</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>0.04</td>\n",
              "      <td>0.19</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>gender</th>\n",
              "      <td>129834.0</td>\n",
              "      <td>1.57</td>\n",
              "      <td>0.49</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_assisted</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.92</td>\n",
              "      <td>0.82</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_assisted_by</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_symptom_period</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.65</td>\n",
              "      <td>1.05</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_previous_surgery</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.97</td>\n",
              "      <td>0.67</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_living_arrangements</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.35</td>\n",
              "      <td>1.04</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_disability</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.81</td>\n",
              "      <td>1.59</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>heart_disease</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.25</td>\n",
              "      <td>2.33</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>high_bp</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>5.46</td>\n",
              "      <td>3.97</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>stroke</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.87</td>\n",
              "      <td>1.02</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>circulation</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.55</td>\n",
              "      <td>1.85</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>lung_disease</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.27</td>\n",
              "      <td>2.31</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>diabetes</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.00</td>\n",
              "      <td>2.64</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>kidney_disease</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.84</td>\n",
              "      <td>1.13</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>nervous_system</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.92</td>\n",
              "      <td>0.80</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>liver_disease</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.95</td>\n",
              "      <td>0.61</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cancer</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.57</td>\n",
              "      <td>1.81</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>depression</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.25</td>\n",
              "      <td>2.33</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>arthritis</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.84</td>\n",
              "      <td>3.36</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_mobility</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.14</td>\n",
              "      <td>1.25</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_self_care</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.55</td>\n",
              "      <td>1.43</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_activity</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.26</td>\n",
              "      <td>1.31</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_discomfort</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.64</td>\n",
              "      <td>1.39</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_anxiety</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.68</td>\n",
              "      <td>1.54</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_eq5d_index_profile</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>23214.76</td>\n",
              "      <td>13893.51</td>\n",
              "      <td>11111.00</td>\n",
              "      <td>21221.00</td>\n",
              "      <td>21231.00</td>\n",
              "      <td>22222.00</td>\n",
              "      <td>99999.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_eq5d_index</th>\n",
              "      <td>131736.0</td>\n",
              "      <td>0.41</td>\n",
              "      <td>0.31</td>\n",
              "      <td>-0.59</td>\n",
              "      <td>0.09</td>\n",
              "      <td>0.59</td>\n",
              "      <td>0.69</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_assisted</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.01</td>\n",
              "      <td>0.79</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_assisted_by</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>8.41</td>\n",
              "      <td>2.05</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_living_arrangements</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.41</td>\n",
              "      <td>1.20</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_disability</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.93</td>\n",
              "      <td>1.47</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_mobility</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.61</td>\n",
              "      <td>1.20</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_self_care</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.32</td>\n",
              "      <td>1.15</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_activity</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.70</td>\n",
              "      <td>1.23</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_discomfort</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.87</td>\n",
              "      <td>1.34</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_anxiety</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.42</td>\n",
              "      <td>1.25</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_satisfaction</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.36</td>\n",
              "      <td>1.33</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_sucess</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.53</td>\n",
              "      <td>1.23</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_allergy</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.53</td>\n",
              "      <td>2.08</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_bleeding</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.80</td>\n",
              "      <td>2.34</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_wound</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.64</td>\n",
              "      <td>2.21</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_urine</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.68</td>\n",
              "      <td>2.26</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_further_surgery</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.06</td>\n",
              "      <td>0.81</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_readmitted</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.02</td>\n",
              "      <td>0.85</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_eq5d_index_profile</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>17587.27</td>\n",
              "      <td>12939.07</td>\n",
              "      <td>11111.00</td>\n",
              "      <td>11111.00</td>\n",
              "      <td>11221.00</td>\n",
              "      <td>21221.00</td>\n",
              "      <td>99999.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_eq5d_index</th>\n",
              "      <td>132962.0</td>\n",
              "      <td>0.75</td>\n",
              "      <td>0.25</td>\n",
              "      <td>-0.59</td>\n",
              "      <td>0.69</td>\n",
              "      <td>0.76</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_eq_5d_index_t1_predicted</th>\n",
              "      <td>124914.0</td>\n",
              "      <td>0.72</td>\n",
              "      <td>0.12</td>\n",
              "      <td>0.03</td>\n",
              "      <td>0.65</td>\n",
              "      <td>0.74</td>\n",
              "      <td>0.82</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_eq_vas</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>154.01</td>\n",
              "      <td>271.43</td>\n",
              "      <td>0.00</td>\n",
              "      <td>55.00</td>\n",
              "      <td>75.00</td>\n",
              "      <td>89.00</td>\n",
              "      <td>999.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t1_eq_vas</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>118.23</td>\n",
              "      <td>196.45</td>\n",
              "      <td>0.00</td>\n",
              "      <td>69.00</td>\n",
              "      <td>80.00</td>\n",
              "      <td>90.00</td>\n",
              "      <td>999.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_eq_vas_t1_predicted</th>\n",
              "      <td>119813.0</td>\n",
              "      <td>72.58</td>\n",
              "      <td>9.46</td>\n",
              "      <td>21.76</td>\n",
              "      <td>65.96</td>\n",
              "      <td>73.17</td>\n",
              "      <td>79.81</td>\n",
              "      <td>93.99</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_pain</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.72</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_night_pain</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.30</td>\n",
              "      <td>1.37</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_washing</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.82</td>\n",
              "      <td>1.02</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_transport</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.13</td>\n",
              "      <td>1.07</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_walking</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.12</td>\n",
              "      <td>1.30</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_standing</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.75</td>\n",
              "      <td>1.11</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_limping</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>0.97</td>\n",
              "      <td>1.29</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_kneeling</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>0.90</td>\n",
              "      <td>1.18</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_work</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.49</td>\n",
              "      <td>1.14</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_confidence</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.91</td>\n",
              "      <td>1.36</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_shopping</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.91</td>\n",
              "      <td>1.39</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_stairs</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.85</td>\n",
              "      <td>1.14</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t0_score</th>\n",
              "      <td>137567.0</td>\n",
              "      <td>18.98</td>\n",
              "      <td>7.75</td>\n",
              "      <td>0.00</td>\n",
              "      <td>13.00</td>\n",
              "      <td>19.00</td>\n",
              "      <td>24.00</td>\n",
              "      <td>48.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_pain</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.67</td>\n",
              "      <td>1.29</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_night_pain</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>2.85</td>\n",
              "      <td>1.34</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_washing</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>3.62</td>\n",
              "      <td>0.90</td>\n",
              "      <td>0.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_transport</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>3.11</td>\n",
              "      <td>1.06</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_walking</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>3.39</td>\n",
              "      <td>1.13</td>\n",
              "      <td>0.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_standing</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>3.12</td>\n",
              "      <td>1.03</td>\n",
              "      <td>0.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_limping</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>3.12</td>\n",
              "      <td>1.21</td>\n",
              "      <td>0.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_kneeling</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>1.65</td>\n",
              "      <td>1.48</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_work</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>3.09</td>\n",
              "      <td>1.11</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_confidence</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>3.53</td>\n",
              "      <td>0.95</td>\n",
              "      <td>0.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_shopping</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>3.21</td>\n",
              "      <td>1.29</td>\n",
              "      <td>0.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_stairs</th>\n",
              "      <td>139236.0</td>\n",
              "      <td>3.06</td>\n",
              "      <td>1.14</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_t1_score</th>\n",
              "      <td>136657.0</td>\n",
              "      <td>35.89</td>\n",
              "      <td>9.47</td>\n",
              "      <td>0.00</td>\n",
              "      <td>30.00</td>\n",
              "      <td>38.00</td>\n",
              "      <td>43.00</td>\n",
              "      <td>48.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>oks_oks_t1_predicted</th>\n",
              "      <td>133855.0</td>\n",
              "      <td>34.43</td>\n",
              "      <td>4.74</td>\n",
              "      <td>7.06</td>\n",
              "      <td>31.55</td>\n",
              "      <td>35.07</td>\n",
              "      <td>38.00</td>\n",
              "      <td>48.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-2f1ef39b-ff6e-4677-80f3-7770d17f883e')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-2f1ef39b-ff6e-4677-80f3-7770d17f883e button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-2f1ef39b-ff6e-4677-80f3-7770d17f883e');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "                                 count      mean       std       min  \\\n",
              "revision_flag                 139236.0      0.04      0.19      0.00   \n",
              "gender                        129834.0      1.57      0.49      1.00   \n",
              "t0_assisted                   139236.0      1.92      0.82      1.00   \n",
              "t0_assisted_by                139236.0      0.00      0.00      0.00   \n",
              "t0_symptom_period             139236.0      2.65      1.05      1.00   \n",
              "t0_previous_surgery           139236.0      1.97      0.67      1.00   \n",
              "t0_living_arrangements        139236.0      1.35      1.04      1.00   \n",
              "t0_disability                 139236.0      1.81      1.59      1.00   \n",
              "heart_disease                 139236.0      8.25      2.33      1.00   \n",
              "high_bp                       139236.0      5.46      3.97      1.00   \n",
              "stroke                        139236.0      8.87      1.02      1.00   \n",
              "circulation                   139236.0      8.55      1.85      1.00   \n",
              "lung_disease                  139236.0      8.27      2.31      1.00   \n",
              "diabetes                      139236.0      8.00      2.64      1.00   \n",
              "kidney_disease                139236.0      8.84      1.13      1.00   \n",
              "nervous_system                139236.0      8.92      0.80      1.00   \n",
              "liver_disease                 139236.0      8.95      0.61      1.00   \n",
              "cancer                        139236.0      8.57      1.81      1.00   \n",
              "depression                    139236.0      8.25      2.33      1.00   \n",
              "arthritis                     139236.0      2.84      3.36      1.00   \n",
              "t0_mobility                   139236.0      2.14      1.25      1.00   \n",
              "t0_self_care                  139236.0      1.55      1.43      1.00   \n",
              "t0_activity                   139236.0      2.26      1.31      1.00   \n",
              "t0_discomfort                 139236.0      2.64      1.39      1.00   \n",
              "t0_anxiety                    139236.0      1.68      1.54      1.00   \n",
              "t0_eq5d_index_profile         139236.0  23214.76  13893.51  11111.00   \n",
              "t0_eq5d_index                 131736.0      0.41      0.31     -0.59   \n",
              "t1_assisted                   139236.0      2.01      0.79      1.00   \n",
              "t1_assisted_by                139236.0      8.41      2.05      1.00   \n",
              "t1_living_arrangements        139236.0      1.41      1.20      1.00   \n",
              "t1_disability                 139236.0      1.93      1.47      1.00   \n",
              "t1_mobility                   139236.0      1.61      1.20      1.00   \n",
              "t1_self_care                  139236.0      1.32      1.15      1.00   \n",
              "t1_activity                   139236.0      1.70      1.23      1.00   \n",
              "t1_discomfort                 139236.0      1.87      1.34      1.00   \n",
              "t1_anxiety                    139236.0      1.42      1.25      1.00   \n",
              "t1_satisfaction               139236.0      2.36      1.33      1.00   \n",
              "t1_sucess                     139236.0      1.53      1.23      1.00   \n",
              "t1_allergy                    139236.0      2.53      2.08      1.00   \n",
              "t1_bleeding                   139236.0      2.80      2.34      1.00   \n",
              "t1_wound                      139236.0      2.64      2.21      1.00   \n",
              "t1_urine                      139236.0      2.68      2.26      1.00   \n",
              "t1_further_surgery            139236.0      2.06      0.81      1.00   \n",
              "t1_readmitted                 139236.0      2.02      0.85      1.00   \n",
              "t1_eq5d_index_profile         139236.0  17587.27  12939.07  11111.00   \n",
              "t1_eq5d_index                 132962.0      0.75      0.25     -0.59   \n",
              "oks_eq_5d_index_t1_predicted  124914.0      0.72      0.12      0.03   \n",
              "t0_eq_vas                     139236.0    154.01    271.43      0.00   \n",
              "t1_eq_vas                     139236.0    118.23    196.45      0.00   \n",
              "oks_eq_vas_t1_predicted       119813.0     72.58      9.46     21.76   \n",
              "oks_t0_pain                   139236.0      0.58      0.72      0.00   \n",
              "oks_t0_night_pain             139236.0      1.30      1.37      0.00   \n",
              "oks_t0_washing                139236.0      2.82      1.02      0.00   \n",
              "oks_t0_transport              139236.0      2.13      1.07      0.00   \n",
              "oks_t0_walking                139236.0      2.12      1.30      0.00   \n",
              "oks_t0_standing               139236.0      1.75      1.11      0.00   \n",
              "oks_t0_limping                139236.0      0.97      1.29      0.00   \n",
              "oks_t0_kneeling               139236.0      0.90      1.18      0.00   \n",
              "oks_t0_work                   139236.0      1.49      1.14      0.00   \n",
              "oks_t0_confidence             139236.0      1.91      1.36      0.00   \n",
              "oks_t0_shopping               139236.0      1.91      1.39      0.00   \n",
              "oks_t0_stairs                 139236.0      1.85      1.14      0.00   \n",
              "oks_t0_score                  137567.0     18.98      7.75      0.00   \n",
              "oks_t1_pain                   139236.0      2.67      1.29      0.00   \n",
              "oks_t1_night_pain             139236.0      2.85      1.34      0.00   \n",
              "oks_t1_washing                139236.0      3.62      0.90      0.00   \n",
              "oks_t1_transport              139236.0      3.11      1.06      0.00   \n",
              "oks_t1_walking                139236.0      3.39      1.13      0.00   \n",
              "oks_t1_standing               139236.0      3.12      1.03      0.00   \n",
              "oks_t1_limping                139236.0      3.12      1.21      0.00   \n",
              "oks_t1_kneeling               139236.0      1.65      1.48      0.00   \n",
              "oks_t1_work                   139236.0      3.09      1.11      0.00   \n",
              "oks_t1_confidence             139236.0      3.53      0.95      0.00   \n",
              "oks_t1_shopping               139236.0      3.21      1.29      0.00   \n",
              "oks_t1_stairs                 139236.0      3.06      1.14      0.00   \n",
              "oks_t1_score                  136657.0     35.89      9.47      0.00   \n",
              "oks_oks_t1_predicted          133855.0     34.43      4.74      7.06   \n",
              "\n",
              "                                   25%       50%       75%       max  \n",
              "revision_flag                     0.00      0.00      0.00      1.00  \n",
              "gender                            1.00      2.00      2.00      2.00  \n",
              "t0_assisted                       2.00      2.00      2.00      9.00  \n",
              "t0_assisted_by                    0.00      0.00      0.00      0.00  \n",
              "t0_symptom_period                 2.00      2.00      3.00      9.00  \n",
              "t0_previous_surgery               2.00      2.00      2.00      9.00  \n",
              "t0_living_arrangements            1.00      1.00      1.00      9.00  \n",
              "t0_disability                     1.00      2.00      2.00      9.00  \n",
              "heart_disease                     9.00      9.00      9.00      9.00  \n",
              "high_bp                           1.00      9.00      9.00      9.00  \n",
              "stroke                            9.00      9.00      9.00      9.00  \n",
              "circulation                       9.00      9.00      9.00      9.00  \n",
              "lung_disease                      9.00      9.00      9.00      9.00  \n",
              "diabetes                          9.00      9.00      9.00      9.00  \n",
              "kidney_disease                    9.00      9.00      9.00      9.00  \n",
              "nervous_system                    9.00      9.00      9.00      9.00  \n",
              "liver_disease                     9.00      9.00      9.00      9.00  \n",
              "cancer                            9.00      9.00      9.00      9.00  \n",
              "depression                        9.00      9.00      9.00      9.00  \n",
              "arthritis                         1.00      1.00      1.00      9.00  \n",
              "t0_mobility                       2.00      2.00      2.00      9.00  \n",
              "t0_self_care                      1.00      1.00      2.00      9.00  \n",
              "t0_activity                       2.00      2.00      2.00      9.00  \n",
              "t0_discomfort                     2.00      2.00      3.00      9.00  \n",
              "t0_anxiety                        1.00      1.00      2.00      9.00  \n",
              "t0_eq5d_index_profile         21221.00  21231.00  22222.00  99999.00  \n",
              "t0_eq5d_index                     0.09      0.59      0.69      1.00  \n",
              "t1_assisted                       2.00      2.00      2.00      9.00  \n",
              "t1_assisted_by                    9.00      9.00      9.00      9.00  \n",
              "t1_living_arrangements            1.00      1.00      2.00      9.00  \n",
              "t1_disability                     1.00      2.00      2.00      9.00  \n",
              "t1_mobility                       1.00      1.00      2.00      9.00  \n",
              "t1_self_care                      1.00      1.00      1.00      9.00  \n",
              "t1_activity                       1.00      2.00      2.00      9.00  \n",
              "t1_discomfort                     1.00      2.00      2.00      9.00  \n",
              "t1_anxiety                        1.00      1.00      1.00      9.00  \n",
              "t1_satisfaction                   1.00      2.00      3.00      9.00  \n",
              "t1_sucess                         1.00      1.00      2.00      9.00  \n",
              "t1_allergy                        2.00      2.00      2.00      9.00  \n",
              "t1_bleeding                       2.00      2.00      2.00      9.00  \n",
              "t1_wound                          2.00      2.00      2.00      9.00  \n",
              "t1_urine                          2.00      2.00      2.00      9.00  \n",
              "t1_further_surgery                2.00      2.00      2.00      9.00  \n",
              "t1_readmitted                     2.00      2.00      2.00      9.00  \n",
              "t1_eq5d_index_profile         11111.00  11221.00  21221.00  99999.00  \n",
              "t1_eq5d_index                     0.69      0.76      1.00      1.00  \n",
              "oks_eq_5d_index_t1_predicted      0.65      0.74      0.82      1.00  \n",
              "t0_eq_vas                        55.00     75.00     89.00    999.00  \n",
              "t1_eq_vas                        69.00     80.00     90.00    999.00  \n",
              "oks_eq_vas_t1_predicted          65.96     73.17     79.81     93.99  \n",
              "oks_t0_pain                       0.00      0.00      1.00      9.00  \n",
              "oks_t0_night_pain                 0.00      1.00      2.00      9.00  \n",
              "oks_t0_washing                    2.00      3.00      4.00      9.00  \n",
              "oks_t0_transport                  2.00      2.00      2.00      9.00  \n",
              "oks_t0_walking                    2.00      2.00      3.00      9.00  \n",
              "oks_t0_standing                   1.00      2.00      2.00      9.00  \n",
              "oks_t0_limping                    0.00      1.00      1.00      9.00  \n",
              "oks_t0_kneeling                   0.00      1.00      1.00      9.00  \n",
              "oks_t0_work                       1.00      1.00      2.00      9.00  \n",
              "oks_t0_confidence                 1.00      2.00      3.00      9.00  \n",
              "oks_t0_shopping                   1.00      2.00      3.00      9.00  \n",
              "oks_t0_stairs                     1.00      2.00      2.00      9.00  \n",
              "oks_t0_score                     13.00     19.00     24.00     48.00  \n",
              "oks_t1_pain                       2.00      3.00      4.00      9.00  \n",
              "oks_t1_night_pain                 2.00      3.00      4.00      9.00  \n",
              "oks_t1_washing                    3.00      4.00      4.00      9.00  \n",
              "oks_t1_transport                  2.00      3.00      4.00      9.00  \n",
              "oks_t1_walking                    3.00      4.00      4.00      9.00  \n",
              "oks_t1_standing                   3.00      3.00      4.00      9.00  \n",
              "oks_t1_limping                    3.00      3.00      4.00      9.00  \n",
              "oks_t1_kneeling                   0.00      2.00      3.00      9.00  \n",
              "oks_t1_work                       2.00      3.00      4.00      9.00  \n",
              "oks_t1_confidence                 3.00      4.00      4.00      9.00  \n",
              "oks_t1_shopping                   3.00      4.00      4.00      9.00  \n",
              "oks_t1_stairs                     2.00      3.00      4.00      9.00  \n",
              "oks_t1_score                     30.00     38.00     43.00     48.00  \n",
              "oks_oks_t1_predicted             31.55     35.07     38.00     48.00  "
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xwtwom3PeOsN"
      },
      "source": [
        "# Making the data 'tidy'\n",
        "\n",
        "This step is something that could take a lot of effort, but in our case we are lucky to be working with 'tidy' data already. \n",
        "* There is only one table to work with, so no merging or joining is needed\n",
        "* Our rows correspond to the 'unit of analysis', just like we want them to\n",
        "* Every column neatly contains one piece of information about our 'unit of analyses', so for every available replacement"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Z_n6cIqefAIL"
      },
      "source": [
        "# Selecting relevant information\n",
        "\n",
        "The next step is selecting the available features (at the intended moment of modeling) and the relevant features (i.e. we exclude the obviously irrelevant features such as IDs, telephone numbers for example), and select a subset of the rows if necessary."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xDUZ03lChl_d",
        "outputId": "14d633b8-9d35-40bc-bb3a-0b13373c34ef"
      },
      "source": [
        "# List all the available features\n",
        "list(df_raw.columns)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['provider_code',\n",
              " 'procedure',\n",
              " 'revision_flag',\n",
              " 'year',\n",
              " 'age_band',\n",
              " 'gender',\n",
              " 't0_assisted',\n",
              " 't0_assisted_by',\n",
              " 't0_symptom_period',\n",
              " 't0_previous_surgery',\n",
              " 't0_living_arrangements',\n",
              " 't0_disability',\n",
              " 'heart_disease',\n",
              " 'high_bp',\n",
              " 'stroke',\n",
              " 'circulation',\n",
              " 'lung_disease',\n",
              " 'diabetes',\n",
              " 'kidney_disease',\n",
              " 'nervous_system',\n",
              " 'liver_disease',\n",
              " 'cancer',\n",
              " 'depression',\n",
              " 'arthritis',\n",
              " 't0_mobility',\n",
              " 't0_self_care',\n",
              " 't0_activity',\n",
              " 't0_discomfort',\n",
              " 't0_anxiety',\n",
              " 't0_eq5d_index_profile',\n",
              " 't0_eq5d_index',\n",
              " 't1_assisted',\n",
              " 't1_assisted_by',\n",
              " 't1_living_arrangements',\n",
              " 't1_disability',\n",
              " 't1_mobility',\n",
              " 't1_self_care',\n",
              " 't1_activity',\n",
              " 't1_discomfort',\n",
              " 't1_anxiety',\n",
              " 't1_satisfaction',\n",
              " 't1_sucess',\n",
              " 't1_allergy',\n",
              " 't1_bleeding',\n",
              " 't1_wound',\n",
              " 't1_urine',\n",
              " 't1_further_surgery',\n",
              " 't1_readmitted',\n",
              " 't1_eq5d_index_profile',\n",
              " 't1_eq5d_index',\n",
              " 'oks_eq_5d_index_t1_predicted',\n",
              " 't0_eq_vas',\n",
              " 't1_eq_vas',\n",
              " 'oks_eq_vas_t1_predicted',\n",
              " 'oks_t0_pain',\n",
              " 'oks_t0_night_pain',\n",
              " 'oks_t0_washing',\n",
              " 'oks_t0_transport',\n",
              " 'oks_t0_walking',\n",
              " 'oks_t0_standing',\n",
              " 'oks_t0_limping',\n",
              " 'oks_t0_kneeling',\n",
              " 'oks_t0_work',\n",
              " 'oks_t0_confidence',\n",
              " 'oks_t0_shopping',\n",
              " 'oks_t0_stairs',\n",
              " 'oks_t0_score',\n",
              " 'oks_t1_pain',\n",
              " 'oks_t1_night_pain',\n",
              " 'oks_t1_washing',\n",
              " 'oks_t1_transport',\n",
              " 'oks_t1_walking',\n",
              " 'oks_t1_standing',\n",
              " 'oks_t1_limping',\n",
              " 'oks_t1_kneeling',\n",
              " 'oks_t1_work',\n",
              " 'oks_t1_confidence',\n",
              " 'oks_t1_shopping',\n",
              " 'oks_t1_stairs',\n",
              " 'oks_t1_score',\n",
              " 'oks_oks_t1_predicted']"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5vPSnDGah0LR"
      },
      "source": [
        "You see that the list contains more features then last week. Features that are dependant on the moment of measurement are indicated with either `t0`, or `t1`.\n",
        "\n",
        "There are also some extra outcomes beside the OKS score:\n",
        "\n",
        "- Oxford Knee Score (OKS): a 12-item questionnaire that assess daily functioning c.q. disability due to knee osteoarthritis. Items are scored from [0,4], where higher is better (no disability).  \n",
        "  - OKS question on pain and night pain, both on scale from [0,4]\n",
        "  - OKS , higher is better. The OKS is a 12-item questionaire that assess daily functioning c.q. disability.\n",
        "- EQ5D: generic quality of life PROMs along 5 dimensions on a 3-point Likert scale [1,3], lower is better. Dimensions are problems activity, anxiety, discomfort, mobility and self-care.\n",
        "- EQ-VAS: general reported health on a scale from [0,100], higher is better.\n",
        "\n",
        "We will explore the PROMs data to see which target variable Y is meaningful to assess the outcome of knee replacement. Note PROMs are measured at T0 (prior to surgery) and T1 (six months after surgery).\n",
        "- [NHS PROMs methodology](https://github.com/dkapitan/jads-nhs-proms/blob/master/references/nhs/proms_guide_v12.pdf)\n",
        "- [NHS PROMs data dictionairy](https://github.com/dkapitan/jads-nhs-proms/blob/master/references/nhs/proms_data_dictionary.pdf)\n",
        "\n",
        "There are a number of `t1` variables that we cannot use at the time when we want the model to make a prediction, so we can exclude those variables from the analysis. For now we will leave in all the `t1` variables relating to the OKS that we will need when constructing our outcome Y. We will remove those after defining Y."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nuZoo-n0iUMD",
        "outputId": "c1c2c6ca-9e6d-41d5-d472-2926852963fb"
      },
      "source": [
        "# remove t1 features that are not available for prediction\n",
        "to_remove = ['t1_assisted', 't1_assisted_by', 't1_living_arrangements', 't1_disability', 't1_mobility', 't1_self_care',\n",
        "             't1_activity', 't1_discomfort', 't1_anxiety', 't1_satisfaction', 't1_sucess', 't1_allergy', 't1_bleeding',\n",
        "             't1_wound', 't1_urine', 't1_further_surgery', 't1_readmitted', 't1_eq5d_index_profile', 't1_eq5d_index']\n",
        "\n",
        "df_temp = df_raw.drop(to_remove, axis=1)\n",
        "list(df_temp.columns)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['provider_code',\n",
              " 'procedure',\n",
              " 'revision_flag',\n",
              " 'year',\n",
              " 'age_band',\n",
              " 'gender',\n",
              " 't0_assisted',\n",
              " 't0_assisted_by',\n",
              " 't0_symptom_period',\n",
              " 't0_previous_surgery',\n",
              " 't0_living_arrangements',\n",
              " 't0_disability',\n",
              " 'heart_disease',\n",
              " 'high_bp',\n",
              " 'stroke',\n",
              " 'circulation',\n",
              " 'lung_disease',\n",
              " 'diabetes',\n",
              " 'kidney_disease',\n",
              " 'nervous_system',\n",
              " 'liver_disease',\n",
              " 'cancer',\n",
              " 'depression',\n",
              " 'arthritis',\n",
              " 't0_mobility',\n",
              " 't0_self_care',\n",
              " 't0_activity',\n",
              " 't0_discomfort',\n",
              " 't0_anxiety',\n",
              " 't0_eq5d_index_profile',\n",
              " 't0_eq5d_index',\n",
              " 'oks_eq_5d_index_t1_predicted',\n",
              " 't0_eq_vas',\n",
              " 't1_eq_vas',\n",
              " 'oks_eq_vas_t1_predicted',\n",
              " 'oks_t0_pain',\n",
              " 'oks_t0_night_pain',\n",
              " 'oks_t0_washing',\n",
              " 'oks_t0_transport',\n",
              " 'oks_t0_walking',\n",
              " 'oks_t0_standing',\n",
              " 'oks_t0_limping',\n",
              " 'oks_t0_kneeling',\n",
              " 'oks_t0_work',\n",
              " 'oks_t0_confidence',\n",
              " 'oks_t0_shopping',\n",
              " 'oks_t0_stairs',\n",
              " 'oks_t0_score',\n",
              " 'oks_t1_pain',\n",
              " 'oks_t1_night_pain',\n",
              " 'oks_t1_washing',\n",
              " 'oks_t1_transport',\n",
              " 'oks_t1_walking',\n",
              " 'oks_t1_standing',\n",
              " 'oks_t1_limping',\n",
              " 'oks_t1_kneeling',\n",
              " 'oks_t1_work',\n",
              " 'oks_t1_confidence',\n",
              " 'oks_t1_shopping',\n",
              " 'oks_t1_stairs',\n",
              " 'oks_t1_score',\n",
              " 'oks_oks_t1_predicted']"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Bt1GQyai888m"
      },
      "source": [
        "In addition, domain experts deem a number of `t0` variables to be irrelevant: 't0_assisted_by' and 't0_eq5d_index_profile'. Can you remove these features from the data?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EIEOBWbJ9ESX",
        "outputId": "118f94d9-77f7-4909-b804-1cda83e65237"
      },
      "source": [
        "# remove irrelevant features\n",
        "to_remove_t0 = [ 't0_assisted_by','t0_eq5d_index_profile']\n",
        "\n",
        "df_temp = df_temp.drop(to_remove_t0, axis=1)\n",
        "list(df_temp.columns)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['provider_code',\n",
              " 'procedure',\n",
              " 'revision_flag',\n",
              " 'year',\n",
              " 'age_band',\n",
              " 'gender',\n",
              " 't0_assisted',\n",
              " 't0_symptom_period',\n",
              " 't0_previous_surgery',\n",
              " 't0_living_arrangements',\n",
              " 't0_disability',\n",
              " 'heart_disease',\n",
              " 'high_bp',\n",
              " 'stroke',\n",
              " 'circulation',\n",
              " 'lung_disease',\n",
              " 'diabetes',\n",
              " 'kidney_disease',\n",
              " 'nervous_system',\n",
              " 'liver_disease',\n",
              " 'cancer',\n",
              " 'depression',\n",
              " 'arthritis',\n",
              " 't0_mobility',\n",
              " 't0_self_care',\n",
              " 't0_activity',\n",
              " 't0_discomfort',\n",
              " 't0_anxiety',\n",
              " 't0_eq5d_index',\n",
              " 'oks_eq_5d_index_t1_predicted',\n",
              " 't0_eq_vas',\n",
              " 't1_eq_vas',\n",
              " 'oks_eq_vas_t1_predicted',\n",
              " 'oks_t0_pain',\n",
              " 'oks_t0_night_pain',\n",
              " 'oks_t0_washing',\n",
              " 'oks_t0_transport',\n",
              " 'oks_t0_walking',\n",
              " 'oks_t0_standing',\n",
              " 'oks_t0_limping',\n",
              " 'oks_t0_kneeling',\n",
              " 'oks_t0_work',\n",
              " 'oks_t0_confidence',\n",
              " 'oks_t0_shopping',\n",
              " 'oks_t0_stairs',\n",
              " 'oks_t0_score',\n",
              " 'oks_t1_pain',\n",
              " 'oks_t1_night_pain',\n",
              " 'oks_t1_washing',\n",
              " 'oks_t1_transport',\n",
              " 'oks_t1_walking',\n",
              " 'oks_t1_standing',\n",
              " 'oks_t1_limping',\n",
              " 'oks_t1_kneeling',\n",
              " 'oks_t1_work',\n",
              " 'oks_t1_confidence',\n",
              " 'oks_t1_shopping',\n",
              " 'oks_t1_stairs',\n",
              " 'oks_t1_score',\n",
              " 'oks_oks_t1_predicted']"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UwkQXtX3jflV"
      },
      "source": [
        "During data understanding, we saw that the data was scarce on patients in the oldest and youngest age-brackets. For the sake of the example, let's suppose that we decide to take these age-brackets out of scope, meaning we exclude these records from our data."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p8DnlSLcjfA2",
        "outputId": "cb68221f-7102-4393-bb82-7e6e4947432c"
      },
      "source": [
        "# describe age_band information\n",
        "df_temp['age_band'].value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "70 to 79     55056\n",
              "60 to 69     45361\n",
              "80 to 89     15883\n",
              "50 to 59     13260\n",
              "40 to 49       250\n",
              "90 to 120       24\n",
              "Name: age_band, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Can you exclude rows corresponding to the highest and lowest age-bracket?"
      ],
      "metadata": {
        "id": "hjNNHd52a2TL"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "39FuO5_FkGOJ",
        "outputId": "ef4763de-47f2-4667-96ad-bb7011bea18b"
      },
      "source": [
        "# Exclude rows corresponding to the highest and lowest age-bracket\n",
        "df_temp = df_temp.drop(df_temp[df_temp.age_band == \"40 to 49\"].index)\n",
        "df_temp = df_temp.drop(df_temp[df_temp.age_band == \"90 to 120\"].index)\n",
        "df_temp['age_band'].value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "70 to 79     55056\n",
              "60 to 69     45361\n",
              "80 to 89     15883\n",
              "50 to 59     13260\n",
              "40 to 49         0\n",
              "90 to 120        0\n",
              "Name: age_band, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QZqJb_Zipjsb",
        "outputId": "7cb14bbf-ff86-4620-86c5-5cbd57d0eb2c"
      },
      "source": [
        "# unused levels should be removed to further clean / prepare the data\n",
        "df_temp.age_band = df_temp.age_band.cat.remove_unused_categories()\n",
        "df_temp['age_band'].value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "70 to 79    55056\n",
              "60 to 69    45361\n",
              "80 to 89    15883\n",
              "50 to 59    13260\n",
              "Name: age_band, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ueq8uUDQqRw5"
      },
      "source": [
        "# Cleaning Data\n",
        "\n",
        "In this step we handle missing values, correct any inconsistencies found in the data and set the appropriate data types. These data types are important for many of the models we will be training, as they make certain assumptions when a feature is for example of a 'category' type. So it's always good to use data types that match the actual nature of the data.\n",
        "\n",
        "We start off with a straightforward imputation strategy, where we impute \n",
        "* numerical variables with the median value, and\n",
        "* categorical variables with the most occurring level"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_gOGhai7qf0M",
        "outputId": "ab9ccaaa-cc5d-486f-e21f-1044193d179d"
      },
      "source": [
        "# First we map all the missings in the data\n",
        "df_temp.isnull().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "provider_code                       0\n",
              "procedure                           0\n",
              "revision_flag                       0\n",
              "year                                0\n",
              "age_band                         9402\n",
              "gender                           9402\n",
              "t0_assisted                         0\n",
              "t0_symptom_period                   0\n",
              "t0_previous_surgery                 0\n",
              "t0_living_arrangements              0\n",
              "t0_disability                       0\n",
              "heart_disease                       0\n",
              "high_bp                             0\n",
              "stroke                              0\n",
              "circulation                         0\n",
              "lung_disease                        0\n",
              "diabetes                            0\n",
              "kidney_disease                      0\n",
              "nervous_system                      0\n",
              "liver_disease                       0\n",
              "cancer                              0\n",
              "depression                          0\n",
              "arthritis                           0\n",
              "t0_mobility                         0\n",
              "t0_self_care                        0\n",
              "t0_activity                         0\n",
              "t0_discomfort                       0\n",
              "t0_anxiety                          0\n",
              "t0_eq5d_index                    7485\n",
              "oks_eq_5d_index_t1_predicted    14279\n",
              "t0_eq_vas                           0\n",
              "t1_eq_vas                           0\n",
              "oks_eq_vas_t1_predicted         19379\n",
              "oks_t0_pain                         0\n",
              "oks_t0_night_pain                   0\n",
              "oks_t0_washing                      0\n",
              "oks_t0_transport                    0\n",
              "oks_t0_walking                      0\n",
              "oks_t0_standing                     0\n",
              "oks_t0_limping                      0\n",
              "oks_t0_kneeling                     0\n",
              "oks_t0_work                         0\n",
              "oks_t0_confidence                   0\n",
              "oks_t0_shopping                     0\n",
              "oks_t0_stairs                       0\n",
              "oks_t0_score                     1669\n",
              "oks_t1_pain                         0\n",
              "oks_t1_night_pain                   0\n",
              "oks_t1_washing                      0\n",
              "oks_t1_transport                    0\n",
              "oks_t1_walking                      0\n",
              "oks_t1_standing                     0\n",
              "oks_t1_limping                      0\n",
              "oks_t1_kneeling                     0\n",
              "oks_t1_work                         0\n",
              "oks_t1_confidence                   0\n",
              "oks_t1_shopping                     0\n",
              "oks_t1_stairs                       0\n",
              "oks_t1_score                     2575\n",
              "oks_oks_t1_predicted             5362\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nfwQn62bFc8O"
      },
      "source": [
        "These are actually not all the missing. For some features a 9 is used to indicate a missing value. We need to identify those variables to be able to impute those 9s that actually refer to a missing value "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 741
        },
        "id": "7NqUIzwnJft_",
        "outputId": "73353c83-f66c-48a3-aa49-6362464f5c82"
      },
      "source": [
        "# list of all columns which contain 9s\n",
        "cols_with_9 = [col for col in df_temp.columns if (df_temp[col]==9).any()]\n",
        "\n",
        "# columns where 9 is *not* a sentinel value\n",
        "comorb = ['heart_disease', 'high_bp', 'stroke', 'circulation', 'lung_disease', 'diabetes',\n",
        "           'kidney_disease', 'nervous_system', 'liver_disease', 'cancer', 'depression', 'arthritis']\n",
        "cols_keep_9 = comorb + [ 't0_eq_vas', 't1_eq_vas','oks_t0_score', 'oks_t1_score']\n",
        "\n",
        "# columns with 9 as sentinel value (the ^ operator performs an XOR test, keeping only features that are in cols_with_9, but not in cols_keep_9)\n",
        "cols_sentinel_9 = list(set(cols_with_9) ^ set(cols_keep_9))\n",
        "df_temp[cols_sentinel_9].head(20)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-51d2e05d-c8ac-494e-8360-e5da3a83df10\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>oks_t0_pain</th>\n",
              "      <th>t0_mobility</th>\n",
              "      <th>oks_t0_washing</th>\n",
              "      <th>oks_t1_pain</th>\n",
              "      <th>oks_t1_night_pain</th>\n",
              "      <th>oks_t0_limping</th>\n",
              "      <th>oks_t1_transport</th>\n",
              "      <th>oks_t1_washing</th>\n",
              "      <th>oks_t1_standing</th>\n",
              "      <th>t0_self_care</th>\n",
              "      <th>oks_t0_stairs</th>\n",
              "      <th>t0_assisted</th>\n",
              "      <th>oks_t1_confidence</th>\n",
              "      <th>t0_symptom_period</th>\n",
              "      <th>oks_t0_walking</th>\n",
              "      <th>oks_t1_shopping</th>\n",
              "      <th>t0_activity</th>\n",
              "      <th>oks_t0_kneeling</th>\n",
              "      <th>oks_t0_work</th>\n",
              "      <th>oks_t1_limping</th>\n",
              "      <th>oks_t1_work</th>\n",
              "      <th>t0_anxiety</th>\n",
              "      <th>oks_t0_standing</th>\n",
              "      <th>t0_previous_surgery</th>\n",
              "      <th>oks_t0_shopping</th>\n",
              "      <th>oks_t1_stairs</th>\n",
              "      <th>oks_t0_confidence</th>\n",
              "      <th>oks_t1_walking</th>\n",
              "      <th>t0_discomfort</th>\n",
              "      <th>oks_t0_transport</th>\n",
              "      <th>oks_t1_kneeling</th>\n",
              "      <th>t0_disability</th>\n",
              "      <th>t0_living_arrangements</th>\n",
              "      <th>oks_t0_night_pain</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>9</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>9</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>9</td>\n",
              "      <td>9</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>9</td>\n",
              "      <td>9</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>9</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>9</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>9</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>9</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-51d2e05d-c8ac-494e-8360-e5da3a83df10')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-51d2e05d-c8ac-494e-8360-e5da3a83df10 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-51d2e05d-c8ac-494e-8360-e5da3a83df10');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "    oks_t0_pain  t0_mobility  oks_t0_washing  oks_t1_pain  oks_t1_night_pain  \\\n",
              "0             0            2               2            3                  3   \n",
              "1             1            9               4            4                  4   \n",
              "2             1            9               4            4                  4   \n",
              "3             1            9               2            3                  2   \n",
              "4             0            2               4            2                  2   \n",
              "5             0            2               4            1                  1   \n",
              "6             1            1               4            3                  4   \n",
              "7             0            2               2            4                  4   \n",
              "8             0            2               3            3                  3   \n",
              "9             1            2               4            4                  4   \n",
              "10            1            2               2            2                  2   \n",
              "11            1            2               3            1                  2   \n",
              "12            0            2               3            3                  4   \n",
              "13            0            2               4            3                  3   \n",
              "14            1            2               4            4                  4   \n",
              "15            1            2               3            9                  9   \n",
              "16            1            2               3            2                  2   \n",
              "17            0            2               1            1                  1   \n",
              "18            1            1               4            4                  4   \n",
              "19            1            2               2            1                  2   \n",
              "\n",
              "    oks_t0_limping  oks_t1_transport  oks_t1_washing  oks_t1_standing  \\\n",
              "0                0                 3               4                3   \n",
              "1                1                 3               4                4   \n",
              "2                2                 4               4                4   \n",
              "3                2                 4               4                4   \n",
              "4                0                 2               4                3   \n",
              "5                3                 4               3                3   \n",
              "6                3                 4               4                3   \n",
              "7                1                 3               4                3   \n",
              "8                2                 2               4                3   \n",
              "9                3                 4               4                4   \n",
              "10               1                 3               4                3   \n",
              "11               2                 2               3                2   \n",
              "12               1                 4               4                3   \n",
              "13               2                 3               4                3   \n",
              "14               1                 4               4                4   \n",
              "15               2                 9               9                9   \n",
              "16               1                 3               4                4   \n",
              "17               1                 2               3                3   \n",
              "18               2                 4               4                4   \n",
              "19               1                 2               0                3   \n",
              "\n",
              "    t0_self_care  oks_t0_stairs  t0_assisted  oks_t1_confidence  \\\n",
              "0              1              2            2                  4   \n",
              "1              9              4            2                  3   \n",
              "2              9              4            2                  4   \n",
              "3              9              2            2                  3   \n",
              "4              1              0            1                  3   \n",
              "5              1              3            1                  4   \n",
              "6              1              2            2                  4   \n",
              "7              2              1            2                  4   \n",
              "8              1              1            2                  3   \n",
              "9              1              3            1                  4   \n",
              "10             2              2            2                  4   \n",
              "11             1              3            1                  3   \n",
              "12             1              2            2                  4   \n",
              "13             1              3            2                  3   \n",
              "14             1              1            1                  4   \n",
              "15             2              1            1                  9   \n",
              "16             1              2            2                  4   \n",
              "17             1              1            2                  4   \n",
              "18             1              3            2                  4   \n",
              "19             2              2            2                  4   \n",
              "\n",
              "    t0_symptom_period  oks_t0_walking  oks_t1_shopping  t0_activity  \\\n",
              "0                   2               2                4            2   \n",
              "1                   2               0                4            9   \n",
              "2                   2               2                4            9   \n",
              "3                   2               2                4            9   \n",
              "4                   3               2                2            2   \n",
              "5                   4               3                3            2   \n",
              "6                   3               4                4            1   \n",
              "7                   2               2                4            3   \n",
              "8                   3               2                0            2   \n",
              "9                   2               2                4            2   \n",
              "10                  2               2                4            2   \n",
              "11                  2               3                2            1   \n",
              "12                  2               2                4            2   \n",
              "13                  4               3                3            2   \n",
              "14                  2               3                4            3   \n",
              "15                  2               2                9            2   \n",
              "16                  2               3                4            2   \n",
              "17                  3               2                4            2   \n",
              "18                  4               3                4            1   \n",
              "19                  2               2                3            2   \n",
              "\n",
              "    oks_t0_kneeling  oks_t0_work  oks_t1_limping  oks_t1_work  t0_anxiety  \\\n",
              "0                 1            1               4            4           1   \n",
              "1                 4            4               4            3           9   \n",
              "2                 3            3               4            4           9   \n",
              "3                 0            1               4            3           9   \n",
              "4                 1            1               3            2           1   \n",
              "5                 1            2               3            2           1   \n",
              "6                 2            3               4            3           1   \n",
              "7                 0            0               4            4           1   \n",
              "8                 1            2               3            3           1   \n",
              "9                 0            2               4            4           1   \n",
              "10                1            2               3            3           1   \n",
              "11                1            3               0            1           1   \n",
              "12                0            1               4            4           1   \n",
              "13                2            3               3            3           1   \n",
              "14                0            1               4            4           1   \n",
              "15                0            2               9            9           1   \n",
              "16                2            2               4            3           1   \n",
              "17                1            2               3            3           1   \n",
              "18                1            2               4            4           1   \n",
              "19                0            1               2            3           3   \n",
              "\n",
              "    oks_t0_standing  t0_previous_surgery  oks_t0_shopping  oks_t1_stairs  \\\n",
              "0                 2                    2                2              3   \n",
              "1                 4                    2                4              4   \n",
              "2                 4                    2                3              4   \n",
              "3                 2                    2                3              3   \n",
              "4                 1                    2                0              1   \n",
              "5                 2                    2                2              3   \n",
              "6                 2                    2                3              3   \n",
              "7                 1                    2                1              4   \n",
              "8                 1                    2                0              1   \n",
              "9                 3                    2                3              4   \n",
              "10                1                    2                2              4   \n",
              "11                3                    2                3              1   \n",
              "12                1                    2                3              4   \n",
              "13                3                    2                3              0   \n",
              "14                2                    2                1              3   \n",
              "15                3                    1                1              9   \n",
              "16                3                    2                3              3   \n",
              "17                2                    2                3              2   \n",
              "18                2                    9                3              3   \n",
              "19                2                    2                0              2   \n",
              "\n",
              "    oks_t0_confidence  oks_t1_walking  t0_discomfort  oks_t0_transport  \\\n",
              "0                   2               3              2                 2   \n",
              "1                   3               4              9                 4   \n",
              "2                   3               4              9                 4   \n",
              "3                   3               1              9                 2   \n",
              "4                   0               2              9                 2   \n",
              "5                   4               2              9                 4   \n",
              "6                   3               4              2                 3   \n",
              "7                   0               4              2                 1   \n",
              "8                   1               1              3                 2   \n",
              "9                   2               4              2                 4   \n",
              "10                  3               3              2                 2   \n",
              "11                  3               2              2                 3   \n",
              "12                  4               4              3                 2   \n",
              "13                  2               4              2                 2   \n",
              "14                  1               4              2                 3   \n",
              "15                  4               9              2                 2   \n",
              "16                  3               4              1                 2   \n",
              "17                  3               3              2                 2   \n",
              "18                  3               4              2                 4   \n",
              "19                  2               3              2                 2   \n",
              "\n",
              "    oks_t1_kneeling  t0_disability  t0_living_arrangements  oks_t0_night_pain  \n",
              "0                 2              1                       2                  1  \n",
              "1                 3              2                       2                  4  \n",
              "2                 2              2                       1                  4  \n",
              "3                 1              1                       2                  2  \n",
              "4                 2              1                       2                  1  \n",
              "5                 1              2                       1                  1  \n",
              "6                 2              2                       1                  2  \n",
              "7                 3              1                       1                  1  \n",
              "8                 1              1                       2                  1  \n",
              "9                 4              2                       1                  2  \n",
              "10                2              2                       1                  1  \n",
              "11                0              2                       1                  2  \n",
              "12                2              2                       1                  2  \n",
              "13                2              2                       1                  1  \n",
              "14                3              9                       1                  2  \n",
              "15                9              2                       2                  4  \n",
              "16                2              2                       9                  1  \n",
              "17                1              2                       1                  1  \n",
              "18                1              2                       1                  1  \n",
              "19                0              1                       1                  1  "
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MHXCCci5uIIZ"
      },
      "source": [
        "Now we are ready to impute the missings using the median (for numeric variables) and most frequent value (for categorical variables). However, we first need to be aware of so-called 'data-leaking': \n",
        "\n",
        "When we impute missing values using information from other data instances, we are also using information from data instances that will end up in our test set. This will give our process of model development an unfair advantage, and will result in an overly optimistic assessment of model performance. \n",
        "\n",
        "Even though the degree of data-leakage is very small when imputing using the median and the most frequent value, the correct thing to do is to first split the data into a training and test set. Sklearn has a useful function for this."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pZCVOTIukAj7"
      },
      "source": [
        "# randomly splitting the data into a train and test set\n",
        "df_train, df_test = train_test_split(df_temp, test_size=0.3, random_state=0)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GihvVg4ukaTq"
      },
      "source": [
        "Now we can impute the values in the training set, using the readily available functionality in the sklearn module. We work with a copy of the training data to be able to check our work and compare the imputed and not imputed versions of the data."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 363
        },
        "id": "NSwOZIodZqNJ",
        "outputId": "65266165-61d6-44be-db48-a30dbb52f2b0"
      },
      "source": [
        "from sklearn.impute import SimpleImputer\n",
        "\n",
        "impute_median = SimpleImputer(strategy='median')\n",
        "impute_most_frequent = SimpleImputer(strategy='most_frequent')\n",
        "\n",
        "# use copy of original data\n",
        "dfc = df_train.copy()\n",
        "\n",
        "# replace 9 with np.nan \n",
        "dfc.loc[:,cols_sentinel_9] = df_train.loc[:,cols_sentinel_9].replace(9, np.nan)\n",
        "\n",
        "# determine most frequent value for imputation\n",
        "impute_most_frequent.fit(dfc[cols_sentinel_9])\n",
        "pd.DataFrame({'columns': cols_sentinel_9, 'most_frequent': impute_most_frequent.statistics_}).head(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-a33aeeb1-00d7-4517-8162-624bdc256e81\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>columns</th>\n",
              "      <th>most_frequent</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>oks_t0_pain</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>t0_mobility</td>\n",
              "      <td>2.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>oks_t0_washing</td>\n",
              "      <td>4.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>oks_t1_pain</td>\n",
              "      <td>3.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>oks_t1_night_pain</td>\n",
              "      <td>4.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>oks_t0_limping</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>oks_t1_transport</td>\n",
              "      <td>4.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>oks_t1_washing</td>\n",
              "      <td>4.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>oks_t1_standing</td>\n",
              "      <td>3.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>t0_self_care</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a33aeeb1-00d7-4517-8162-624bdc256e81')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-a33aeeb1-00d7-4517-8162-624bdc256e81 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-a33aeeb1-00d7-4517-8162-624bdc256e81');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "             columns  most_frequent\n",
              "0        oks_t0_pain            0.0\n",
              "1        t0_mobility            2.0\n",
              "2     oks_t0_washing            4.0\n",
              "3        oks_t1_pain            3.0\n",
              "4  oks_t1_night_pain            4.0\n",
              "5     oks_t0_limping            0.0\n",
              "6   oks_t1_transport            4.0\n",
              "7     oks_t1_washing            4.0\n",
              "8    oks_t1_standing            3.0\n",
              "9       t0_self_care            1.0"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Uy6ofUZam50J"
      },
      "source": [
        "All the 9s in the relevant columns have now been replaced by missings. Can you print the missing values for all the columns that had 9 as a sentinel value?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LNvITQdWmd32",
        "outputId": "47352601-145c-4fb6-ddb8-0743fa13f3aa"
      },
      "source": [
        "# Print number of missing in columns with 9 as sentinal value\n",
        "dfc[cols_sentinel_9].isnull().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "oks_t0_pain                133\n",
              "t0_mobility               3045\n",
              "oks_t0_washing              87\n",
              "oks_t1_pain                881\n",
              "oks_t1_night_pain          611\n",
              "oks_t0_limping             989\n",
              "oks_t1_transport           843\n",
              "oks_t1_washing             843\n",
              "oks_t1_standing            703\n",
              "t0_self_care              3113\n",
              "oks_t0_stairs              941\n",
              "t0_assisted               1084\n",
              "oks_t1_confidence          654\n",
              "t0_symptom_period          844\n",
              "oks_t0_walking            1007\n",
              "oks_t1_shopping            696\n",
              "t0_activity               3157\n",
              "oks_t0_kneeling            977\n",
              "oks_t0_work                951\n",
              "oks_t1_limping             759\n",
              "oks_t1_work                697\n",
              "t0_anxiety                3638\n",
              "oks_t0_standing           1010\n",
              "t0_previous_surgery        745\n",
              "oks_t0_shopping            936\n",
              "oks_t1_stairs              676\n",
              "oks_t0_confidence          937\n",
              "oks_t1_walking             883\n",
              "t0_discomfort             3936\n",
              "oks_t0_transport           944\n",
              "oks_t1_kneeling            757\n",
              "t0_disability             4116\n",
              "t0_living_arrangements    1479\n",
              "oks_t0_night_pain          917\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aaJpE3B3nBWJ"
      },
      "source": [
        "And these can now be imputed using the most frequent values:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TX9u06GEbYS1"
      },
      "source": [
        "# assign most frequent data to the missing values for the selected columns\n",
        "dfc.loc[:, cols_sentinel_9] = impute_most_frequent.transform(dfc[cols_sentinel_9])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a6l6CQ7BneiY"
      },
      "source": [
        "After which all the missings in these columns have been replaced. Can you print the missing again?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Gk5fTJz_njE_",
        "outputId": "d3cd897b-46f7-4f21-af75-b1f5320568e7"
      },
      "source": [
        "dfc[cols_sentinel_9].isnull().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "oks_t0_pain               0\n",
              "t0_mobility               0\n",
              "oks_t0_washing            0\n",
              "oks_t1_pain               0\n",
              "oks_t1_night_pain         0\n",
              "oks_t0_limping            0\n",
              "oks_t1_transport          0\n",
              "oks_t1_washing            0\n",
              "oks_t1_standing           0\n",
              "t0_self_care              0\n",
              "oks_t0_stairs             0\n",
              "t0_assisted               0\n",
              "oks_t1_confidence         0\n",
              "t0_symptom_period         0\n",
              "oks_t0_walking            0\n",
              "oks_t1_shopping           0\n",
              "t0_activity               0\n",
              "oks_t0_kneeling           0\n",
              "oks_t0_work               0\n",
              "oks_t1_limping            0\n",
              "oks_t1_work               0\n",
              "t0_anxiety                0\n",
              "oks_t0_standing           0\n",
              "t0_previous_surgery       0\n",
              "oks_t0_shopping           0\n",
              "oks_t1_stairs             0\n",
              "oks_t0_confidence         0\n",
              "oks_t1_walking            0\n",
              "t0_discomfort             0\n",
              "oks_t0_transport          0\n",
              "oks_t1_kneeling           0\n",
              "t0_disability             0\n",
              "t0_living_arrangements    0\n",
              "oks_t0_night_pain         0\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "O798gQs9nsYQ"
      },
      "source": [
        "But these were not the only categorical variables with missing information. We have to do apply the same procedure to age_band and gender"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ivAgp4JtoNED",
        "outputId": "da61bebb-ad9b-4026-9c23-3e05f21bab82"
      },
      "source": [
        "# Print number of missings\n",
        "dfc[['age_band', 'gender']].isnull().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "age_band    6551\n",
              "gender      6551\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Can you find the most frequent values for the 'age_band' and 'gender' features?"
      ],
      "metadata": {
        "id": "ZDhG4nQJu0Mp"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 112
        },
        "id": "TRH3MEEVoNhj",
        "outputId": "43072d29-85d1-488d-ce73-656b16818748"
      },
      "source": [
        "# Find most frequent values\n",
        "impute_most_frequent.fit(dfc[['age_band', 'gender']])\n",
        "pd.DataFrame({'columns': ['age_band', 'gender'], 'most_frequent': impute_most_frequent.statistics_})"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-eb7d03b3-53f6-4680-8a53-c29c6de620b3\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>columns</th>\n",
              "      <th>most_frequent</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>age_band</td>\n",
              "      <td>70 to 79</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>gender</td>\n",
              "      <td>2.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-eb7d03b3-53f6-4680-8a53-c29c6de620b3')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-eb7d03b3-53f6-4680-8a53-c29c6de620b3 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-eb7d03b3-53f6-4680-8a53-c29c6de620b3');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "    columns most_frequent\n",
              "0  age_band      70 to 79\n",
              "1    gender           2.0"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "And can you assign these most frequent values for the 'age_band' and 'gender' features?"
      ],
      "metadata": {
        "id": "s62xV8NRvG13"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_mCMZtr3qQbn"
      },
      "source": [
        "# assign most frequent data to the missing values for the age and gender\n",
        "dfc.loc[:, ['age_band', 'gender']] = impute_most_frequent.transform(dfc[['age_band', 'gender']])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "l8T5MF2rqaqa",
        "outputId": "de9c8111-fb0f-4c28-ebb8-0ff43eddfd59"
      },
      "source": [
        "# and our check:\n",
        "dfc[['age_band', 'gender']].isnull().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "age_band    0\n",
              "gender      0\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nmex5XUbsxtO"
      },
      "source": [
        "The eq_vas variables used 999 as sentinal values, so we can apply a similar procedure to these variables. As these variables contain numeric values, we use the median for imputation, rather than the most frequent value:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xKJtMS0H3GRK",
        "outputId": "8c083421-bb9f-468c-bf99-68b4852f561d"
      },
      "source": [
        "# same procedure for 999 sentinel values, using impute median\n",
        "eq_vas = ['t0_eq_vas', 't1_eq_vas']\n",
        "dfc.loc[:, eq_vas] = dfc.loc[:,eq_vas].replace(999, np.nan)\n",
        "\n",
        "\n",
        "impute_median.fit(dfc.loc[:,eq_vas])\n",
        "dfc.loc[:,eq_vas] = impute_median.transform(dfc[eq_vas])\n",
        "impute_median.statistics_"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([70., 80.])"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zKtWW3u5ttTR",
        "outputId": "2b5bd08c-4dcf-47b6-db37-58d4da8dcae8"
      },
      "source": [
        "dfc.isnull().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "provider_code                       0\n",
              "procedure                           0\n",
              "revision_flag                       0\n",
              "year                                0\n",
              "age_band                            0\n",
              "gender                              0\n",
              "t0_assisted                         0\n",
              "t0_symptom_period                   0\n",
              "t0_previous_surgery                 0\n",
              "t0_living_arrangements              0\n",
              "t0_disability                       0\n",
              "heart_disease                       0\n",
              "high_bp                             0\n",
              "stroke                              0\n",
              "circulation                         0\n",
              "lung_disease                        0\n",
              "diabetes                            0\n",
              "kidney_disease                      0\n",
              "nervous_system                      0\n",
              "liver_disease                       0\n",
              "cancer                              0\n",
              "depression                          0\n",
              "arthritis                           0\n",
              "t0_mobility                         0\n",
              "t0_self_care                        0\n",
              "t0_activity                         0\n",
              "t0_discomfort                       0\n",
              "t0_anxiety                          0\n",
              "t0_eq5d_index                    5279\n",
              "oks_eq_5d_index_t1_predicted    10058\n",
              "t0_eq_vas                           0\n",
              "t1_eq_vas                           0\n",
              "oks_eq_vas_t1_predicted         13669\n",
              "oks_t0_pain                         0\n",
              "oks_t0_night_pain                   0\n",
              "oks_t0_washing                      0\n",
              "oks_t0_transport                    0\n",
              "oks_t0_walking                      0\n",
              "oks_t0_standing                     0\n",
              "oks_t0_limping                      0\n",
              "oks_t0_kneeling                     0\n",
              "oks_t0_work                         0\n",
              "oks_t0_confidence                   0\n",
              "oks_t0_shopping                     0\n",
              "oks_t0_stairs                       0\n",
              "oks_t0_score                     1159\n",
              "oks_t1_pain                         0\n",
              "oks_t1_night_pain                   0\n",
              "oks_t1_washing                      0\n",
              "oks_t1_transport                    0\n",
              "oks_t1_walking                      0\n",
              "oks_t1_standing                     0\n",
              "oks_t1_limping                      0\n",
              "oks_t1_kneeling                     0\n",
              "oks_t1_work                         0\n",
              "oks_t1_confidence                   0\n",
              "oks_t1_shopping                     0\n",
              "oks_t1_stairs                       0\n",
              "oks_t1_score                     1776\n",
              "oks_oks_t1_predicted             3754\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FFaphcjZyCqe"
      },
      "source": [
        "There are still a few variables containing missings:\n",
        "- t0_eq5d_index is a numeric variable which can be imputed using the median\n",
        "- we will leave the missings in the t1_variables for now (oks_t1_score, oks_oks_t1_predicted, oks_eq_vas_t1_predicted and oks_eq_5d_index_t1_predicted)\n",
        "- for the sake of the example, we can decide to exclude all instances where oks_t0_scores are missing, which has the implication that our model can only be applied in clinical practice to those patients for which oks_t0_scores are available."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DbKLbH8jzSiC",
        "outputId": "e20cab0b-1a40-4735-cee7-1d97720af20a"
      },
      "source": [
        "# impute median for t0_eq5d_index\n",
        "impute_median.fit(dfc.loc[:,'t0_eq5d_index'].to_frame())\n",
        "dfc.loc[:,'t0_eq5d_index'] = impute_median.transform(dfc['t0_eq5d_index'].to_frame())\n",
        "impute_median.statistics_"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.587], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hIvgQRHx1W8e"
      },
      "source": [
        "# remove instances with missing oks_t0_scores\n",
        "dfc = dfc.dropna(subset=['oks_t0_score'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tcOKQqLq3626"
      },
      "source": [
        "## Convert data types\n",
        "\n",
        "As we have seen last week, some features that actually indicate boolean variables were sometimes indicates with other values (like 1 and 9). We now transform variables into the suitable data type (True and False), such that the data are in the proper format for building models."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2s0BJW7zcmtJ"
      },
      "source": [
        "# transform into boolean\n",
        "dfc['female'] = dfc.loc[:,'gender'].replace({1: False, 2: True})\n",
        "dfc.loc[:, comorb] = dfc.loc[:, comorb].replace({9: False, 1: True})\n",
        "boolean = ['t0_assisted', 't0_previous_surgery', 't0_disability']\n",
        "dfc.loc[:, boolean] = dfc.loc[:, boolean].replace({1: True, 2: False})"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "miNa_eJS4iU6"
      },
      "source": [
        "With the function below, we can describe how often True and False occurs within these boolean values: "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 551
        },
        "id": "3s3U5oq00ton",
        "outputId": "ab0dcba5-0a60-4e17-bba2-383207213ba6"
      },
      "source": [
        "# helper function for counting boolean attribues\n",
        "def count_boolean(series):\n",
        "    '''\n",
        "    Returns absolute and normalized value counts of pd.series as a dataframe with \n",
        "    index = series.name\n",
        "    columns with absolute and normalized counts of each value\n",
        "    '''\n",
        "    try:\n",
        "        count = series.value_counts().to_frame().transpose()\n",
        "        norm = series.value_counts(normalize=True).to_frame().transpose()\n",
        "        return count.join(norm, lsuffix='_count', rsuffix='_normalized') \n",
        "    except:\n",
        "        print('Error: expecting a pandas.Series object as input. \\n' + count_boolean.__doc__)\n",
        "        return None\n",
        "\n",
        "pd.concat([count_boolean(dfc[col]) for col in ['female'] + comorb + boolean]).round(2)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-dbb5778d-557e-4981-8da3-96b99647480b\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>True_count</th>\n",
              "      <th>False_count</th>\n",
              "      <th>True_normalized</th>\n",
              "      <th>False_normalized</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>female</th>\n",
              "      <td>57483</td>\n",
              "      <td>38631</td>\n",
              "      <td>0.60</td>\n",
              "      <td>0.40</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>heart_disease</th>\n",
              "      <td>8936</td>\n",
              "      <td>87178</td>\n",
              "      <td>0.09</td>\n",
              "      <td>0.91</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>high_bp</th>\n",
              "      <td>42362</td>\n",
              "      <td>53752</td>\n",
              "      <td>0.44</td>\n",
              "      <td>0.56</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>stroke</th>\n",
              "      <td>1588</td>\n",
              "      <td>94526</td>\n",
              "      <td>0.02</td>\n",
              "      <td>0.98</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>circulation</th>\n",
              "      <td>5417</td>\n",
              "      <td>90697</td>\n",
              "      <td>0.06</td>\n",
              "      <td>0.94</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>lung_disease</th>\n",
              "      <td>8763</td>\n",
              "      <td>87351</td>\n",
              "      <td>0.09</td>\n",
              "      <td>0.91</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>diabetes</th>\n",
              "      <td>11920</td>\n",
              "      <td>84194</td>\n",
              "      <td>0.12</td>\n",
              "      <td>0.88</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>kidney_disease</th>\n",
              "      <td>1923</td>\n",
              "      <td>94191</td>\n",
              "      <td>0.02</td>\n",
              "      <td>0.98</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>nervous_system</th>\n",
              "      <td>976</td>\n",
              "      <td>95138</td>\n",
              "      <td>0.01</td>\n",
              "      <td>0.99</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>liver_disease</th>\n",
              "      <td>533</td>\n",
              "      <td>95581</td>\n",
              "      <td>0.01</td>\n",
              "      <td>0.99</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cancer</th>\n",
              "      <td>5171</td>\n",
              "      <td>90943</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.95</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>depression</th>\n",
              "      <td>8962</td>\n",
              "      <td>87152</td>\n",
              "      <td>0.09</td>\n",
              "      <td>0.91</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>arthritis</th>\n",
              "      <td>74007</td>\n",
              "      <td>22107</td>\n",
              "      <td>0.77</td>\n",
              "      <td>0.23</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_assisted</th>\n",
              "      <td>14672</td>\n",
              "      <td>81442</td>\n",
              "      <td>0.15</td>\n",
              "      <td>0.85</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_previous_surgery</th>\n",
              "      <td>7459</td>\n",
              "      <td>88655</td>\n",
              "      <td>0.08</td>\n",
              "      <td>0.92</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>t0_disability</th>\n",
              "      <td>50603</td>\n",
              "      <td>45511</td>\n",
              "      <td>0.53</td>\n",
              "      <td>0.47</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-dbb5778d-557e-4981-8da3-96b99647480b')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-dbb5778d-557e-4981-8da3-96b99647480b button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-dbb5778d-557e-4981-8da3-96b99647480b');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "                     True_count  False_count  True_normalized  \\\n",
              "female                    57483        38631             0.60   \n",
              "heart_disease              8936        87178             0.09   \n",
              "high_bp                   42362        53752             0.44   \n",
              "stroke                     1588        94526             0.02   \n",
              "circulation                5417        90697             0.06   \n",
              "lung_disease               8763        87351             0.09   \n",
              "diabetes                  11920        84194             0.12   \n",
              "kidney_disease             1923        94191             0.02   \n",
              "nervous_system              976        95138             0.01   \n",
              "liver_disease               533        95581             0.01   \n",
              "cancer                     5171        90943             0.05   \n",
              "depression                 8962        87152             0.09   \n",
              "arthritis                 74007        22107             0.77   \n",
              "t0_assisted               14672        81442             0.15   \n",
              "t0_previous_surgery        7459        88655             0.08   \n",
              "t0_disability             50603        45511             0.53   \n",
              "\n",
              "                     False_normalized  \n",
              "female                           0.40  \n",
              "heart_disease                    0.91  \n",
              "high_bp                          0.56  \n",
              "stroke                           0.98  \n",
              "circulation                      0.94  \n",
              "lung_disease                     0.91  \n",
              "diabetes                         0.88  \n",
              "kidney_disease                   0.98  \n",
              "nervous_system                   0.99  \n",
              "liver_disease                    0.99  \n",
              "cancer                           0.95  \n",
              "depression                       0.91  \n",
              "arthritis                        0.23  \n",
              "t0_assisted                      0.85  \n",
              "t0_previous_surgery              0.92  \n",
              "t0_disability                    0.47  "
            ]
          },
          "metadata": {},
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Z1U4vEAc44Py"
      },
      "source": [
        "Finally, a few variables need to be defined as a categorical variable:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EwECD0NUlc3Z"
      },
      "source": [
        "# convert to categories\n",
        "dfc.loc[:, ['provider_code', 'age_band']] = dfc.loc[:, ['provider_code', 'age_band']].astype('category')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "68fM2bgo5Vwe"
      },
      "source": [
        "#Constructing data\n",
        "\n",
        "In this final step of the data preparation, we get to the part where we make adjustments to the data:\n",
        "- Defining the outcome Y\n",
        "- Redefine and/or construct features (feature engineering, possibly handling correlated input features)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "poHEugvVqrNz"
      },
      "source": [
        "## Selecting outcome Y\n",
        "\n",
        "Last week we already looked at defining a outcome way that can be used for predicting treatment success. This was based on knowledge from domain experts, which use the oks_score and whether the oks_score at t1 is above a certain threshold. We defined an outcome as good when the improvement in oks_score was larger than a minimum change threshold (chosen to be set at 13), OR when the absolute oks_score at t1 was above a certain threshold (chosen to be set at 43).\n",
        "\n",
        "Here we will we have a look at an laternative (data-driven) approach to defining an outcome measure based on the data. \n",
        "\n",
        "In the [narrative on outcomes of osteoarthritis](https://github.com/dkapitan/jads-nhs-proms/blob/master/references/hunter2019osteaoarthritis.pdf) it is stated that the [European League Against Rheumatism (EULAR)](https://eular.org) standard considers three primary symptoms, namely:\n",
        "- Hip/knee pain\n",
        "- Early morning stiffness (EMS)\n",
        "- Functional limitations\n",
        "\n",
        "Since we have no information on EMS, we take pain and functional limitations as the primary outcome parameters. We now turn to the NHS dataset to choose which (engineered) features are available as a measure for pain and physical functioning at `t1`."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3-K-eCza7I2H"
      },
      "source": [
        "## Feature engineering\n",
        "\n",
        "We start off with adding three simple features\n",
        "- a variable representing total pain, which is the sum of 'pain' and 'night_pain', \n",
        "- a variable representing total functioning, which is the sum of the OKS-scores on all the non-pain dimensions, and\n",
        "- a variable representing the total number of comorbidities\n",
        "\n",
        "It is common to start simple and return to and put more effort into feature engineering if model performance is deemed insufficient"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Bd1s7X0A7x1g"
      },
      "source": [
        "# define total pain and total number of comorbidities\n",
        "def oks_questions(t='t0'):\n",
        "  return [f'oks_{t}_pain', f'oks_{t}_night_pain', f'oks_{t}_washing',\n",
        "          f'oks_{t}_transport', f'oks_{t}_walking', f'oks_{t}_standing',\n",
        "          f'oks_{t}_limping', f'oks_{t}_kneeling', f'oks_{t}_work',\n",
        "          f'oks_{t}_confidence', f'oks_{t}_shopping', f'oks_{t}_stairs', ]\n",
        "\n",
        "# combine that contain 'pain' to contstruct oks_pain_total and the rest for oks_functioning_total\n",
        "for t in (\"t0\", \"t1\"):\n",
        "    dfc[f\"oks_{t}_pain_total\"] = dfc[f\"oks_{t}_pain\"] + dfc[f\"oks_{t}_night_pain\"]\n",
        "    dfc[f\"oks_{t}_functioning_total\"] = (\n",
        "        dfc.loc[:, [col for col in oks_questions(t) if \"pain\" not in col]]\n",
        "        .sum(axis=1)\n",
        "    )\n",
        "dfc[\"n_comorb\"] = dfc.loc[:, comorb].sum()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1tb9sSKIj5lR"
      },
      "source": [
        "### Pain measures\n",
        "Let's plot a histogram of the different pain measure scores at t1 to get an idea of the distributions."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F6-bQHmOj5lR",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 498
        },
        "outputId": "f6b58432-9d6c-43bb-dec1-97d3fe5876a6"
      },
      "source": [
        "dfc[['oks_t1_pain', 'oks_t1_night_pain', 'oks_t1_pain_total']].hist(bins=range(0,9), figsize=S);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x576 with 4 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bozCf8hOj5lS"
      },
      "source": [
        "### Physical functioning\n",
        "The same for the functioning measures."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sQjwIiYUj5lS",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 736
        },
        "outputId": "f537dcd6-c1bb-4493-df57-eb8abf8dd20b"
      },
      "source": [
        "dfc.loc[:, [col for col in oks_questions(\"t1\") if \"pain\" not in col],].hist(figsize=M);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 864x864 with 12 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "And its total score."
      ],
      "metadata": {
        "id": "ez2jI2Fomz-4"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Nlx8N4mqj5lT",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 501
        },
        "outputId": "4f8d03c9-7f84-4eeb-c457-4f14c6151bef"
      },
      "source": [
        "dfc['oks_t1_functioning_total'].hist(figsize=S);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PEbpIP8Kj5lT"
      },
      "source": [
        "### Combined outcome Y\n",
        "\n",
        "Defining the appropriate outcome Y that meaningfully captures clinical change is a well-known challenge, see for example [development of psychometric tests](https://github.com/dkapitan/jads-nhs-proms/blob/master/references/wise2004methods.pdf). \n",
        "\n",
        "Although this should always be defined with the help of domain experts, a data-driven approach to gather input on this decision could be to consider [the intersection between the 'pre' and 'post' histograms](http://blog.datadive.net/histogram-intersection-for-change-detection/), in this case pain and functioning measured at `t0` and `t1`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "32xFSvRuj5lU",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 505
        },
        "outputId": "2c09741b-a527-46d1-b54d-a21eedf90570"
      },
      "source": [
        "dfc[['oks_t0_pain_total', 'oks_t1_pain_total']].plot.hist(bins=range(0,9), alpha=0.5, figsize=S);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zArTSiNYj5lX",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 501
        },
        "outputId": "86b5b231-95ad-4a91-ba1f-6db88b9b75c3"
      },
      "source": [
        "dfc[['oks_t0_functioning_total', 'oks_t1_functioning_total']].plot.hist(bins=range(0,40), alpha=0.5, figsize=S);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aOSVpYGMj5lX"
      },
      "source": [
        "Using the histograms above, one way to combine pain and functioning into one outcome measure is to use the cut-off points:\n",
        "- `oks_pain_total >= 4`: good outcome with less pain\n",
        "- `oks_functioning_total >= 26`: good outcome with improved functioning\n",
        "\n",
        "The figures below visualizes the combination of both measures to show that most people move from scoring poor on both pain and functioning at `t0` to scoring well on both of these measures at `t1`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FokKxo-gj5lZ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "d6277192-95ad-4278-ddde-9ea4f35fafbd"
      },
      "source": [
        "BINS = 8\n",
        "SAMPLE = 1000\n",
        "\n",
        "# https://stackoverflow.com/questions/45540886/reduce-line-width-of-seaborn-timeseries-plot\n",
        "sns.set(rc={\"lines.linewidth\": 0.7})\n",
        "fig1_layout = {\n",
        "    'kind': \"hex\",\n",
        "    'height': 8,\n",
        "    'ratio': 3,\n",
        "    'marginal_kws': {'bins': BINS, 'rug': False},\n",
        "}\n",
        "j1 = (sns.jointplot(\n",
        "            'oks_t0_pain_total', 'oks_t0_functioning_total',\n",
        "            data=dfc[['oks_t0_pain_total', 'oks_t0_functioning_total']].head(SAMPLE),\n",
        "            **fig1_layout)\n",
        "         .plot_joint(sns.kdeplot, color='lawngreen'))\n",
        "j2 = (sns.jointplot(\n",
        "            'oks_t1_pain_total', 'oks_t1_functioning_total',\n",
        "            data=dfc[['oks_t1_pain_total', 'oks_t1_functioning_total']].head(SAMPLE), \n",
        "            **fig1_layout)\n",
        "         .plot_joint(sns.kdeplot, color='lawngreen' ));"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x576 with 3 Axes>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x576 with 3 Axes>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a4v8SiYBj5lZ"
      },
      "source": [
        "So let's add boolean columns indicating whether pain is 'good' (i.e. no pain) and functioning is good, both at `t0` and `t1`."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tyK1rMy1j5lZ"
      },
      "source": [
        "# add boolean columns\n",
        "for t in ('t0', 't1'):\n",
        "    dfc[f'y_{t}_pain_good'] = dfc[f'oks_{t}_pain_total'] >= 4\n",
        "    dfc[f'y_{t}_functioning_good'] = dfc[f'oks_{t}_functioning_total'] >= 26"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MEoek8trj5la",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 363
        },
        "outputId": "9c19c098-43f6-4286-aed8-0c3f4e7c7542"
      },
      "source": [
        "# inspect whether new features are correct\n",
        "dfc[[col for col in dfc.columns if '_good' in col]].head(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-c8d2521d-d5b3-4a03-881b-4e8e56f4a643\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>y_t0_pain_good</th>\n",
              "      <th>y_t0_functioning_good</th>\n",
              "      <th>y_t1_pain_good</th>\n",
              "      <th>y_t1_functioning_good</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>8990</th>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>True</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>30721</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>132380</th>\n",
              "      <td>True</td>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>26737</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>85796</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>127528</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>53492</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>41139</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>66788</th>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>True</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>138709</th>\n",
              "      <td>False</td>\n",
              "      <td>True</td>\n",
              "      <td>False</td>\n",
              "      <td>False</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-c8d2521d-d5b3-4a03-881b-4e8e56f4a643')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-c8d2521d-d5b3-4a03-881b-4e8e56f4a643 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-c8d2521d-d5b3-4a03-881b-4e8e56f4a643');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "        y_t0_pain_good  y_t0_functioning_good  y_t1_pain_good  \\\n",
              "8990             False                   True            True   \n",
              "30721            False                  False            True   \n",
              "132380            True                  False            True   \n",
              "26737            False                  False            True   \n",
              "85796            False                  False            True   \n",
              "127528           False                  False            True   \n",
              "53492            False                  False           False   \n",
              "41139            False                  False            True   \n",
              "66788            False                  False            True   \n",
              "138709           False                   True           False   \n",
              "\n",
              "        y_t1_functioning_good  \n",
              "8990                     True  \n",
              "30721                    True  \n",
              "132380                   True  \n",
              "26737                    True  \n",
              "85796                   False  \n",
              "127528                   True  \n",
              "53492                   False  \n",
              "41139                    True  \n",
              "66788                    True  \n",
              "138709                  False  "
            ]
          },
          "metadata": {},
          "execution_count": 38
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EQ3FB6weCt0B"
      },
      "source": [
        "To get an idea how often the different combinations of True and False occur for these scores, before and after treatment, it is useful to print a frequency table. The crosstab function from Pandas is very useful for this. It describes the frequency with which the different True/False combinations occur in the data."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ugAMLFTjj5la",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "outputId": "d3e20e5c-eaff-43ef-e304-5fbbfa030b5a"
      },
      "source": [
        "pd.crosstab(dfc.y_t0_pain_good, dfc.y_t0_functioning_good, normalize=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-6cdf63bf-ccd6-4afc-af32-2beb7d4bf433\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>y_t0_functioning_good</th>\n",
              "      <th>False</th>\n",
              "      <th>True</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>y_t0_pain_good</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>False</th>\n",
              "      <td>0.81</td>\n",
              "      <td>0.07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>True</th>\n",
              "      <td>0.08</td>\n",
              "      <td>0.05</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6cdf63bf-ccd6-4afc-af32-2beb7d4bf433')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-6cdf63bf-ccd6-4afc-af32-2beb7d4bf433 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-6cdf63bf-ccd6-4afc-af32-2beb7d4bf433');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "y_t0_functioning_good  False  True\n",
              "y_t0_pain_good                    \n",
              "False                   0.81  0.07\n",
              "True                    0.08  0.05"
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Can you calculate the crosstab for at t1?"
      ],
      "metadata": {
        "id": "U8GQAnalNGnG"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g4LaJtHBj5la",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "outputId": "3cf538ff-7bed-4ea0-f908-356d36c2b953"
      },
      "source": [
        "pd.crosstab(dfc.y_t1_pain_good, dfc.y_t1_functioning_good, normalize=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-e477a149-c570-413e-a768-b89860d12d51\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>y_t1_functioning_good</th>\n",
              "      <th>False</th>\n",
              "      <th>True</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>y_t1_pain_good</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>False</th>\n",
              "      <td>0.15</td>\n",
              "      <td>0.06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>True</th>\n",
              "      <td>0.08</td>\n",
              "      <td>0.71</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e477a149-c570-413e-a768-b89860d12d51')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-e477a149-c570-413e-a768-b89860d12d51 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-e477a149-c570-413e-a768-b89860d12d51');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ],
            "text/plain": [
              "y_t1_functioning_good  False  True\n",
              "y_t1_pain_good                    \n",
              "False                   0.15  0.06\n",
              "True                    0.08  0.71"
            ]
          },
          "metadata": {},
          "execution_count": 40
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "On average, there seems to be a clear improvement based on the thresholds we have chosen. 71% has a good score on both engineered outcome measure, which is similar to the outcomes we defined last week based on domain expertise. Such an automatic, data-driven, approach can thus be a good solution if domain expertise is lacking or undefined."
      ],
      "metadata": {
        "id": "UkSbRV9JqvyT"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UXRVdLXMUrCy"
      },
      "source": [
        "## Conclusion and reflection\n",
        "\n",
        "Today we have worked through the phase of data preparation. This involved the following steps: \n",
        "\n",
        "### Making the data 'tidy'\n",
        "\n",
        "- Each variable must have its own column\n",
        "- Each observation must have its own row\n",
        "- Each value must have its own cell\n",
        "\n",
        "### Selecting relevant information \n",
        "\n",
        "- Select appropriate observations \n",
        "- Select available features\n",
        "- Select useful features\n",
        "\n",
        "### Cleaning data\n",
        "\n",
        "- Clean inconsistencies\n",
        "- Handle missing data\n",
        "- Setting appropriate data types\n",
        "\n",
        "### Constructing data\n",
        "\n",
        "- Define outcome Y (clustered classes, categories)\n",
        "- Redefine and/or construct features (feature engineering)\n",
        "- Decide if and how to handle correlated input features\n",
        "\n",
        "\n",
        "You should expect to be returning to this stage of data preparation multiple times, trying to tweak the data such that model performance can further improve. As such, always make sure to set up your script / workflow in such a way that adjustments can be easily made."
      ]
    }
  ]
}