term_project_final.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "## Final Project\n",
        "### 20182650 Jang Jaehoon\n"
      ],
      "metadata": {
        "id": "kBr7i0eo7FUA"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Load Data using github"
      ],
      "metadata": {
        "id": "Xb0_1Ud37T-a"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!git clone https://github.com/leehahoon/question_quality_assessment.git git_dir"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kccZjbwj7KsU",
        "outputId": "c003fdb0-acba-426c-c9aa-41a9a83777a7"
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Cloning into 'git_dir'...\n",
            "remote: Enumerating objects: 983, done.\u001b[K\n",
            "remote: Counting objects: 100% (30/30), done.\u001b[K\n",
            "remote: Compressing objects: 100% (25/25), done.\u001b[K\n",
            "remote: Total 983 (delta 7), reused 26 (delta 5), pack-reused 953\u001b[K\n",
            "Receiving objects: 100% (983/983), 51.28 MiB | 25.04 MiB/s, done.\n",
            "Resolving deltas: 100% (9/9), done.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "data = pd.read_csv('git_dir/data/train_data/train_task_3_4.csv')\n",
        "data.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "0cQhBP1M7ZVg",
        "outputId": "9df5c5e4-099e-4001-af3a-ce16a5bd1b39"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   QuestionId  UserId  AnswerId  IsCorrect  CorrectAnswer  AnswerValue\n",
              "0         898    2111    280203          1              2            2\n",
              "1         767    3062     55638          1              3            3\n",
              "2         165    1156    386475          1              2            2\n",
              "3         490    1653    997498          1              4            4\n",
              "4         298    3912    578636          1              3            3"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-46fcf999-5f72-403a-bba2-a1d2eafb6ce3\">\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>QuestionId</th>\n",
              "      <th>UserId</th>\n",
              "      <th>AnswerId</th>\n",
              "      <th>IsCorrect</th>\n",
              "      <th>CorrectAnswer</th>\n",
              "      <th>AnswerValue</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>898</td>\n",
              "      <td>2111</td>\n",
              "      <td>280203</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>767</td>\n",
              "      <td>3062</td>\n",
              "      <td>55638</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>165</td>\n",
              "      <td>1156</td>\n",
              "      <td>386475</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>490</td>\n",
              "      <td>1653</td>\n",
              "      <td>997498</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>298</td>\n",
              "      <td>3912</td>\n",
              "      <td>578636</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-46fcf999-5f72-403a-bba2-a1d2eafb6ce3')\"\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-46fcf999-5f72-403a-bba2-a1d2eafb6ce3 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-46fcf999-5f72-403a-bba2-a1d2eafb6ce3');\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",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Get correct rate inverse score"
      ],
      "metadata": {
        "id": "OtsLWNIEP73Z"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "result_df = pd.DataFrame()\n",
        "result_df['Correct_Rate'] = data.groupby('QuestionId').agg(['mean'])['IsCorrect']\n",
        "result_df['Correct_Rate_Inverse'] = 1-result_df['Correct_Rate']\n"
      ],
      "metadata": {
        "id": "S8_72J79P0LK"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Get subject score"
      ],
      "metadata": {
        "id": "D2j0y7jAQEWa"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import ast\n",
        "\n",
        "question_metadata = pd.read_csv('git_dir/data/metadata/question_metadata_task_3_4.csv')\n",
        "subj_metadata = pd.read_csv('git_dir/data/metadata/subject_metadata.csv')\n",
        "\n",
        "question_data = question_metadata.sort_values('QuestionId')\n",
        "question_data['SubjectList'] = question_data['SubjectId'].apply(lambda x: ast.literal_eval(x))\n",
        "question_data['SubjectLen'] = question_data['SubjectList'].apply(len)\n",
        "question_data.set_index('QuestionId', inplace=True)\n",
        "question_data.head()\n",
        "# get SubjectList's length\n",
        "\n",
        "result_df['Subject_Score'] = 10-question_data['SubjectLen']\n",
        "result_df\n",
        "\n",
        "# The longer the length, the more diverse the problem area.\n",
        "# so, 10 - (SubjectLen)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 455
        },
        "id": "iFU06Md5QPhk",
        "outputId": "38443a2b-b6fe-48da-fb3b-01a7747a596f"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            Correct_Rate  Correct_Rate_Inverse  Subject_Score\n",
              "QuestionId                                                   \n",
              "0               0.443457              0.556543              6\n",
              "1               0.571429              0.428571              6\n",
              "2               0.385214              0.614786              4\n",
              "3               0.808757              0.191243              6\n",
              "4               0.401408              0.598592              6\n",
              "...                  ...                   ...            ...\n",
              "943             0.566528              0.433472              6\n",
              "944             0.142857              0.857143              6\n",
              "945             0.422336              0.577664              6\n",
              "946             0.459459              0.540541              6\n",
              "947             0.617124              0.382876              6\n",
              "\n",
              "[948 rows x 3 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-72a2c3b2-4489-4d3a-8c5b-48032935147e\">\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>Correct_Rate</th>\n",
              "      <th>Correct_Rate_Inverse</th>\n",
              "      <th>Subject_Score</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>QuestionId</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.443457</td>\n",
              "      <td>0.556543</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.571429</td>\n",
              "      <td>0.428571</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.385214</td>\n",
              "      <td>0.614786</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.808757</td>\n",
              "      <td>0.191243</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.401408</td>\n",
              "      <td>0.598592</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>943</th>\n",
              "      <td>0.566528</td>\n",
              "      <td>0.433472</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>944</th>\n",
              "      <td>0.142857</td>\n",
              "      <td>0.857143</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>945</th>\n",
              "      <td>0.422336</td>\n",
              "      <td>0.577664</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>946</th>\n",
              "      <td>0.459459</td>\n",
              "      <td>0.540541</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>947</th>\n",
              "      <td>0.617124</td>\n",
              "      <td>0.382876</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>948 rows × 3 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-72a2c3b2-4489-4d3a-8c5b-48032935147e')\"\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-72a2c3b2-4489-4d3a-8c5b-48032935147e 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-72a2c3b2-4489-4d3a-8c5b-48032935147e');\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",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Get answer value entropy"
      ],
      "metadata": {
        "id": "-MlJgVJHMdZf"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import scipy as sp\n",
        "from scipy.stats import multinomial\n",
        "import os\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import argparse\n",
        "\n",
        "result_df['AnswerValue_Entropy'] = data.groupby('QuestionId')['AnswerValue'].agg(lambda x:multinomial.entropy(1,x.value_counts(normalize=True)).mean())\n",
        "# reference : starter_kit - sample_model_entropy\n",
        "# A smaller value means the answer to the question is easier to predict."
      ],
      "metadata": {
        "id": "aCjLm45bMvpw"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Get question correct entropy"
      ],
      "metadata": {
        "id": "DHplVP-qMv_G"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "result_df['IsCorrect_Entropy'] = data.groupby('QuestionId')['IsCorrect'].agg(lambda x:multinomial.entropy(1,x.value_counts(normalize=True)).mean())\n",
        "# Calculates the entropy of the polynomial distribution of the IsCorrect column for each question\n",
        "# A smaller value means that the correct answer for that question is easier to predict."
      ],
      "metadata": {
        "id": "oeYK_O0-M21X"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Get characters in images"
      ],
      "metadata": {
        "id": "Kt1pKAPq4jcD"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Reference -> image2text.ipynb\n",
        "data = pd.read_csv('git_dir/data/metadata/imagesText.csv')\n",
        "data.set_index('QuestionId', inplace=True)\n",
        "result_df['QuestionTextLength'] = data['QuestionTextLength']\n",
        "result_df.head()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "id": "hUewAbH94uT0",
        "outputId": "427e4218-c059-4df2-be67-8029662a9cbd"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            Correct_Rate  Correct_Rate_Inverse  Subject_Score  \\\n",
              "QuestionId                                                      \n",
              "0               0.443457              0.556543              6   \n",
              "1               0.571429              0.428571              6   \n",
              "2               0.385214              0.614786              4   \n",
              "3               0.808757              0.191243              6   \n",
              "4               0.401408              0.598592              6   \n",
              "\n",
              "            AnswerValue_Entropy  IsCorrect_Entropy  QuestionTextLength  \n",
              "QuestionId                                                              \n",
              "0                      1.194456           0.686739                 141  \n",
              "1                      0.876006           0.682908                  46  \n",
              "2                      1.331663           0.666559                 105  \n",
              "3                      0.695423           0.488020                  53  \n",
              "4                      1.324321           0.673579                 125  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-0c282e91-020b-40c1-9e5b-dc41ed698ab8\">\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>Correct_Rate</th>\n",
              "      <th>Correct_Rate_Inverse</th>\n",
              "      <th>Subject_Score</th>\n",
              "      <th>AnswerValue_Entropy</th>\n",
              "      <th>IsCorrect_Entropy</th>\n",
              "      <th>QuestionTextLength</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>QuestionId</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.443457</td>\n",
              "      <td>0.556543</td>\n",
              "      <td>6</td>\n",
              "      <td>1.194456</td>\n",
              "      <td>0.686739</td>\n",
              "      <td>141</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.571429</td>\n",
              "      <td>0.428571</td>\n",
              "      <td>6</td>\n",
              "      <td>0.876006</td>\n",
              "      <td>0.682908</td>\n",
              "      <td>46</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.385214</td>\n",
              "      <td>0.614786</td>\n",
              "      <td>4</td>\n",
              "      <td>1.331663</td>\n",
              "      <td>0.666559</td>\n",
              "      <td>105</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.808757</td>\n",
              "      <td>0.191243</td>\n",
              "      <td>6</td>\n",
              "      <td>0.695423</td>\n",
              "      <td>0.488020</td>\n",
              "      <td>53</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.401408</td>\n",
              "      <td>0.598592</td>\n",
              "      <td>6</td>\n",
              "      <td>1.324321</td>\n",
              "      <td>0.673579</td>\n",
              "      <td>125</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0c282e91-020b-40c1-9e5b-dc41ed698ab8')\"\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-0c282e91-020b-40c1-9e5b-dc41ed698ab8 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-0c282e91-020b-40c1-9e5b-dc41ed698ab8');\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",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Rank features"
      ],
      "metadata": {
        "id": "PhhNtsPK4Zoc"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "rank_df = pd.DataFrame()\n",
        "\n",
        "rank_df['Correct_Rate_Inverse'] = result_df['Correct_Rate_Inverse'].rank(method='first', ascending=False).astype('int16')\n",
        "rank_df['CR_inverse+SS'] = (result_df['Correct_Rate_Inverse'] + result_df['Subject_Score']).rank(method='first', ascending=False).astype('int16')\n",
        "rank_df['QuestionTextLength'] = result_df['QuestionTextLength'].rank(method='first', ascending=False).astype('int16')\n",
        "rank_df['QuestionTextLength_Inverse'] = result_df['QuestionTextLength'].rank(method='first', ascending=True).astype('int16')\n",
        "rank_df['CR_inverse+QuestionTextLength'] = (result_df['Correct_Rate_Inverse'] + result_df['QuestionTextLength']).rank(method='first', ascending=False).astype('int16')\n",
        "rank_df['CR_inverse+SS+QuestionTextLength'] = (result_df['Correct_Rate_Inverse'] + result_df['QuestionTextLength'] + result_df['Subject_Score']).rank(method='first', ascending=False).astype('int16')\n",
        "\n",
        "rank_df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "id": "3jUk-AJi5agz",
        "outputId": "4f96d380-f0cd-49b3-e05a-758b098d9aed"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            Correct_Rate_Inverse  CR_inverse+SS  QuestionTextLength  \\\n",
              "QuestionId                                                            \n",
              "0                            355            345                 322   \n",
              "1                            619            608                 847   \n",
              "2                            228            948                 494   \n",
              "3                            922            908                 818   \n",
              "4                            273            264                 387   \n",
              "\n",
              "            QuestionTextLength_Inverse  CR_inverse+QuestionTextLength  \\\n",
              "QuestionId                                                              \n",
              "0                                  624                            322   \n",
              "1                                   98                            848   \n",
              "2                                  443                            497   \n",
              "3                                  130                            819   \n",
              "4                                  557                            389   \n",
              "\n",
              "            CR_inverse+SS+QuestionTextLength  \n",
              "QuestionId                                    \n",
              "0                                        322  \n",
              "1                                        848  \n",
              "2                                        512  \n",
              "3                                        819  \n",
              "4                                        389  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-0e4db651-6809-4880-8f7f-f0ff1318a301\">\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>Correct_Rate_Inverse</th>\n",
              "      <th>CR_inverse+SS</th>\n",
              "      <th>QuestionTextLength</th>\n",
              "      <th>QuestionTextLength_Inverse</th>\n",
              "      <th>CR_inverse+QuestionTextLength</th>\n",
              "      <th>CR_inverse+SS+QuestionTextLength</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>QuestionId</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>355</td>\n",
              "      <td>345</td>\n",
              "      <td>322</td>\n",
              "      <td>624</td>\n",
              "      <td>322</td>\n",
              "      <td>322</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>619</td>\n",
              "      <td>608</td>\n",
              "      <td>847</td>\n",
              "      <td>98</td>\n",
              "      <td>848</td>\n",
              "      <td>848</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>228</td>\n",
              "      <td>948</td>\n",
              "      <td>494</td>\n",
              "      <td>443</td>\n",
              "      <td>497</td>\n",
              "      <td>512</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>922</td>\n",
              "      <td>908</td>\n",
              "      <td>818</td>\n",
              "      <td>130</td>\n",
              "      <td>819</td>\n",
              "      <td>819</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>273</td>\n",
              "      <td>264</td>\n",
              "      <td>387</td>\n",
              "      <td>557</td>\n",
              "      <td>389</td>\n",
              "      <td>389</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0e4db651-6809-4880-8f7f-f0ff1318a301')\"\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-0e4db651-6809-4880-8f7f-f0ff1318a301 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-0e4db651-6809-4880-8f7f-f0ff1318a301');\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",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "rank_df['Correct_Rate_Inverse'].to_csv('Correct_Rate_Inverse.csv', header=True, index_label='QuestionId')\n",
        "rank_df['CR_inverse+SS'].to_csv('CR_inverse+SS.csv', header=True, index_label='QuestionId')\n",
        "rank_df['QuestionTextLength'].to_csv('QuestionTextLength.csv', header=True, index_label='QuestionId')\n",
        "rank_df['QuestionTextLength_Inverse'].to_csv('QuestionTextLength_Inverse.csv', header=True, index_label='QuestionId')\n",
        "rank_df['CR_inverse+QuestionTextLength'].to_csv('CR_inverse+QuestionTextLength.csv', header=True, index_label='QuestionId')\n",
        "rank_df['CR_inverse+SS+QuestionTextLength'].to_csv('CR_inverse+SS+QuestionTextLength.csv', header=True, index_label='QuestionId')\n",
        "\n"
      ],
      "metadata": {
        "id": "2QUSu902-9HD"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Evaluate (Public)"
      ],
      "metadata": {
        "id": "jjB5Ia9U7aqq"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# load evaluators' score on \"validation\" set\n",
        "eval_validation = pd.read_csv('git_dir/data/test_data/quality_response_remapped_public.csv')\n",
        "print(len(eval_validation))\n",
        "eval_validation\n",
        "\n",
        "# take the average of 5 evaulators` scores\n",
        "eval_validation['score'] = eval_validation.filter(regex='^T', axis = 1).mean(axis=1)\n"
      ],
      "metadata": {
        "id": "H30moPrCQnZB",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c7d63d2d-070e-4232-82f7-f9fb02c2adce"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "25\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# decide preference\n",
        "import numpy as np\n",
        "def calc_preference(scores):\n",
        "    preference = np.ones(len(scores), dtype=int)\n",
        "\n",
        "    idx_two = scores > 1.5\n",
        "    preference[idx_two] = 2\n",
        "\n",
        "    return list(preference)\n",
        "\n",
        "eval_validation['preference'] = calc_preference(eval_validation['score'])\n",
        "# eval_dev[]\n",
        "# eval_dev['score']"
      ],
      "metadata": {
        "id": "oseMpobFRnzz"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def evaluate(truth, submission):\n",
        "    # extract ranking\n",
        "    left = list(truth.left)\n",
        "    right = list(truth.right)\n",
        "#     if len(left) != len(right):\n",
        "#     message = 'left and right lengths are not the same'\n",
        "#     sys.exit(message)\n",
        "\n",
        "    submission_left = []\n",
        "    submission_right = []\n",
        "    submission_preference = []\n",
        "    for idx in range(len(left)):\n",
        "        submission_left.append(left[idx])\n",
        "        submission_right.append(right[idx])\n",
        "        ranking_left = submission[submission.QuestionId==left[idx]].ranking.values[0]\n",
        "        ranking_right = submission[submission.QuestionId==right[idx]].ranking.values[0]\n",
        "        preference = 1 if ranking_left < ranking_right else 2\n",
        "        submission_preference.append(preference)\n",
        "    #print(submission_preference)\n",
        "    return submission_preference"
      ],
      "metadata": {
        "id": "snC4sUM0Rq-b"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "csv_filelist = ['Correct_Rate_Inverse', 'CR_inverse+SS', 'QuestionTextLength', 'QuestionTextLength_Inverse', 'CR_inverse+QuestionTextLength', \\\n",
        "                'CR_inverse+SS+QuestionTextLength']\n",
        "templates = []\n",
        "for csv_file in csv_filelist:\n",
        "  templates.append(pd.read_csv(csv_file+'.csv'))\n",
        "\n",
        "for template in templates:\n",
        "  template.columns = ['QuestionId', 'ranking']\n",
        "\n",
        "print(templates[0])\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kHLbDXlIALDV",
        "outputId": "e1eb8622-2d11-4e99-d257-4aa1443019e8"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "     QuestionId  ranking\n",
            "0             0      355\n",
            "1             1      619\n",
            "2             2      228\n",
            "3             3      922\n",
            "4             4      273\n",
            "..          ...      ...\n",
            "943         943      613\n",
            "944         944        1\n",
            "945         945      310\n",
            "946         946      395\n",
            "947         947      710\n",
            "\n",
            "[948 rows x 2 columns]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "eval_pub_result = {}\n",
        "for idx in range(len(templates)):\n",
        "  eval_pub_result[csv_filelist[idx]] = evaluate(eval_validation, templates[idx])\n",
        "\n",
        "print(eval_pub_result)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "76VASVknCQTv",
        "outputId": "3f498bea-bb0c-45e6-e03f-fb1a17aa4a9a"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "{'Correct_Rate_Inverse': [2, 2, 2, 2, 2, 2, 1, 2, 1, 2, 2, 2, 1, 2, 1, 2, 1, 1, 2, 1, 1, 2, 1, 2, 1], 'CR_inverse+SS': [2, 2, 2, 2, 2, 2, 1, 2, 1, 2, 2, 2, 1, 2, 1, 2, 1, 2, 2, 1, 1, 2, 1, 2, 1], 'QuestionTextLength': [2, 2, 2, 1, 2, 2, 1, 2, 2, 1, 1, 2, 1, 1, 1, 1, 2, 1, 1, 1, 2, 2, 1, 2, 1], 'QuestionTextLength_Inverse': [1, 1, 1, 2, 1, 1, 2, 1, 1, 2, 2, 1, 2, 2, 2, 2, 1, 2, 2, 2, 1, 1, 2, 1, 2], 'CR_inverse+QuestionTextLength': [2, 2, 2, 1, 2, 2, 1, 2, 2, 1, 1, 2, 1, 1, 1, 1, 2, 1, 1, 1, 2, 2, 1, 2, 1], 'CR_inverse+SS+QuestionTextLength': [2, 2, 2, 1, 2, 2, 1, 2, 2, 1, 1, 2, 1, 1, 1, 1, 2, 1, 1, 1, 2, 2, 1, 2, 1]}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def getCompareDF(eval_validation, estimate_df):\n",
        "  df = pd.DataFrame()\n",
        "  df['truth'] = eval_validation['preference']\n",
        "  df['estimate'] = estimate_df\n",
        "  df['evaluate'] = df['truth'] == df['estimate']\n",
        "  df['score'] = df['evaluate'].apply(lambda x: 1 if x else 0)\n",
        "  total_score = df['score'].sum()\n",
        "  total_percent = total_score/df['score'].count()\n",
        "  return (total_score, total_percent)\n",
        "\n",
        "for key, value in eval_pub_result.items():\n",
        "  score, percent = getCompareDF(eval_validation, value)\n",
        "  print(f'{key}: {score}, {percent}')\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LbF3jgK8Dnm-",
        "outputId": "0528963c-10a5-42a6-f249-a8b69c31a9f7"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Correct_Rate_Inverse: 19, 0.76\n",
            "CR_inverse+SS: 20, 0.8\n",
            "QuestionTextLength: 14, 0.56\n",
            "QuestionTextLength_Inverse: 11, 0.44\n",
            "CR_inverse+QuestionTextLength: 14, 0.56\n",
            "CR_inverse+SS+QuestionTextLength: 14, 0.56\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Evaluate (Private)"
      ],
      "metadata": {
        "id": "BDLh2W80G5Ff"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "eval_test = pd.read_csv('git_dir/data/test_data/quality_response_remapped_private.csv')\n",
        "eval_test['score'] = eval_test.filter(regex='^T', axis = 1).mean(axis=1)\n",
        "eval_test['preference'] = calc_preference(eval_test['score'])\n",
        "\n",
        "eval_priv_result = {}\n",
        "for idx in range(len(templates)):\n",
        "  eval_priv_result[csv_filelist[idx]] = evaluate(eval_test, templates[idx])\n",
        "\n",
        "for key, value in eval_priv_result.items():\n",
        "  score, percent = getCompareDF(eval_test, value)\n",
        "  print(f'{key}: {score}, {percent}')\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DF2SgbZYG6Qx",
        "outputId": "68cf59c2-17f7-42b9-c4ec-1c6e5017ed11"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Correct_Rate_Inverse: 17, 0.68\n",
            "CR_inverse+SS: 17, 0.68\n",
            "QuestionTextLength: 15, 0.6\n",
            "QuestionTextLength_Inverse: 10, 0.4\n",
            "CR_inverse+QuestionTextLength: 15, 0.6\n",
            "CR_inverse+SS+QuestionTextLength: 15, 0.6\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Select best subset"
      ],
      "metadata": {
        "id": "Pc7LhfSAOJKY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import itertools\n",
        "\n",
        "# Calculate the combination of all columns\n",
        "combinations = []\n",
        "column_combinations = []\n",
        "num_columns = len(result_df.columns)\n",
        "\n",
        "for r in range(1, num_columns+1):\n",
        "    for cols in itertools.combinations(result_df.columns, r):\n",
        "        combination_sum = result_df[list(cols)].sum(axis=1)\n",
        "        combinations.append(combination_sum)\n",
        "        column_combinations.append(list(cols))  # Save column name combinations\n",
        "\n",
        "subset_df = pd.concat(combinations, axis=1)\n",
        "subset_df.columns = ['Combination_{}'.format(i) for i in range(len(combinations))]\n",
        "\n",
        "print(column_combinations[62])\n",
        "\n",
        "subset_df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 349
        },
        "id": "MdTwdpg2OJcW",
        "outputId": "11259d08-eeb5-4a2f-bf57-858b5dd2a4a4"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            Combination_0  Combination_1  Combination_2  Combination_3  \\\n",
              "QuestionId                                                               \n",
              "0                0.443457       0.556543              6       1.194456   \n",
              "1                0.571429       0.428571              6       0.876006   \n",
              "2                0.385214       0.614786              4       1.331663   \n",
              "3                0.808757       0.191243              6       0.695423   \n",
              "4                0.401408       0.598592              6       1.324321   \n",
              "\n",
              "            Combination_4  Combination_5  Combination_6  Combination_7  \\\n",
              "QuestionId                                                               \n",
              "0                0.686739            141            1.0       6.443457   \n",
              "1                0.682908             46            1.0       6.571429   \n",
              "2                0.666559            105            1.0       4.385214   \n",
              "3                0.488020             53            1.0       6.808757   \n",
              "4                0.673579            125            1.0       6.401408   \n",
              "\n",
              "            Combination_8  Combination_9  ...  Combination_53  Combination_54  \\\n",
              "QuestionId                                ...                                   \n",
              "0                1.637913       1.130196  ...      148.243283      143.437739   \n",
              "1                1.447434       1.254337  ...       53.111480       47.987485   \n",
              "2                1.716877       1.051773  ...      110.281345      107.613008   \n",
              "3                1.504180       1.296777  ...       59.679262       54.374685   \n",
              "4                1.725730       1.074987  ...      132.272170      127.596491   \n",
              "\n",
              "            Combination_55  Combination_56  Combination_57  Combination_58  \\\n",
              "QuestionId                                                                   \n",
              "0               148.881195        8.881195      149.194456      148.686739   \n",
              "1                53.558914        8.558914       53.876006       53.682908   \n",
              "2               110.998222        6.998222      111.331663      110.666559   \n",
              "3                60.183442        8.183442       60.695423       60.488020   \n",
              "4               132.997900        8.997900      133.324321      132.673579   \n",
              "\n",
              "            Combination_59  Combination_60  Combination_61  Combination_62  \n",
              "QuestionId                                                                  \n",
              "0               143.881195      149.324652      149.437739      149.881195  \n",
              "1                48.558914       54.130342       53.987485       54.558914  \n",
              "2               107.998222      111.383436      111.613008      111.998222  \n",
              "3                55.183442       60.992200       60.374685       61.183442  \n",
              "4               127.997900      133.399308      133.596491      133.997900  \n",
              "\n",
              "[5 rows x 63 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-cd6aca4d-ce7c-4d3a-bd5c-792679af4428\">\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>Combination_0</th>\n",
              "      <th>Combination_1</th>\n",
              "      <th>Combination_2</th>\n",
              "      <th>Combination_3</th>\n",
              "      <th>Combination_4</th>\n",
              "      <th>Combination_5</th>\n",
              "      <th>Combination_6</th>\n",
              "      <th>Combination_7</th>\n",
              "      <th>Combination_8</th>\n",
              "      <th>Combination_9</th>\n",
              "      <th>...</th>\n",
              "      <th>Combination_53</th>\n",
              "      <th>Combination_54</th>\n",
              "      <th>Combination_55</th>\n",
              "      <th>Combination_56</th>\n",
              "      <th>Combination_57</th>\n",
              "      <th>Combination_58</th>\n",
              "      <th>Combination_59</th>\n",
              "      <th>Combination_60</th>\n",
              "      <th>Combination_61</th>\n",
              "      <th>Combination_62</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>QuestionId</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.443457</td>\n",
              "      <td>0.556543</td>\n",
              "      <td>6</td>\n",
              "      <td>1.194456</td>\n",
              "      <td>0.686739</td>\n",
              "      <td>141</td>\n",
              "      <td>1.0</td>\n",
              "      <td>6.443457</td>\n",
              "      <td>1.637913</td>\n",
              "      <td>1.130196</td>\n",
              "      <td>...</td>\n",
              "      <td>148.243283</td>\n",
              "      <td>143.437739</td>\n",
              "      <td>148.881195</td>\n",
              "      <td>8.881195</td>\n",
              "      <td>149.194456</td>\n",
              "      <td>148.686739</td>\n",
              "      <td>143.881195</td>\n",
              "      <td>149.324652</td>\n",
              "      <td>149.437739</td>\n",
              "      <td>149.881195</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.571429</td>\n",
              "      <td>0.428571</td>\n",
              "      <td>6</td>\n",
              "      <td>0.876006</td>\n",
              "      <td>0.682908</td>\n",
              "      <td>46</td>\n",
              "      <td>1.0</td>\n",
              "      <td>6.571429</td>\n",
              "      <td>1.447434</td>\n",
              "      <td>1.254337</td>\n",
              "      <td>...</td>\n",
              "      <td>53.111480</td>\n",
              "      <td>47.987485</td>\n",
              "      <td>53.558914</td>\n",
              "      <td>8.558914</td>\n",
              "      <td>53.876006</td>\n",
              "      <td>53.682908</td>\n",
              "      <td>48.558914</td>\n",
              "      <td>54.130342</td>\n",
              "      <td>53.987485</td>\n",
              "      <td>54.558914</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.385214</td>\n",
              "      <td>0.614786</td>\n",
              "      <td>4</td>\n",
              "      <td>1.331663</td>\n",
              "      <td>0.666559</td>\n",
              "      <td>105</td>\n",
              "      <td>1.0</td>\n",
              "      <td>4.385214</td>\n",
              "      <td>1.716877</td>\n",
              "      <td>1.051773</td>\n",
              "      <td>...</td>\n",
              "      <td>110.281345</td>\n",
              "      <td>107.613008</td>\n",
              "      <td>110.998222</td>\n",
              "      <td>6.998222</td>\n",
              "      <td>111.331663</td>\n",
              "      <td>110.666559</td>\n",
              "      <td>107.998222</td>\n",
              "      <td>111.383436</td>\n",
              "      <td>111.613008</td>\n",
              "      <td>111.998222</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.808757</td>\n",
              "      <td>0.191243</td>\n",
              "      <td>6</td>\n",
              "      <td>0.695423</td>\n",
              "      <td>0.488020</td>\n",
              "      <td>53</td>\n",
              "      <td>1.0</td>\n",
              "      <td>6.808757</td>\n",
              "      <td>1.504180</td>\n",
              "      <td>1.296777</td>\n",
              "      <td>...</td>\n",
              "      <td>59.679262</td>\n",
              "      <td>54.374685</td>\n",
              "      <td>60.183442</td>\n",
              "      <td>8.183442</td>\n",
              "      <td>60.695423</td>\n",
              "      <td>60.488020</td>\n",
              "      <td>55.183442</td>\n",
              "      <td>60.992200</td>\n",
              "      <td>60.374685</td>\n",
              "      <td>61.183442</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.401408</td>\n",
              "      <td>0.598592</td>\n",
              "      <td>6</td>\n",
              "      <td>1.324321</td>\n",
              "      <td>0.673579</td>\n",
              "      <td>125</td>\n",
              "      <td>1.0</td>\n",
              "      <td>6.401408</td>\n",
              "      <td>1.725730</td>\n",
              "      <td>1.074987</td>\n",
              "      <td>...</td>\n",
              "      <td>132.272170</td>\n",
              "      <td>127.596491</td>\n",
              "      <td>132.997900</td>\n",
              "      <td>8.997900</td>\n",
              "      <td>133.324321</td>\n",
              "      <td>132.673579</td>\n",
              "      <td>127.997900</td>\n",
              "      <td>133.399308</td>\n",
              "      <td>133.596491</td>\n",
              "      <td>133.997900</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 63 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-cd6aca4d-ce7c-4d3a-bd5c-792679af4428')\"\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-cd6aca4d-ce7c-4d3a-bd5c-792679af4428 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-cd6aca4d-ce7c-4d3a-bd5c-792679af4428');\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",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# get ranking of all columns\n",
        "subset_rank_df = pd.DataFrame()\n",
        "subset_rank_df = subset_df.rank(method='first', ascending=False).astype('int16')\n",
        "\n",
        "subset_rank_df['QuestionId'] = subset_rank_df.index  # Add QuestionId Column\n",
        "subset_rank_df = subset_rank_df[['QuestionId'] + subset_rank_df.columns[:-1].tolist()]  # Move the QuestionId column to the front"
      ],
      "metadata": {
        "id": "d8vCN_c1SOxX"
      },
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "subset_templates = []\n",
        "for i in range(len(combinations)):\n",
        "  selected_columns = ['QuestionId', 'Combination_'+str(i)]\n",
        "  subset_templates.append(subset_rank_df[selected_columns])\n",
        "\n",
        "for template in subset_templates:\n",
        "  template.columns = ['QuestionId', 'ranking']\n",
        "\n",
        "print(subset_templates[0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eiJEOdhoX0_K",
        "outputId": "d6d41e9d-5ba2-44ec-ff79-1afa1361a24d"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "            QuestionId  ranking\n",
            "QuestionId                     \n",
            "0                    0      594\n",
            "1                    1      328\n",
            "2                    2      721\n",
            "3                    3       27\n",
            "4                    4      676\n",
            "...                ...      ...\n",
            "943                943      336\n",
            "944                944      948\n",
            "945                945      639\n",
            "946                946      554\n",
            "947                947      239\n",
            "\n",
            "[948 rows x 2 columns]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Eval Public"
      ],
      "metadata": {
        "id": "Eo8CeEuVaIIJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "eval_pub_subset_result = []\n",
        "for idx in range(len(subset_templates)):\n",
        "  eval_pub_subset_result.append(evaluate(eval_validation, subset_templates[idx]))\n"
      ],
      "metadata": {
        "id": "RqwcgOyuYcSj"
      },
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "for idx in range(len(subset_templates)):\n",
        "  score, percent = getCompareDF(eval_validation, eval_pub_subset_result[idx])\n",
        "  print(f'{column_combinations[idx]}: {score}, {percent}')\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GxXPV-DDYl0s",
        "outputId": "d7166e2b-2c95-4f70-dbd3-8302c493ed68"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['Correct_Rate']: 6, 0.24\n",
            "['Correct_Rate_Inverse']: 19, 0.76\n",
            "['Subject_Score']: 12, 0.48\n",
            "['AnswerValue_Entropy']: 21, 0.84\n",
            "['IsCorrect_Entropy']: 19, 0.76\n",
            "['QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Correct_Rate_Inverse']: 12, 0.48\n",
            "['Correct_Rate', 'Subject_Score']: 6, 0.24\n",
            "['Correct_Rate', 'AnswerValue_Entropy']: 19, 0.76\n",
            "['Correct_Rate', 'IsCorrect_Entropy']: 6, 0.24\n",
            "['Correct_Rate', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate_Inverse', 'Subject_Score']: 20, 0.8\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy']: 21, 0.84\n",
            "['Correct_Rate_Inverse', 'IsCorrect_Entropy']: 19, 0.76\n",
            "['Correct_Rate_Inverse', 'QuestionTextLength']: 14, 0.56\n",
            "['Subject_Score', 'AnswerValue_Entropy']: 22, 0.88\n",
            "['Subject_Score', 'IsCorrect_Entropy']: 19, 0.76\n",
            "['Subject_Score', 'QuestionTextLength']: 14, 0.56\n",
            "['AnswerValue_Entropy', 'IsCorrect_Entropy']: 21, 0.84\n",
            "['AnswerValue_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score']: 12, 0.48\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'AnswerValue_Entropy']: 21, 0.84\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'IsCorrect_Entropy']: 19, 0.76\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Subject_Score', 'AnswerValue_Entropy']: 20, 0.8\n",
            "['Correct_Rate', 'Subject_Score', 'IsCorrect_Entropy']: 6, 0.24\n",
            "['Correct_Rate', 'Subject_Score', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 20, 0.8\n",
            "['Correct_Rate', 'AnswerValue_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy']: 22, 0.88\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'IsCorrect_Entropy']: 20, 0.8\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 21, 0.84\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate_Inverse', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 22, 0.88\n",
            "['Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Subject_Score', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy']: 22, 0.88\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'IsCorrect_Entropy']: 19, 0.76\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 21, 0.84\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'AnswerValue_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 20, 0.8\n",
            "['Correct_Rate', 'Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Subject_Score', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 22, 0.88\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 22, 0.88\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 14, 0.56\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Eval private"
      ],
      "metadata": {
        "id": "7mKddtIoaKWV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "eval_priv_subset_result = []\n",
        "for idx in range(len(subset_templates)):\n",
        "  eval_priv_subset_result.append(evaluate(eval_test, subset_templates[idx]))\n",
        "\n",
        "for idx in range(len(subset_templates)):\n",
        "  score, percent = getCompareDF(eval_test, eval_priv_subset_result[idx])\n",
        "  print(f'{column_combinations[idx]}: {score}, {percent}')\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7kaHEeVwaL5i",
        "outputId": "38e48c2a-4214-41bd-9818-29cc92ee2ac6"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['Correct_Rate']: 8, 0.32\n",
            "['Correct_Rate_Inverse']: 17, 0.68\n",
            "['Subject_Score']: 11, 0.44\n",
            "['AnswerValue_Entropy']: 16, 0.64\n",
            "['IsCorrect_Entropy']: 12, 0.48\n",
            "['QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse']: 11, 0.44\n",
            "['Correct_Rate', 'Subject_Score']: 8, 0.32\n",
            "['Correct_Rate', 'AnswerValue_Entropy']: 14, 0.56\n",
            "['Correct_Rate', 'IsCorrect_Entropy']: 8, 0.32\n",
            "['Correct_Rate', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate_Inverse', 'Subject_Score']: 17, 0.68\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy']: 16, 0.64\n",
            "['Correct_Rate_Inverse', 'IsCorrect_Entropy']: 17, 0.68\n",
            "['Correct_Rate_Inverse', 'QuestionTextLength']: 15, 0.6\n",
            "['Subject_Score', 'AnswerValue_Entropy']: 16, 0.64\n",
            "['Subject_Score', 'IsCorrect_Entropy']: 12, 0.48\n",
            "['Subject_Score', 'QuestionTextLength']: 15, 0.6\n",
            "['AnswerValue_Entropy', 'IsCorrect_Entropy']: 15, 0.6\n",
            "['AnswerValue_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score']: 11, 0.44\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'AnswerValue_Entropy']: 16, 0.64\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'IsCorrect_Entropy']: 12, 0.48\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Subject_Score', 'AnswerValue_Entropy']: 14, 0.56\n",
            "['Correct_Rate', 'Subject_Score', 'IsCorrect_Entropy']: 8, 0.32\n",
            "['Correct_Rate', 'Subject_Score', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 13, 0.52\n",
            "['Correct_Rate', 'AnswerValue_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy']: 16, 0.64\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'IsCorrect_Entropy']: 17, 0.68\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 16, 0.64\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate_Inverse', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 15, 0.6\n",
            "['Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Subject_Score', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy']: 16, 0.64\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'IsCorrect_Entropy']: 12, 0.48\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'AnswerValue_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 13, 0.52\n",
            "['Correct_Rate', 'Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Subject_Score', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 16, 0.64\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: 15, 0.6\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Scaling"
      ],
      "metadata": {
        "id": "ekKTfF0dbPYx"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "result_df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "id": "uP2TKzKXbSJ1",
        "outputId": "43d7d388-02a9-458b-9aa7-81b5d288e29f"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            Correct_Rate  Correct_Rate_Inverse  Subject_Score  \\\n",
              "QuestionId                                                      \n",
              "0               0.443457              0.556543              6   \n",
              "1               0.571429              0.428571              6   \n",
              "2               0.385214              0.614786              4   \n",
              "3               0.808757              0.191243              6   \n",
              "4               0.401408              0.598592              6   \n",
              "\n",
              "            AnswerValue_Entropy  IsCorrect_Entropy  QuestionTextLength  \n",
              "QuestionId                                                              \n",
              "0                      1.194456           0.686739                 141  \n",
              "1                      0.876006           0.682908                  46  \n",
              "2                      1.331663           0.666559                 105  \n",
              "3                      0.695423           0.488020                  53  \n",
              "4                      1.324321           0.673579                 125  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-7d8b2946-77ad-41e0-b72c-c6ff3c96b68e\">\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>Correct_Rate</th>\n",
              "      <th>Correct_Rate_Inverse</th>\n",
              "      <th>Subject_Score</th>\n",
              "      <th>AnswerValue_Entropy</th>\n",
              "      <th>IsCorrect_Entropy</th>\n",
              "      <th>QuestionTextLength</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>QuestionId</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.443457</td>\n",
              "      <td>0.556543</td>\n",
              "      <td>6</td>\n",
              "      <td>1.194456</td>\n",
              "      <td>0.686739</td>\n",
              "      <td>141</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.571429</td>\n",
              "      <td>0.428571</td>\n",
              "      <td>6</td>\n",
              "      <td>0.876006</td>\n",
              "      <td>0.682908</td>\n",
              "      <td>46</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.385214</td>\n",
              "      <td>0.614786</td>\n",
              "      <td>4</td>\n",
              "      <td>1.331663</td>\n",
              "      <td>0.666559</td>\n",
              "      <td>105</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.808757</td>\n",
              "      <td>0.191243</td>\n",
              "      <td>6</td>\n",
              "      <td>0.695423</td>\n",
              "      <td>0.488020</td>\n",
              "      <td>53</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.401408</td>\n",
              "      <td>0.598592</td>\n",
              "      <td>6</td>\n",
              "      <td>1.324321</td>\n",
              "      <td>0.673579</td>\n",
              "      <td>125</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-7d8b2946-77ad-41e0-b72c-c6ff3c96b68e')\"\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-7d8b2946-77ad-41e0-b72c-c6ff3c96b68e 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-7d8b2946-77ad-41e0-b72c-c6ff3c96b68e');\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",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "result_df.describe()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "Y0Cii5x_cynH",
        "outputId": "3e8299ca-3ee3-484d-cc47-ecfb4598df2a"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       Correct_Rate  Correct_Rate_Inverse  Subject_Score  AnswerValue_Entropy  \\\n",
              "count    948.000000            948.000000     948.000000           948.000000   \n",
              "mean       0.507043              0.492957       5.983122             1.139295   \n",
              "std        0.156731              0.156731       0.144340             0.204057   \n",
              "min        0.142857              0.000000       4.000000             0.000000   \n",
              "25%        0.386833              0.382436       6.000000             1.037220   \n",
              "50%        0.495265              0.504735       6.000000             1.198014   \n",
              "75%        0.617564              0.613167       6.000000             1.297631   \n",
              "max        1.000000              0.857143       6.000000             1.386294   \n",
              "\n",
              "       IsCorrect_Entropy  QuestionTextLength  \n",
              "count         948.000000          948.000000  \n",
              "mean            0.641264          124.777426  \n",
              "std             0.070357           70.530214  \n",
              "min             0.000000           16.000000  \n",
              "25%             0.618163           71.750000  \n",
              "50%             0.666980          109.000000  \n",
              "75%             0.687197          164.250000  \n",
              "max             0.693147          415.000000  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1e5b1a22-f9dc-4b42-a3cd-a240e6a0a2fd\">\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>Correct_Rate</th>\n",
              "      <th>Correct_Rate_Inverse</th>\n",
              "      <th>Subject_Score</th>\n",
              "      <th>AnswerValue_Entropy</th>\n",
              "      <th>IsCorrect_Entropy</th>\n",
              "      <th>QuestionTextLength</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>948.000000</td>\n",
              "      <td>948.000000</td>\n",
              "      <td>948.000000</td>\n",
              "      <td>948.000000</td>\n",
              "      <td>948.000000</td>\n",
              "      <td>948.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.507043</td>\n",
              "      <td>0.492957</td>\n",
              "      <td>5.983122</td>\n",
              "      <td>1.139295</td>\n",
              "      <td>0.641264</td>\n",
              "      <td>124.777426</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.156731</td>\n",
              "      <td>0.156731</td>\n",
              "      <td>0.144340</td>\n",
              "      <td>0.204057</td>\n",
              "      <td>0.070357</td>\n",
              "      <td>70.530214</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.142857</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>16.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.386833</td>\n",
              "      <td>0.382436</td>\n",
              "      <td>6.000000</td>\n",
              "      <td>1.037220</td>\n",
              "      <td>0.618163</td>\n",
              "      <td>71.750000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.495265</td>\n",
              "      <td>0.504735</td>\n",
              "      <td>6.000000</td>\n",
              "      <td>1.198014</td>\n",
              "      <td>0.666980</td>\n",
              "      <td>109.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.617564</td>\n",
              "      <td>0.613167</td>\n",
              "      <td>6.000000</td>\n",
              "      <td>1.297631</td>\n",
              "      <td>0.687197</td>\n",
              "      <td>164.250000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.857143</td>\n",
              "      <td>6.000000</td>\n",
              "      <td>1.386294</td>\n",
              "      <td>0.693147</td>\n",
              "      <td>415.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1e5b1a22-f9dc-4b42-a3cd-a240e6a0a2fd')\"\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-1e5b1a22-f9dc-4b42-a3cd-a240e6a0a2fd 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-1e5b1a22-f9dc-4b42-a3cd-a240e6a0a2fd');\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",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "result_df.plot.box()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 448
        },
        "id": "JZb9gBVjcxEJ",
        "outputId": "fda3a9f7-328f-4849-f5b7-dadf75b28639"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: >"
            ]
          },
          "metadata": {},
          "execution_count": 25
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "std_scaler = StandardScaler()\n",
        "std_scaled = std_scaler.fit_transform(result_df)\n",
        "scaled_df = pd.DataFrame(std_scaled, columns=['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score',\t'AnswerValue_Entropy',\t'IsCorrect_Entropy',\t'QuestionTextLength'])\n",
        "scaled_df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "wKUtu8P9c5r9",
        "outputId": "f1e970fd-6928-4f70-f69c-5f11b1ddaf80"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   Correct_Rate  Correct_Rate_Inverse  Subject_Score  AnswerValue_Entropy  \\\n",
              "0     -0.405919              0.405919       0.116991             0.270462   \n",
              "1      0.411018             -0.411018       0.116991            -1.290954   \n",
              "2     -0.777723              0.777723     -13.746451             0.943211   \n",
              "3      1.926059             -1.926059       0.116991            -2.176383   \n",
              "4     -0.674342              0.674342       0.116991             0.907212   \n",
              "\n",
              "   IsCorrect_Entropy  QuestionTextLength  \n",
              "0           0.646696            0.230130  \n",
              "1           0.592216           -1.117521  \n",
              "2           0.359719           -0.280559  \n",
              "3          -2.179250           -1.018221  \n",
              "4           0.459543            0.003157  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9590a377-1491-4fc8-8d0c-86bb482425c8\">\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>Correct_Rate</th>\n",
              "      <th>Correct_Rate_Inverse</th>\n",
              "      <th>Subject_Score</th>\n",
              "      <th>AnswerValue_Entropy</th>\n",
              "      <th>IsCorrect_Entropy</th>\n",
              "      <th>QuestionTextLength</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>-0.405919</td>\n",
              "      <td>0.405919</td>\n",
              "      <td>0.116991</td>\n",
              "      <td>0.270462</td>\n",
              "      <td>0.646696</td>\n",
              "      <td>0.230130</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.411018</td>\n",
              "      <td>-0.411018</td>\n",
              "      <td>0.116991</td>\n",
              "      <td>-1.290954</td>\n",
              "      <td>0.592216</td>\n",
              "      <td>-1.117521</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>-0.777723</td>\n",
              "      <td>0.777723</td>\n",
              "      <td>-13.746451</td>\n",
              "      <td>0.943211</td>\n",
              "      <td>0.359719</td>\n",
              "      <td>-0.280559</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1.926059</td>\n",
              "      <td>-1.926059</td>\n",
              "      <td>0.116991</td>\n",
              "      <td>-2.176383</td>\n",
              "      <td>-2.179250</td>\n",
              "      <td>-1.018221</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>-0.674342</td>\n",
              "      <td>0.674342</td>\n",
              "      <td>0.116991</td>\n",
              "      <td>0.907212</td>\n",
              "      <td>0.459543</td>\n",
              "      <td>0.003157</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9590a377-1491-4fc8-8d0c-86bb482425c8')\"\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-9590a377-1491-4fc8-8d0c-86bb482425c8 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-9590a377-1491-4fc8-8d0c-86bb482425c8');\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",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "scaled_df.plot.box()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 453
        },
        "id": "9uu00h_7d5gA",
        "outputId": "0914bed0-20eb-430f-8fb6-e8c87e1af9fb"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: >"
            ]
          },
          "metadata": {},
          "execution_count": 27
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlMAAAGiCAYAAADDZ0PJAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABaBElEQVR4nO3deVxU5eIG8GdmYIZhFwQRQVDAXXLLBS1RUzRNyOWWS26k1g1NMVTMSu0aJZqWS13LpVIz8xre3MpU0iuUZYI7ArkLLqhssvP+/vA3J0YYlmaGGeT5fj58dM5558x7zpyZ88x73vMemRBCgIiIiIj+FrmpK0BERERUlzFMEREREemBYYqIiIhIDwxTRERERHpgmCIiIiLSA8MUERERkR4YpoiIiIj0wDBFREREpAeGKSIiIiI9MEwRERER6cHkYWrBggWQyWRaf61atar0Od9++y1atWoFKysrtG/fHnv27Kml2hIRERFpM3mYAoC2bdsiLS1N+vvf//6ns2xcXBxGjRqF0NBQnDhxAiEhIQgJCcHp06drscZERERED8lMfaPjBQsWICYmBgkJCdUq/8ILLyA3Nxe7du2SpnXv3h0dOnTAp59+aqRaEhEREVXMwtQVAIDk5GS4u7vDysoKPXr0QFRUFJo2bVph2fj4eISHh2tNCwoKQkxMjM7lFxQUoKCgQHpcWlqKu3fvwtnZGTKZzCDrQERERMYlhEB2djbc3d0hl5vFyTUAZhCmunXrho0bN6Jly5ZIS0vDwoUL8dRTT+H06dOws7MrVz49PR2NGjXSmtaoUSOkp6frfI2oqCgsXLjQ4HUnIiKi2nf16lV4eHiYuhoSk4epQYMGSf/39/dHt27d4OXlhW3btiE0NNQgrxEZGanVmpWZmYmmTZvi6tWrsLe3N8hrEBERkXFlZWXB09OzwsYWUzJ5mHqUo6MjWrRogZSUlArnu7m54ebNm1rTbt68CTc3N53LVKlUUKlU5abb29szTBEREdUx5tZFx3xOOP6/nJwcpKamonHjxhXO79GjBw4cOKA1bf/+/ejRo0dtVI+IiIhIi8nD1BtvvIGff/4Zly5dQlxcHJ5//nkoFAqMGjUKADBu3DhERkZK5V9//XXs27cPy5Ytw/nz57FgwQL8/vvvCAsLM9UqEBERUT1m8tN8165dw6hRo5CRkQEXFxf06tULv/zyC1xcXAAAV65c0eqxHxAQgC1btmD+/PmYN28e/Pz8EBMTg3bt2plqFYiIiKgeM/k4U6aQlZUFBwcHZGZmss8UERFRHWGux2+Tn+YjIiIiqssYpoiIiIj0wDBFREREpAeGKSIiIiI9mPxqPiIiIjKtwsJCrFmzBqmpqfDx8cE///lPKJVKU1erzmCYIiIiqsdmz56N5cuXo7i4WJoWERGBmTNnYsmSJSasWd3BMEVERFRPzZ49G9HR0WjUqBFeeuklNG/eHH/++Se++uorREdHAwADVTVwnCkzGqeCiIiothQWFsLGxgY2NjZwdHTE5cuXpXleXl64f/8+cnNzkZubazan/Mz1+M0O6ERERPXQmjVrUFxcjMzMTPj7+yM+Ph7Z2dmIj4+Hv78/MjMzUVxcjDVr1pi6qmaPYYqIiKgeSk5OBgD0798fMTEx6N69O2xtbdG9e3fExMSgf//+WuVIN4YpIiKiekgmkwEAOnfurHUPXACQy+Xo2LGjVjnSjWGKiIioHurWrRsAYP369VpX8gFAcXExNm7cqFWOdGOYIiIiqoc8PT0BALdu3YKHhwfWrl2LGzduYO3atfDw8MCtW7e0ypFuvJrPjK4GICIiqi0lJSXw9fWFQqHApUuXUFJSIs2zsLCAl5cXSktLkZycDIVCYcKa/sVcj98cZ4qIiKgeUigUWLZsGUaMGIFnn30WPj4+yM/Ph5WVFVJTU7Fnzx5s377dbIKUOWOYIiIiqqeGDRuG7du3Y9asWdi9e7c0vVmzZti+fTuGDRtmwtrVHTzNZ0bNhERERKZQV+7NZ67Hb7ZMERER1WM7duxAeHi41gjoK1aswIcffsiWqWri1XxERET11I4dOzB8+HDcvHlTa/rNmzcxfPhw7Nixw0Q1q1sYpoiIiOqhkpISvPLKKwDKD8ypefzqq69qXeVHFeNpPiIionooNjYWt2/fBgD07dsXzz77LNRqNfLy8rBnzx7s3r0bt27dQmxsLPr162fi2po3hikiIqJ66ODBgwCAFi1a4MyZM1pX83l7e6NFixa4cOECDh48yDBVBYYpIiKieujKlSsAHt7IePDgwYiIiJBapvbu3SuFK0050o1hygzl5eUhIiICycnJ8PPzQ3R0NNRqtamrRUREjxEPDw8AgLW1NU6ePIldu3ZJ85o2bQpra2vk5uZK5Ug3dkA3MyEhIbC2tsbq1avx448/YvXq1bC2tkZISIipq0ZERI+Rhg0bAgByc3ORn5+vdW++/Px85ObmapUj3RimzEhISAh27twJpVKJuXPnIiUlBXPnzoVSqcTOnTsZqIiIyGBcXV2l/2dlZWHKlClwd3fHlClTkJ2dXWE5qhhP85mJvLw8KUhlZ2dLI89GRUVh4cKFsLOzw86dO5GXl8dTfkREpLeMjAzp/48OjaCrHFWMLVNmIiIiAgAQHh5ebgh/pVKJGTNmaJUjIiLSh4uLC4CH9+ErKCjQmldYWIhmzZpplSPd2DJlJpKTkwEAL7/8coXzQ0NDsWTJEqkcERGRPpo0aQIAuHjxIlxdXREYGAhra2s8ePAAsbGxuHjxolY50o1hykz4+fnhxx9/xOeff46oqKhy89etWyeVIyIi0ldAQAAsLCygVCpx9+5dbNu2TZpnYWEBa2trFBYWIiAgwIS1rBtkQghh6krUNnO863ReXh6sra3L9ZkCHja32tnZobCwEA8ePGCfKSIi0ltsbCz69OkD4OGpvD59+sDGxga5ubk4dOiQNDr6oUOHEBgYaMKa/sUcj98A+0yZDbVajeDgYCk4zZkzBxcuXMCcOXOkIBUcHMwgRUREBnH9+nUAD/tMaVqmNmzYgG3btuHu3btSnylNOdKNYcqMxMTESIFqyZIlaNmyJZYsWSIFqZiYGFNXkYiIHhOalqeLFy9WeOGTps+Uphzpxj5TZiYmJoYjoBMRkdE5OztL/+/duzdsbGxw7949NGjQALm5udi3b1+5clQxhikzpFarsWrVKlNXg4iIHmO3bt2S/q8JTlWVo4rxNB8REVE9dPfuXYOWq89M3jIVFRWFHTt24Pz581Cr1QgICMAHH3yAli1b6nzOxo0bMXHiRK1pKpUK+fn5xq6u3vIKS5B6O6fKcvlFJbh2Lw8eDdSwslRUa9k+LrZQK6tXloiI6rfi4mLp/4MGDcLgwYOhVquRl5eH3bt3Y+/eveXKUcVMHqZ+/vlnvPbaa3jyySdRXFyMefPmYcCAATh79ixsbGx0Ps/e3h5JSUnS48qGwjcnqbdzMGTl/4yy7F3TeqFdEwejLJuIiB4v58+fBwA4OTnh7NmzUngCAG9vbzg5OeHu3btSOdLN5GHq0fO0GzduhKurK44fP46nn35a5/NkMhnc3NyMXT2D83Gxxa5pvaosl3IrBzO+ScCKFzrA19W22ssmIiLSqOxsyK17D29mfPfuXbTr+CRemPRPPChVwFpegl8OH8ThAz9I5U5fzyz3fJ4N+YvJw9SjMjMfvmFOTk6VlsvJyYGXlxdKS0vRqVMnvPfee2jbtm2FZQsKCrTuO5SVlWW4CteQWqmoUeuRr6stW5uIiOhvqexsSEbOX1eJH/n5kBSeAEBmoZL+fzpHXeEyeDbkL2YVpkpLSzFjxgz07NkT7dq101muZcuWWL9+Pfz9/ZGZmYmlS5ciICAAZ86cgYeHR7nyUVFRWLhwoTGrTkREZHYqOxuS93JndPVrDIXCAg1dGuJm2g1pnqtrQ9y5eRMlJcU4uG1dhcPz8GzIX8zqdjKvvvoq9u7di//9738VhiJdioqK0Lp1a4waNQrvvvtuufkVtUx5enqa3XD0ZZ2+nokhK//H5E9EREYTEhKCnTt3QqlUot+g5xCf5Yge9vdxYO/3ZjlgtLneTsZsWqbCwsKwa9cuHD58uEZBCgAsLS3RsWNHpKSkVDhfpVJBpVJVOI/qLmNdGcl+AERUX8TExEiBau/O/wAANN3QzS1ImTOThykhBKZNm4bvvvsOsbGx0r2AaqKkpASnTp3Cs88+a4Qakrky1pWRbA0kovpEc+eNSa9OR8zPxxHSuzPWf/Ix77xRAyYPU6+99hq2bNmCnTt3ws7ODunp6QAABwcH6Y0cN24cmjRpgqioKADAokWL0L17d/j6+uL+/fuIjo7G5cuX8fLLL5tsPaj2GevKSPYDIKL6Rq1W483FSxG/8n94c1ovBqkaMnmY+uSTTwAAgYGBWtM3bNiACRMmAACuXLkCufyvwdrv3buHyZMnIz09HQ0aNEDnzp0RFxeHNm3a1Fa1yQzwykgiIjIHJg9T1en/Hhsbq/V4+fLlWL58uZFqRERERFR9vDcfERERkR4YpoiIiIj0wDBFREREpAeGKSIiIiI9MEwRERER6YFhioiIiEgPDFNEREREemCYIiIiItIDwxQRERGRHhimiIiIiPTAMEVERESkB4YpIiIiIj0wTBERERHpgWGKiIiISA8MU0RERER6YJgiIiIi0oOFqStARPS4yCssQertnCrL5ReV4Nq9PHg0UMPKUlGtZfu42EKtrF5ZIqpdDFNERAaSejsHQ1b+zyjL3jWtF9o1cTDKsolIPwxTREQG4uNii13TelVZLuVWDmZ8k4AVL3SAr6tttZdNROaJYYqIyEDUSkWNWo98XW3Z2kT0GGAHdCIiIiI9MEwRERER6YFhioiIiEgPDFNEREREemAHdDI7F+/kIreg2GDLS7mVo/WvodioLNCsoY1Bl0nmy5D7JfdJoscLwxSZlYt3ctFnaaxRlj3jmwSDL/PQG4E8eNUDxtovuU/qZqwBUDn4KRkDw5QB8Zer/jTbrybj71Tl74w2XRXNOEGGbEEj82Xo/bI+75PV/Z7UrI+hVfc9NPfvSjIvDFMGwl+uhmXo8Xe6eBtsUVSPGXK/rI/7pDFbnqurJt+pdeG7kswDw5SB8JcrEVHl2PJsWHWhf2l9aeFjmDIw/nIlIqocW571V5f6l9aHFj6GKSIiojqmLrTy1ZUWPkNgmCIiIqqj2MpnHhimiIioVhSU5ENudR0Xs5IgtzJMa4oxXMzKgdzqOgpK8gHwRtRUNYYpIqIq1IUQUBcCwI3cy7BpthLzjpm6JlWzaQbcyO2Azmhk6qpQHcAwRURUhboSAsw9ALjbeCH34jR89EIH+Bion48xpN7KwevfJMC9j5epq0J1BMMUEVEV6kIIqAsBQKWwQml+EzSzb4k2zubZegYApfmZKM2/DZXCytRVoTrCLMLU6tWrER0djfT0dDzxxBNYuXIlunbtqrP8t99+i7feeguXLl2Cn58fPvjgAzz77LO1WGMiqk/qQghgACAyHZOHqW+++Qbh4eH49NNP0a1bN6xYsQJBQUFISkqCq6trufJxcXEYNWoUoqKiMGTIEGzZsgUhISH4448/0K5dOxOsARlSXeibAtSR/imZmfgm4XiV5fIKS3Dlbq5R6tDUyabK+6C5OVghpF1HqC3URqkDEZGxmTxMffjhh5g8eTImTpwIAPj000+xe/durF+/HnPnzi1X/qOPPsLAgQMREREBAHj33Xexf/9+rFq1Cp9++mmt1p0Mr670TQHMv3/KNwnHsf7S6yatw9Eb1SvnZLMRQX6djVsZIiIjMWmYKiwsxPHjxxEZGSlNk8vleOaZZxAfH1/hc+Lj4xEeHq41LSgoCDExMTpfp6CgAAUFBdLjrKws/SpORlMX+qYAdaN/ygsdOgP4qMpy5tAy9XSzNkZ5fSKi2mDSMHXnzh2UlJSgUSPtX/aNGjXC+fPnK3xOenp6heXT09N1vk5UVBQWLlyof4UrURdOT9WFU1N1oW8KUDf6p7g7OGBm776mrgYR0WPP5Kf5akNkZKRWa1ZWVhY8PT0N+hp15fSUuZ+aIiIiqmtMGqYaNmwIhUKBmzdvak2/efMm3NzcKnyOm5tbjcoDgEqlgkql0r/ClagLp6fqwqkpIiKiusakYUqpVKJz5844cOAAQkJCAAClpaU4cOAAwsLCKnxOjx49cODAAcyYMUOatn//fvTo0aMWaqxbXTg9VRdOTRERUdXYtcS8mPw0X3h4OMaPH48uXbqga9euWLFiBXJzc6Wr+8aNG4cmTZogKioKAPD666+jd+/eWLZsGQYPHoytW7fi999/x9q1a025GkRERLWGXUvMi8nD1AsvvIDbt2/j7bffRnp6Ojp06IB9+/ZJncyvXLkCuVwulQ8ICMCWLVswf/58zJs3D35+foiJieEYU0REZi6vqAQAcPp6psGWmV9Ugmv38uDRQA0ry8qvHK2ulFs5BlmOMbFriXkxeZgCgLCwMJ2n9WJjY8tNGzlyJEaOHGnkWhERkSGl/n9ImbvjlIlrUj02KrM4RFaIXUvMi/nuKURE9FgZ0PbhhUI+rrZQG7AVacY3CVjxQgf4GrCFxkZlgWYNbQy2PHq8MUwREVXB0Ken6uupKScbJV7s2tQoy/Z1tUW7JubZQkOPP4YpIqIq1KXTU+Z8aoroccVPHRFRFQx9eoqnpogeLwxTRERVMNbpKZ6aIno8MEyRWeGl00REVNcwTBkIO6gaRl3qmwKwfwoRmUZd+OFZF445hsIjgYHUpRBgzgGAl04TEVWNxxzz8vivYS1hB1XD4KXTRERVqys/PM39mGMoDFMGwg6qRERUW/jD07zIqy5CRERERLowTBERERHpgWGKiIiISA8MU0RERER6YJgiIiIi0gPDFBEREZEeGKaIiIiI9MAwRURERKQHhikiIiIiPTBMEREREemBYYqIiIhID7w3HxER0WMqr7AEqbdzqlU25VaO1r9V8XGxhVppmJss13UMU0REVCeVlJTgt7gjyD17GL/FCbQeNggKBQ/uZaXezsGQlf+r0XNmfJNQrXK7pvXiDZH/H8MUERHVOTt27MCsWbNw6dIlAMCk76OxyNsby5Ytw7Bhw0xbOTPi42KLXdN6VatsflEJrt3Lg0cDNawsqw6lPi62+lbvscEwRUREdcqOHTswYsQIqFQqrenp6ekYMWIEtm/fzkD1/9RKRY1aj7p4G68ujzOGKSIiMju6+vqUlJRg8pSpEEIgP79Aa97DxwKTp74Cny6BFZ7yYz8fMgaGKSIiMju6+vrkXUrA3Yw7AAC5tT1Unu0hs7SCKMpHwdVTKH2Qibt3bmPA7E+g9u5Q7vns50PGwDBFRERmR1dfnxVRP2EdAKVSiaK8LOQl/RW4ZDIZlEolCgsL8VzDO5hRwfPZz4eMgWGKiIjMjq6+PpeTTgMACgsLy80TQkjTLyedZgsU1RoO2klERHWGlZWVQcsRGQLDFBER1RmlpaUGLUdkCAxTRERUZ5w+fdqg5YgMgWGKiIjqjKysLIOWIzIEhikiIqozeJqPzBHDFBFRLcrJycH00NG4se41TA8djZyc6t1Ulh6Sy6t32KpuOSJD4NAIRES1pGvXrvjtt9+kx4d+uAw7Ozs8+eSTOHbsmAlrVndYWloatByRIZgsTF26dAnvvvsuDh48iPT0dLi7u2Ps2LF48803oVQqdT4vMDAQP//8s9a0qVOn4tNPPzV2lQ1C1y0SHpVyK0fr3+rgbRKITKuyz/eowX1wOvFEhfN+++03tO/QCV/vPqRz2fx8P9S4cWPcvn27WuWIaovJwtT58+dRWlqKf//73/D19cXp06cxefJk5ObmYunSpZU+d/LkyVi0aJH02Nra2tjVNRhdt0jQZcY3CdUuy9skEJmWrs93cX4OrusIUhqnE09gYPQ+WFhVPEI3P98PVff7vi4dF6juM1mYGjhwIAYOHCg9bt68OZKSkvDJJ59UGaasra3h5uZm7Coaha5bJDwqv6gE1+7lwaOBGlaW1fs1ytskEJmWrs932IQXcP3//6+yskJBfr40r+xj3zMbsGrjNzqXTUBBQUHVhWpQjsgQzKrPVGZmJpycnKost3nzZmzatAlubm547rnn8NZbb1X6K6SgoEDrg2XKS2Z13SKhIl28jVsXIjIsXZ/vpNOJ0v/lMpnWvLKPk04nsvWpChwBncyR2VzukJKSgpUrV2Lq1KmVlhs9ejQ2bdqEQ4cOITIyEl999RXGjh1b6XOioqLg4OAg/Xl6ehqy6kRElSp7mX6/fv0QHx+P7OxsxMfHo1+/fhWWo4p5e3sbtByRIRi8ZWru3Ln44IMPKi1z7tw5tGrVSnp8/fp1DBw4ECNHjsTkyZMrfe6UKVOk/7dv3x6NGzdGv379kJqaCh8fnwqfExkZifDwcOlxVlYWAxUR1ZpWrVrh1q1bAIBPPvkEgwYNwo0bN+Du7o69e/di165dUjmqXLt27QxajsgQDB6mZs2ahQkTJlRapnnz5tL/b9y4gT59+iAgIABr166t8et169YNwMOWLV1hSqVSQaVS1XjZZN6MdWUkr5oiQ2vSpIn0/7I/5O7evav1uGw5qtjBgwerXW7evHlGrg3RQwYPUy4uLnBxcalW2evXr6NPnz7o3LkzNmzY8LcGWUtISADAy2DrI2NdGcmrpsjQmjVrZtBy9dnVq1cNWo7IEEzWAf369esIDAyEl5cXli5dqjVuiOZKvevXr6Nfv3748ssv0bVrV6SmpmLLli149tln4ezsjJMnT2LmzJl4+umn4e/vb6pVIRMx1pWRvGqKDK1z584GLVef2dr+9fmUy+Va/czKPi5bjsjYTBam9u/fj5SUFKSkpMDDw0NrnhACAFBUVISkpCQ8ePAAAKBUKvHTTz9hxYoVyM3NhaenJ4YPH4758+fXev3J9HhlJNUVb7/9drXLDRs2zMi1qdv69u2LP/74A8DDY0J+mWEmyj7u27evSepH9ZPJwtSECROq7Fvl7e0tBSvgYV+DR0c/JyIyd1euXJH+/2hrikKhQElJSblyVLHc3Fzp/2WD1KOPy5YjMjazGRqByBjy8vIQFhaGoKAghIWFIS8vz9RVonrI3t4ewMMBhx9tiffw8JDGydOUI6K6hWGKHlshISGwtrbG6tWr8eOPP2L16tWwtrZGSEiIqatG9cyrr74KAHjw4AFOnDiB5cuXIywsDMuXL8cff/whdWXQlCPdyp6tkD0yAGrZx2XLERkbwxQ9lkJCQrBz504olUrMnTsXKSkpmDt3LpRKJXbu3MlARbWq7ACSzs7OmDlzJlatWoWZM2fC2dm5wnJUMUdHR+n/SqVSa17Zx2XLERmbWd1OhsgQ8vLypCCVnZ0tfcFGRUVh4cKFsLOzw86dO5GXlwe1Wm3i2lJ9UN3xozjOVNXKDqFjb28Pd3d3FBQUQKVS4caNG9KV4X9nqB2iv4t7Gz12IiIiAADh4eEV/nKdMWOGVjkiYwsICICFhQUcHBzKjYnn7u4OBwcHWFhYICAgwEQ1rDs0LXlKpRK3b99GYmIizp8/j8TERNy+fVv6zJdt8SMyNoYpeuwkJycDAF5++eUK54eGhmqVIzK2uLg4FBcXIzMzE+np6Vrz0tLSkJmZieLiYsTFxZmohnVHo0aNAACFhYUVztdM15Qjqg0MU/TY8fPzAwB8/vnnFc5ft26dVjkiY0tLS5P+/2in6bKno8qWo4qVbXGqrAM6W6aoNjFM0WMnOjoaAPDhhx+W+/VaWFiIFStWaJUjMjbNgd3JyQnZ2dlaV/NlZWWhQYMGWuVIt5iYGAAP77n66A3rmzZtKp3m05Qjqg3sgE6PHbVajeDgYOzcuRN2dnaYMWMGQkNDsW7dOqxYsQKFhYUIDg5m53OqNadOnQLwsMN0mzZtcPnyZWneihUrYG9vj3v37uHUqVMYMGCAqapZJxw/fhwAUFBQAH9/f8yePRtqtRp5eXnYt2+ftG015YhqA1um6LEUExOD4OBgFBYWYsmSJWjZsiWWLFkiBSn+aqXadOnSJenf/Px8rF27Fjdu3MDatWuRn58vBQBNOdJNM+TBc889h9OnTyMsLAyhoaEICwvDmTNnMGTIEK1yRLWBLVP02IqJiUFeXh4iIiKQnJwMPz8/REdHs0WKap1m/CgvLy/IZDJMmTJFmtesWTN4eXnh8uXLHGeqGsLDw/HTTz9h3759yMzMxK+//oq0tDQ0btwY3bp1g4ODg1SOqLYwTNFjTa1WY9WqVaauBtVz7du3BwBkZ2fjxo0biI+PlwJAjx494O7urlWOdBswYIB0Ws/R0RHDhg3Dk08+if/+978ICgpCUVER1Go1T5dSreJpPiIiI8vIyAAA3L17F15eXrhw4QJ69+6NCxcuwMvLC3fv3tUqR7opFAps2rQJwMMLSrZu3YpZs2Zh69at0gUnmzZtgkKhMGU1qZ5hmCIiMjLNQJ1jxozBnTt3MHXqVDRp0gRTp05FRkYGRo8erVWOqiaTyWBlZaU1zcrKqtxwCUS1QSbq4d0gs7Ky4ODggMzMTN6lnYiMrqSkBL6+vmjYsCFu376tdTWfl5cXXFxckJGRgeTkZLaoVEGzLdu3b4///Oc/OHr0qHTKtGfPnhg+fDhOnz7NbfmYMtfjN1umiIiMTKFQYOTIkfj9998rvJrv999/x4gRI3jwr4YjR47g0qVLmDdvHiwtLREYGIhRo0YhMDAQlpaWiIyMxMWLF3HkyBFTV5XqEYYpIiIjKykpwbfffosuXbrAysoKU6ZMgbu7O6ZMmQK1Wo0uXbpg+/btKCkpMXVVzZ5mlPh27dpVOF8znaPJU21imCIiMjJNa8rKlSuRmpqKQ4cOYcuWLTh06BBSUlLw8ccfszWlmjT9yk6fPl3hfM109j+j2sShEYiIjKxsa4pCoUBgYKDWfLamVN9TTz0Fb29vvPfee4iJidG6t2FpaSmioqLQrFkzPPXUUyasJdU3DFNEREZWtjWle/fu5eazNaX6FAoFli1bhhEjRiA4OBgDBw7Uup3M7t27sX37dvY/o1rFq/nM6GoAIno8lb0CraLWlJCQEF6BVkOzZ8/G8uXLUVxcLE2zsLDAzJkzsWTJEhPWjIzJXI/fbJkiIjKysq0pISEhiIyMRLt27XD69GlERUVh165dbE2pgR07dmDp0qUYPHgwBg0aJLVM7d27F0uXLkX37t0xbNgwU1eT6hG2TJlRsiWix9uOHTswa9YsrRsaN2vWDEuXLuXBv5rYyle/mevxm2HKjN4MInr8lZSU4MiRI9JAk0899RQP+jUQGxuLPn36ID4+Hk8++WS5bXns2DEEBATg0KFD5Tr6U91nrsdvnuYjIqpFFV3NR9WnueIxNTUVo0aN0mrl8/b2xr/+9S+tckS1geNMERFRnaG54nHs2LFo37494uPjkZ2djfj4eLRv3x5jx47VKkdUG3iaz4yaCYmIqHKFhYWwsbGBs7Mzrl27BguLv06wFBcXw8PDAxkZGcjNzYVSqTRhTckYzPX4zZYpIiKqM+Li4lBcXIxbt25h2LBhWi1Tw4YNw61bt1BcXIy4uDhTV5XqEYYpIiKqMzR9ob766iucOnUKAQEBsLe3R0BAAE6fPo2vvvpKqxxRbWAHdCIiqjM0faF8fHyQkpJS4dV8ZcsR1Qb2mTKjc65ERFQ5jjNVv5nr8Zun+YiIqM7QjCa/a9cuhISEaPWZCgkJwa5du7B06VIGKapVPM1HRER1yrBhw7B9+3bMmjULAQEB0vRmzZph+/btHE2eah1P85lRMyEREVUfR5Ovf8z1+M2WKSIiqpM4mjyZC/aZIiIiItKDScOUt7c3ZDKZ1t/7779f6XPy8/Px2muvwdnZGba2thg+fDhu3rxZSzUmIiIi0mbylqlFixYhLS1N+ps2bVql5WfOnInvv/8e3377LX7++WfcuHGDnQ2JiIjIZEzeZ8rOzg5ubm7VKpuZmYl169Zhy5Yt6Nu3LwBgw4YNaN26NX755Rd0797dmFUlIiIiKsfkLVPvv/8+nJ2d0bFjR0RHR6O4uFhn2ePHj6OoqAjPPPOMNK1Vq1Zo2rQp4uPjdT6voKAAWVlZWn9EREREhmDSlqnp06ejU6dOcHJyQlxcHCIjI5GWloYPP/ywwvLp6elQKpVwdHTUmt6oUSOkp6frfJ2oqCgsXLjQkFUnIiIT49AIZC4M3jI1d+7ccp3KH/07f/48ACA8PByBgYHw9/fHK6+8gmXLlmHlypUoKCgwaJ0iIyORmZkp/V29etWgyyciotq1Y8cO+Pr6ok+fPhg9ejT69OkDX19f7Nixw9RVo3rI4GFq1qxZOHfuXKV/zZs3r/C53bp1Q3FxMS5dulThfDc3NxQWFuL+/fta02/evFlpvyuVSgV7e3utPyIiqpt27NiBESNGoH379lq3k2nfvj1GjBjBQEW1zqxGQN+8eTPGjRuHO3fuoEGDBuXmZ2ZmwsXFBV9//TWGDx8OAEhKSkKrVq0QHx9f7Q7o5jqCKhERVY43Oq7fzPX4bbIO6PHx8VixYgUSExPx559/YvPmzZg5cybGjh0rBanr16+jVatWOHbsGADAwcEBoaGhCA8Px6FDh3D8+HFMnDgRPXr04JV8RET1wJEjR3Dp0iXMmzdPK0gBgFwuR2RkJC5evIgjR46YqIZUH5msA7pKpcLWrVuxYMECFBQUoFmzZpg5cybCw8OlMkVFRUhKSsKDBw+kacuXL4dcLsfw4cNRUFCAoKAgrFmzxhSrQEREtSwtLQ0A0K5duwrna6ZryhHVBpOFqU6dOuGXX36ptIy3tzcePQtpZWWF1atXY/Xq1casHhGRUeTl5SEiIgLJycnw8/NDdHQ01Gq1qatVZzRu3BgAcPr06QrPSJw+fVqrHFFtMKs+U7XFXM+5EtHjLSQkBDt37iw3PTg4GDExMbVfoTqIfabqN3M9fpt80E4iovpAE6SUSiXmzp2LlJQUzJ07F0qlEjt37kRISIipq1gnKBQKLFu2DLt27UJISIjW1XwhISHYtWsXli5dyiBFtYotU2aUbIno8ZSXlwdra2solUpkZ2dDqVRK8woLC2FnZ4fCwkI8ePCAp/yqaceOHZg1a5bWUDrNmjXD0qVLeb/Wx5i5Hr/ZMkVEZGQREREAHg5UXDZIAYBSqcSMGTO0ylHVhg0bhpSUFBw6dAhbtmzBoUOHkJyczCBFJmHyGx0TET3ukpOTAQAvv/xyhfNDQ0OxZMkSqRxVj0KhQGBgoKmrQcSWKSIiY/Pz8wMAfP755xXOX7dunVY5Iqpb2GfKjM65EtHjiX2miAzDXI/fbJkiIjIytVqN4OBgKTjNmTMHFy5cwJw5c6QgFRwczCBFVEexZcqMki0RPd44zhSRfsz1+M0O6EREtSQmJoYjoBM9hhimiIhqkVqtxqpVq0xdDSIyIPaZIiIiItIDwxQRERGRHhimiIiIiPTAMEVERESkB4YpIiIiIj0wTBERERHpgWGKiIiISA8MU0RERER6YJgiIiIi0gPDFBEREZEeGKaIiIiI9MAwRUTVkpKSAqVSCZlMBqVSiZSUFFNXiYjILPBGx0RUJblcDiGE9LioqAh+fn6QyWQoLS01Yc2IiEyPLVNEVKmyQcra2hrR0dGwtrYGAAghIJfza4SI6jd+CxKRTikpKVKQSktLQ25uLt544w3k5uYiLS0NwMNAxVN+RFSfMUwRkU5t2rQB8LBFys3NTWuem5ub1EKlKUdEVB8xTBGRTkVFRQCAhQsXVjh/3rx5WuWIiOojmSjbq7SeyMrKgoODAzIzM2Fvb2/q6hCZLaVSiaKiIlhbWyM3N7fcfBsbGzx48ACWlpYoLCw0QQ2JqD4x1+M3W6aISKezZ88CAB48eID09HSteenp6Xjw4IFWOSKi+ohhioh08vX1hUwmAwA0btwYNjY2WLx4MWxsbNC4cWMAgEwmg6+vrymrSURkUjzNZ0bNhETm6tFxpjQ4zhQR1SZzPX6zZYqIqlRaWork5GRYWloCACwtLZGcnMwgRUQEjoBORNXk6+vLTuZERBVgyxQRERGRHhimiIiIiPRgsjAVGxsLmUxW4d9vv/2m83mBgYHlyr/yyiu1WHMiIiKiv5isz1RAQIB0by+Nt956CwcOHECXLl0qfe7kyZOxaNEi6bHmlhZEREREtc1kYUqpVGrd66uoqAg7d+7EtGnTpHFtdKnoPmFEREREpmA2fab++9//IiMjAxMnTqyy7ObNm9GwYUO0a9cOkZGR0ijMuhQUFCArK0vrj4iIiMgQzGZohHXr1iEoKAgeHh6Vlhs9ejS8vLzg7u6OkydPYs6cOUhKSsKOHTt0PicqKkrnjVqJiIiI9GHwEdDnzp2LDz74oNIy586dQ6tWraTH165dg5eXF7Zt24bhw4fX6PUOHjyIfv36ISUlBT4+PhWWKSgoQEFBgfQ4KysLnp6eZjeCKhEREelmriOgG7xlatasWZgwYUKlZZo3b671eMOGDXB2dsbQoUNr/HrdunUDgErDlEqlgkqlqvGyiYiIiKpi8DDl4uICFxeXapcXQmDDhg0YN26cdKuKmkhISAAA6aarRERERLXJ5B3QDx48iIsXL+Lll18uN+/69eto1aoVjh07BgBITU3Fu+++i+PHj+PSpUv473//i3HjxuHpp5+Gv79/bVediIiIyPQd0NetW4eAgACtPlQaRUVFSEpKkq7WUyqV+Omnn7BixQrk5ubC09MTw4cPx/z582u72kREREQAjNABvS4w1w5sREREpJu5Hr9NfpqPiIiIqC5jmCIiIiLSA8MUERERkR4YpoiIiIj0wDBFREREpAeGKSIiIiI9MEwRERER6YFhioiIiEgPDFNEREREemCYIiIiItIDwxQRERGRHhimiIiIiPTAMEVERESkB4YpIiIiIj0wTBERERHpgWGKiIiISA8MU0RERER6YJgiIiIi0gPDFBER1Ul5eXkICwtDUFAQwsLCkJeXZ+oqUT0lE0IIU1eitmVlZcHBwQGZmZmwt7c3dXWIiKiGQkJCsHPnznLTg4ODERMTU/sVolphrsdvtkwREVGdoglSSqUSc+fORUpKCubOnQulUomdO3ciJCTE1FWkeoYtU2aUbImIqHJ5eXmwtraGUqlEdnY2lEqlNK+wsBB2dnYoLCzEgwcPoFarTVhTMgZzPX6zZYqIiOqMiIgIAEB4eLhWkAIApVKJGTNmaJUjqg0MU0REVGckJycDAF5++eUK54eGhmqVI6oNDFNERFRn+Pn5AQA+//zzCuevW7dOqxxRbWCfKTM650pERJVjn6n6zVyP32yZIiKiOkOtViM4OFgKTnPmzMGFCxcwZ84cKUgFBwczSFGtYsuUGSVbIiKqHo4zVT+Z6/HbwtQVICIiqqmYmBjk5eUhIiICycnJ8PPzQ3R0NFukyCQYpoiIqE5Sq9VYtWqVqatBxD5TRERERPpgmCIiIiLSA8MUERERkR4YpoiIiIj0wDBFREREpAeGKSIiIiI9GC1MLV68GAEBAbC2toajo2OFZa5cuYLBgwfD2toarq6uiIiIQHFxcaXLvXv3LsaMGQN7e3s4OjoiNDQUOTk5RlgDIiIioqoZLUwVFhZi5MiRePXVVyucX1JSgsGDB6OwsBBxcXH44osvsHHjRrz99tuVLnfMmDE4c+YM9u/fj127duHw4cOYMmWKMVaBiIiIqEpGv53Mxo0bMWPGDNy/f19r+t69ezFkyBDcuHEDjRo1AgB8+umnmDNnDm7fvq1180qNc+fOoU2bNvjtt9/QpUsXAMC+ffvw7LPP4tq1a3B3d69Wncx1OHoiIiLSzVyP3ybrMxUfH4/27dtLQQoAgoKCkJWVhTNnzuh8jqOjoxSkAOCZZ56BXC7Hr7/+qvO1CgoKkJWVpfVHREREZAgmC1Pp6elaQQqA9Dg9PV3nc1xdXbWmWVhYwMnJSedzACAqKgoODg7Sn6enp561JyIiInqoRmFq7ty5kMlklf6dP3/eWHX92yIjI5GZmSn9Xb161dRVIiIiosdEjW50PGvWLEyYMKHSMs2bN6/Wstzc3HDs2DGtaTdv3pTm6XrOrVu3tKYVFxfj7t27Op8DACqVCiqVqlr1IiIiIqqJGoUpFxcXuLi4GOSFe/TogcWLF+PWrVvSqbv9+/fD3t4ebdq00fmc+/fv4/jx4+jcuTMA4ODBgygtLUW3bt0MUi8iIiKimjBan6krV64gISEBV65cQUlJCRISEpCQkCCNCTVgwAC0adMGL730EhITE/HDDz9g/vz5eO2116RWpGPHjqFVq1a4fv06AKB169YYOHAgJk+ejGPHjuHo0aMICwvDiy++WO0r+YiIiIgMqUYtUzXx9ttv44svvpAed+zYEQBw6NAhBAYGQqFQYNeuXXj11VfRo0cP2NjYYPz48Vi0aJH0nAcPHiApKQlFRUXStM2bNyMsLAz9+vWDXC7H8OHD8fHHHxtrNYiIiIgqZfRxpsyRuY5TQURERLqZ6/Gb9+YjIiIi0gPDFBEREZEeGKaIiIiI9MAwRURERKQHhikiIiIiPTBMEREREemBYYqIiIhIDwxTRERERHpgmCIiIiLSA8MUERERkR4YpoiIiIj0wDBFREREpAeGKSIiIiI9MEwRERER6YFhioiIiEgPDFNEREREemCYIiIiItIDwxQRERGRHhimiIiIiPTAMEVERESkB4YpIiIiIj0wTBERERHpgWGKiIiISA8MU0RERER6YJgiIiIi0gPDFBEREZEeGKaIiIiI9MAwRURERKQHhikiIiIiPTBMEREREemBYYqIiIhIDwxTRERERHpgmCIiIiLSA8MUERERkR4YpoiIiIj0wDBFREREpAejhanFixcjICAA1tbWcHR0LDc/MTERo0aNgqenJ9RqNVq3bo2PPvqoyuV6e3tDJpNp/b3//vtGWAMiIiKiqlkYa8GFhYUYOXIkevTogXXr1pWbf/z4cbi6umLTpk3w9PREXFwcpkyZAoVCgbCwsEqXvWjRIkyePFl6bGdnZ/D6ExEREVWH0cLUwoULAQAbN26scP6kSZO0Hjdv3hzx8fHYsWNHlWHKzs4Obm5uBqknERERkT7Mqs9UZmYmnJycqiz3/vvvw9nZGR07dkR0dDSKi4srLV9QUICsrCytPyIiIiJDMFrLVE3FxcXhm2++we7duystN336dHTq1AlOTk6Ii4tDZGQk0tLS8OGHH+p8TlRUlNRSRkRERGRINWqZmjt3brnO34/+nT9/vsaVOH36NIKDg/HOO+9gwIABlZYNDw9HYGAg/P398corr2DZsmVYuXIlCgoKdD4nMjISmZmZ0t/Vq1drXEciIiKiitSoZWrWrFmYMGFCpWWaN29eowqcPXsW/fr1w5QpUzB//vwaPRcAunXrhuLiYly6dAktW7assIxKpYJKparxsomIiIiqUqMw5eLiAhcXF4O9+JkzZ9C3b1+MHz8eixcv/lvLSEhIgFwuh6urq8HqRURERFRdRuszdeXKFdy9exdXrlxBSUkJEhISAAC+vr6wtbXF6dOn0bdvXwQFBSE8PBzp6ekAAIVCIQW2Y8eOYdy4cThw4ACaNGmC+Ph4/Prrr+jTpw/s7OwQHx+PmTNnYuzYsWjQoIGxVoWIiIhIJ6OFqbfffhtffPGF9Lhjx44AgEOHDiEwMBDbt2/H7du3sWnTJmzatEkq5+XlhUuXLgEAHjx4gKSkJBQVFQF4eLpu69atWLBgAQoKCtCsWTPMnDkT4eHhxloNIiIiokrJhBDC1JWobVlZWXBwcEBmZibs7e1NXR0iIiKqBnM9fpvVOFNEREREdQ3DFBEREZEeGKaIiIiI9MAwRURERKQHhikiIiIiPTBMEREREemBYYqIiIhIDwxTRERERHpgmCIiIiLSA8MUERERkR4YpoiIiIj0YLQbHRPR46WwsBBr1qxBamoqfHx88M9//hNKpdLU1SIiMjmGKSKq0uzZs7F8+XIUFxdL0yIiIjBz5kwsWbLEhDUjIjI9nuYjokrNnj0b0dHRcHZ2xmeffYa0tDR89tlncHZ2RnR0NGbPnm3qKhIRmZRMCCFMXYnalpWVBQcHB2RmZsLe3t7U1SEyW4WFhbCxsYGzszOuXbsGC4u/GrOLi4vh4eGBjIwM5Obm8pQfERmduR6/2TJFRDqtWbMGxcXF+Ne//qUVpADAwsICixYtQnFxMdasWWOiGhIRmR7DFBHplJqaCgAYMmRIhfM10zXliIjqI4YpItLJx8cHALBr164K52uma8oREdVH7DNlRudcicwN+0wRkTkx1+M3W6aISCelUomZM2fi5s2b8PDwwNq1a3Hjxg2sXbsWHh4euHnzJmbOnMkgRUT1GseZIqJKacaRWr58OaZOnSpNt7CwQEREBMeZIqJ6j6f5zKiZkMiccQR0IjI1cz1+M0yZ0ZtBREREupnr8Zt9poiIiIj0wDBFREREpAeGKSIiIiI9MEwRERER6YFhioiIiEgPDFNEREREemCYIiIiItIDwxQRERGRHhimiIiIiPRQL+/Npxn0PSsry8Q1ISIiourSHLfN7eYt9TJMZWdnAwA8PT1NXBMiIiKqqezsbDg4OJi6GpJ6eW++0tJS3LhxA3Z2dpDJZKauToWysrLg6emJq1evmtX9h+oibkvD4bY0DG5Hw+G2NJy6sC2FEMjOzoa7uzvkcvPpqVQvW6bkcjk8PDxMXY1qsbe3N9uduq7htjQcbkvD4HY0HG5LwzH3bWlOLVIa5hPriIiIiOoghikiIiIiPTBMmSmVSoV33nkHKpXK1FWp87gtDYfb0jC4HQ2H29JwuC3/vnrZAZ2IiIjIUNgyRURERKQHhikiIiIiPTBMEREREemBYYqIiIhIDwxTRHVQbGwsZDIZ7t+/r1cZMq7AwEDMmDHD1NWgOk4mkyEmJsbU1XhsTZgwASEhIXoto16GqfT0dEybNg3NmzeHSqWCp6cnnnvuORw4cMDUVStn48aNcHR0rFZZzXqp1WrIZDLIZDIoFAp4eHggKiqqxjeG9Pb2xooVK2pe6WqoyXppBAYGSutlZWWFFi1amM161fSgefv2bbz66qto2rQpVCoV3NzcEBQUhKNHjxqsTgEBAUhLSzPYaME1DWfx8fFQKBTw8vIy6noaw3PPPYeBAwdWOO/IkSOQyWQ4efJkLdeqagsWLJA+I2X/WrVqpfM5WVlZePPNN9GqVStYWVnBxcUFMpkMy5YtM7ubyQJ/70fChAkTKtwuZd/jq1evYtKkSXB3d4dSqYSXlxdef/11ZGRk/O3XrakFCxagQ4cO5aanpaVh0KBBBnmNS5cuVbgtyv5t3Ljxby9f13YyRGDRl2bdExISDL7senc7mUuXLqFnz55wdHREdHQ02rdvj6KiIvzwww947bXXcP78+Rovs7CwEEqlstz0oqIiWFpaGqLaVSq7Xs2bN0fbtm3x2muv4aeffsLnn3+Ot99+G46Ojnj11VervUxdX6S1uV6Pmjx5MhYtWoSCggIcPHgQU6ZMqfF6mYPhw4ejsLAQX3zxBZo3b46bN2/iwIED0hf331FSUgKZTCbdr0qpVMLNzc1QVa6xdevWwc3NDdeuXcPWrVvRtWtXg6xnVXR9Hmvy/NDQUAwfPhzXrl0rd+upDRs2oEuXLvD399e3qkbRtm1btG3bFpmZmdJB0cKi4q/6+/fvo1evXsjMzMS//vUvPPnkk0hPT0e/fv3w0UcfITQ0VOePnsq286P7YnWeY2wDBw7Ehg0btKZpxlP6888/0aNHD7Ro0QJff/01mjVrhjNnziAiIgJ79+7FL7/8Uu3XMcY6GvJz7OnpibS0NOnx0qVLsW/fPvz000/SNHO8XYvZE/XMoEGDRJMmTUROTk65effu3RNCCHH58mUxdOhQYWNjI+zs7MTIkSNFenq6VO6dd94RTzzxhPjss8+Et7e3kMlkQgghAIg1a9aI5557TlhbW4t33nlHCCFETEyM6Nixo1CpVKJZs2ZiwYIFoqioSOt1p0yZIlxdXYVKpRJt27YV33//vTh06JAAoPWnWWZl69W7d2/x+uuvay2/U6dOIigoSFovGxsb4e7uLho2bChsbGxEly5dxEsvvSStl5WVVbnX1qzXxIkTRa9evYSlpaWwtLQUCoVCeHt7G2W9ynp0vYQQolOnTuL555+XHqekpIihQ4cKV1dXab3279+vtYxHX1vjyJEjolevXsLKykp4eHiIadOmVbifVKduXl5eYvHixWLixInC1tZWeHp6in//+9/SdgEgXnzxRa1l3Lp1S1hYWIitW7cKAOLXX38Vs2bNEu7u7kKtVgsAYvny5UIIIW1Da2tr4eXlJWQymQAg9u3bJw4dOiSefPJJoVKpBADRtWtXcenSJWkdW7duLZW3t7cX8+bNk963/Px8MXv2bOHh4SGUSqXw8fERn3/+ubh48WK57TZ+/Hid2yM7O1vY2toKAKJPnz5i8eLF0jxN3X/66SfRuXNnoVarRaNGjYSTk5O0n3z88cciMDBQ2NraCrVaLaysrISFhYVo2rSpsLGxEd9++620PEtLS2FrayteeuklYWdnJwYOHCiUSqXYv3+/6NWrl1CpVMLa2lpYWVkJOzs70adPH5GQkCA9v6LPc1FRkWjUqJF49913K1yvTz75RNy5c0e8+OKL0vvTrl07sWXLlkr3CwDiu+++0yrj4OAgNmzYID2+cuWKGDlypHBwcBANGjQQQ4cOFRcvXtS5rcvSrMv48eNFcHCwEEKIb7/9VrRr105YWVkJJycnYWVlJd555x0xceJEYWFhIWQymfjggw+06lj2r3fv3uLu3buiefPm0mdepVKJJk2aCCGEOHnypGjVqpUAIGxtbYWjo6NQKBTi4sWLwsvLS3To0EHan1QqlbCzsxNDhw4VPXv2lOqkUqlERkaGVIf8/Hzh6+srrK2t9doPNcpuj4oMHDhQABCrV68WISEhQq1WC19fX7Fx40ZhbW0txowZo/N1FQqF6Nu3r3j99deFs7Oz6N69uxg5cqSwsbERCoVCyGQy0bBhQzFnzhxRVFQkfT7lcrlQKpXCzc1N2NraChsbm3Kv4e7uLr0nmv3miSeeEFOmTBF9+vSRtt/kyZNFdnZ2ufWNjo4Wbm5uwsnJSfzzn/8UhYWFOvcZjZKSEvHee+8Jb29vYWVlJfz9/aXPW2lpqejXr58YMGCAKC0tFUIIkZGRIZo0aSLeeuutSt+fqt6DU6dOiYEDBwobGxvh6uoqxo4dK27fvi3N7927t5g2bZqIiIgQDRo0EI0aNSp33Dh37pzo2bOnUKlUonXr1mL//v1a266ifbum20uXehWmMjIyhEwmE++9957OMiUlJaJDhw6iV69e4vfffxe//PKL6Ny5s7TRhXi489nY2IiBAweKP/74QyQmJgohHr5Rrq6uYv369SI1NVVcvnxZHD58WNjb24uNGzeK1NRU8eOPP0rBQ/N63bt3F23bthU//vijSE1NFd9//73Ys2ePKCgoECtWrBD29vYiLS1NpKWlaX1gdK1X2S/w0tJScfjwYaFWq4Wjo6O0Xl999ZVo2rSp6NKli7hw4YKYP3++UCgUwtraWgwcOFAcOnRIuLq6ikWLFgkAomHDhmL9+vXi4MGDQq1Wi7CwMGFrayvefPNN0aZNGzFgwACDr9ejKlova2tr8cILL0hlEhISxKeffipOnTolrZeVlZW4fPmytK08PDzEokWLpNcW4mEIs7GxEcuXLxcXLlwQR48eFR07dhQTJkyosl6P1k2Ih2HKyclJrF69WiQnJ4uoqCghl8vF+fPnRVFRkXRQycvLk56zcuVK0bRpU/Hnn38KAOL5558XAQEB4vDhw+KPP/4QAISlpaW4cOGCViBt06aN+PLLL0VgYKBo2rSpsLe3F2+88YbYtGmTdIC4fPmySElJEWq1WqhUKvH++++Lbdu2CV9fX2Frayu9b//4xz+Ep6en2LFjh0hNTRU//fST2Lp1qyguLhb/+c9/BACRlJQk0tLSxP3793Vuj3Xr1onOnTsLW1tbMXToUNG8eXPpy1dT927duomDBw8Kf39/YW1tLdq0aSPtJ15eXmLs2LFi+/btQiaTiX/84x9i586dYsOGDUIul4u+ffsKIYS4e/eutB1mz54tUlJSRHh4uOjcubP0fgYEBIiePXuKli1bimHDholZs2YJZ2dn6eCt6/McEREhfHx8pHoLIcT69euFWq0W9+/fF9euXRPR0dHixIkTIjU1VXz88cdCoVCIX3/9Ved+UVWYKiwsFK1btxaTJk0SJ0+eFGfPnhWjR48WLVu2FAUFBVXuh4+GqRs3bggLCwvx4YcfiosXL4qTJ08KJycn0aBBA7Fy5Urh4OAgunbtKu2bQghx7NgxKeympaWJjIwMMXToUOHg4CDUarUYPHiw6Nmzp2jatKm4d++eaNy4sejcubOwsLAQbdu2FY0bNxYhISEiNzdXeHl5CUtLS6FUKsXgwYPFnj17xGeffSYAiGeeeUZcuHBBHDx4UMjlchEYGCitx9ChQwUAsXDhQr32Q43KDuSa708AwsPDQ2zZskUkJyeL6dOnC1tbWzFu3DjRoEEDsX37dunzVPZ1FQqFUKlUIiIiQpw6dUr4+PiIF154QVhZWYkXX3xRDBkyRDRp0kQ4OzuLt956Szg4OIg33nhDdO3aVVhbW4v+/fuLgwcPivXr1wsLCwvh7u4u0tLSxPHjx4VMJpPej++++076HnB1dRXDhg0Tp06dEgcOHBDNmjXTCpXjx48X9vb24pVXXhHnzp0T33//vbC2thZr167Vuc9o/Otf/xKtWrUS+/btE6mpqWLDhg1CpVKJ2NhYIYQQ165dEw0aNBArVqwQQggxcuRI0bVrV1FUVFTp+1PZe3Dv3j3h4uIiIiMjxblz58Qff/wh+vfvL/r06SOV6d27t7C3txcLFiwQFy5cEF988YWQyWTixx9/FEIIUVxcLFq2bCn69+8vEhISxJEjR0TXrl21PnMV7ds13V661Ksw9euvvwoAYseOHTrL/Pjjj0KhUIgrV65I086cOSMAiGPHjgkhHu58lpaW4tatW1rPBSBmzJihNa1fv37lwttXX30lGjduLIQQ4ocffhByuVwkJSVVWJ8NGzYIBweHGq1X7969haWlpbCxsRGWlpbSQbiq9XJxcREKhUJaLy8vL7F8+XKt9QoNDRVTpkzRWq8jR44IuVwu1q9fb9D1elRF62VlZSWOHj1a6fPatm0rVq5cKT3WrFdZmvUqS7NeZQNPZXV7NEyNHTtWelxaWipcXV3FJ598IoR4GDYACKVSKQICAkRkZKTw9/cXc+bMkX7dyeVycf36dSHEX61ZnTp1EpGRkVphStPKkpGRIbVgxcbGSmU0La6hoaHC3d1da388cuSIkMlkws3NTSQlJQkAWi15ZT26vMoEBASIFStWiO3btwtHR0cBQLRt21ZERkaKzz//XPpC0+wna9euFQCkbW1nZyc2btwoRo8eLfr376+17H79+gmVSiWEeNjqq/l1r9m2zzzzjPTr/ciRI8Le3l7k5+drvZ8+Pj5SS6Guz/O5c+cEAHHo0CFp2lNPPaX1vj5q8ODBYtasWdLjmoapr776SrRs2VIrwBUUFAi1Wi1++OGHSra4kNZFLpcLCwsLoVAopBbm0aNHS2U0++bNmzcFALFs2TKtfVOz/504cUIIIcSFCxcEADFo0CDRqFEjUVBQIO7cuSPUarWYMmWKaNCggfj000+lfXH37t1CLpeL9PR04eXlJTw9PYWTk5PIzc0VQjzcD3v16iVsbW1FSUmJEEKIkJAQ6f3X7Idubm5a20GjJvuhxvjx44VCoZBa5TV/ixcvFr/88ov0WZo/f770nJycHAFATJkyRfp+BSA2bdqktWyFQiGaNm0qhPjr/YuMjJTeR837FxYWJrU+xcbGit69e4vWrVtrraOmVUVj0KBB4tVXX5X2m2nTpokWLVqIBg0aaLWal93mmvX18vISxcXFUpmRI0dq/fDUKBum8vPzhbW1tYiLi9MqExoaKkaNGiU93rZtm7CyshJz584VNjY24sKFC9I8Xe9PZWHq3XffFQMGDNCadvXqVSmUCfHws9SrVy+tMk8++aSYM2eOEEKIvXv3CgsLC+kHshCiXMvUo/t22bpVd3vpUq/6TIlqdKY8d+4cPD094enpKU1r06YNHB0dce7cOTz55JMAAC8vL7i4uJR7fpcuXbQeJyYm4ujRo1i8eLE0raSkBPn5+Xjw4AESEhLg4eGBFi1a/N3VqnC9xowZgzfffBP37t3DO++8A5lMhrNnz0rrlZOTg/Xr10Mul6N3796Qy+XIzc2Fo6NjpeuVmJiIkydPorCwEAcOHMC8efOkMq+++ioKCgoMtl4VeXS9AgICEBAQIM3PycnBggULsHv3bqSlpaG4uBh5eXm4cuVKpcvVrNfmzZulaUIIlJaW4uLFi2jdunWN61q2T41MJoObmxtu3boFAJg0aRK2bdsGCwsLdOvWDTExMTh58iRGjhwpPae0tFTafpr3OCEhAb6+vlIZS0tL6XWcnJzQsmVLAEBQUBA6duwI4OGFCY6OjkhMTMSNGzcwb948rfdNU+bXX3+FQqFA7969a7yuZSUlJeHYsWP47rvv4OrqisGDB+OFF17A5cuXERsbiw8++EDaPhs2bICHh4e0f926dQtNmzZFeHg4Xn75ZVhZWaFbt25ITU2Fj48PAEgXi6Snp+Pnn3+GlZUVOnXqhNjYWISGhiIuLg5NmjTBxo0bsXHjRhQWFsLKykqqn5OTEwoKCpCamipNq+jz3KpVKwQEBGD9+vUIDAxESkoKjhw5gkWLFgF4+Dl+7733sG3bNly/fh2FhYUoKCiAtbX13952iYmJSElJgZ2dndb0/Px8rfpWpmXLlmjVqhWys7OxatUqTJo0Cf/9738xcuRIDBgwACUlJfD395f2qUf3zUedO3cOFhYWaNiwIdq3bw+lUglnZ2e0bNkSp06dwhNPPAGVSgWlUgl/f394e3ujtLQUSUlJAICGDRvC19dX2i6JiYlITExEUVER7OzsIJPJUFJSAgD45ZdfpHpMnToVMpms5htRhz59+uCTTz7Rmubk5ITk5GTpcdnPrI2NDezt7ZGdnQ0AlfYT9fb2BvDX+7dkyRLIZDLpfczPz4e9vT1yc3MxcuRIBAUFwcbGBr6+vkhPT0fjxo0BAB4eHjh69ChKSkqgUCgwefJkTJo0CcDDvqpbtmxBly5d4O7uDhsbG+n1e/bsKW3zRo0aAXjYd06hUEhlGjdujFOnTlW6jVJSUvDgwQP0799fa3phYaH0fQIAI0eOxHfffYf3338fn3zyCfz8/CpdblUSExNx6NAh2NralpuXmpoqfQ8+2k+xcePG0v6SlJQET09Prf5lXbt2rXYd/s72KqtehSk/Pz/IZLK/1cn8UWV35Mqm5+TkYOHChRg2bFi5slZWVlCr1XrXpaL1cnBwkA6627ZtK/fhe+ONN7B//36o1WpERkbihRdeQK9evaQvNV3rlZOTg6lTp2Lt2rWYPn06goKCpDKNGzeGUqk02HpV5NH18vX1Rffu3fHMM89ordfSpUvh6+sLtVqNESNGoLCwsNLlatZr+vTp5eY1bdr0b9X10S9fmUyG0tJS6fG4ceMwffp0fPfdd5DL5bh48SI+++wzjBs3DgAgl8tx/PhxKBQKZGRkoHv37vjyyy/Rr18/6b22srIqd8AJCQnB+vXrsWbNGvzyyy/o0qULfvrpJ+Tk5EChUCA8PFzrfQMevndlDyr6WLduHYqLi+Hu7i5NE0JApVIhLS0No0ePxp49e2BpaSntJ5p10GyfBQsWYPTo0ejbty8uXryINm3aYOvWrXj++efh5eUFADh8+LAUptq0aYNNmzbht99+Q1FREWQyGaZOnQqVSoWvvvpKKyRr9tOyHat1fZ5DQ0Mxbdo0rF69Ghs2bICPj48UNqOjo/HRRx9hxYoVaN++PWxsbDBjxoxK9zWZTFbux09RUZH0/5ycHHTu3FmrvhoV/cipiFKphL29PUpLS9GyZUv873//Q1xcHH788UesXLkSN27cQFZWFlxcXODo6Ijz58+X2zd10bWdAEhXET/q0c7vOTk5GDFiBL7++mvs2bMHTZo0AQAEBwfjyJEj0gFV8zkwFE14eZSvr6/0vlT0mb1x4wZcXFykej36/mn2bc26de7cGQ4ODrCzs5N+OAAPf7C89957WLp0KSIjIzF8+HBcvHgRLVq0wP79+9G9e/dydXvuueekZf/+++8oKipCixYtqnWQr+r7pyI5OTkAgN27d0vvi0bZmx8/ePBA+m4yxPdGTk4OnnvuOa3tpaEJmsDfW6fq0nfZ9WpoBCcnJwQFBWH16tXIzc0tN//+/fto3bo1rl69iqtXr0rTz549i/v376NNmzY1fs1OnTohKSkJvr6+5f7kcjn8/f1x7do1XLhwocLnK5VKnQGnuutVXFyMESNGID09XWqhOXr0KJ599lnk5uYiKCgIbm5u5S5lrei1O3XqhLNnz6Jz587IyMhAv379pL82bdoYdL2qYmtri9dffx1vvPGG9AV39OhRTJgwAc8//zzat28PNzc3XLp0qdrrVdH7ZKyrj4KDg5Gfn499+/Zhy5YtCAgIQG5urnTQLC0txa1bt+Dr64usrCwAQJMmTbR+eRUXF0v/v3fvHi5cuIDWrVujY8eOGDNmDACgdevW2LJlCzp16gRbW9ty75vmvXviiSdQWlqKn3/+ucL6arZDZe9bcXExvvzySyxbtgwJCQnSX2JiItzd3fH1119LYQiAtJ9cvny53LJatGiB3r17w8fHB8OGDZOuxIqLi4OdnR2+//57nDlzBiqVCo0bN0ZBQQH+/e9/o0uXLujSpQvOnj2LAQMG4M6dO/Dx8Sm3nzZs2LDK9+gf//gH5HI5tmzZgi+//BKTJk2SAsPRo0cRHByMsWPH4oknnkDz5s117u8aLi4uWldSJScn48GDB9LjTp06ITk5Ga6uruX2w797hZVMJkPPnj2xcOFCnDhxAjKZDKdOnYJcLseLL76IzZs3awU6zfucnZ2N4uJitG7dGsXFxbhz545UJiMjA0lJSfD390diYiIKCgqkeUePHoVcLpdaSYGHLQ95eXnSOp44cQK2trZ46qmnpPULCwvDpk2bcOLECQAPr7CrSHX2w5pwdnaWWmLKrgfw8DMYHx+PCRMmSK/76PtX9oCref/8/f1x5swZ+Pj4SOt38uRJ2NnZwcPDAx07dkTTpk3h4uKCdu3aYcuWLdKyVSqV1EJiYWGB8ePHAwAOHDiAF198Ee3bt0diYqLWd31F2/zvaNOmDVQqFa5cuVJu/yt7tmbWrFmQy+XYu3cvPv74Yxw8eFCa93fen06dOuHMmTPw9vYu97qVBfiyWrZsiatXr+LmzZvStN9++02rjKH3nbLqVZgCgNWrV6OkpARdu3bFf/7zHyQnJ+PcuXP4+OOP0aNHDzzzzDNo3749xowZgz/++APHjh3DuHHj0Lt373Kn8Krj7bffxpdffomFCxfizJkzOHfuHLZu3Yr58+cDAHr37o2nn34aw4cPx/79+3Hx4kXs3bsX+/btA/Cw+TgnJwcHDhzAnTt3tL54da3X7du3ce/ePa31io6OlsZV+eOPP+Ds7Ix169ahc+fOsLS0xOjRo8v94vL29sbhw4cBQDqYz5kzB3FxcXB2dsaXX36J6dOnY+XKlRg9erRR1qsqU6dOxYULF/Cf//wHwMNWuh07dkgH8NGjR5f7daFZr+vXr0sHCM16hYWFISEhAcnJydi5cyfCwsL+Vr0qk5GRgb59++K7775D7969MXv2bJw9exbHjh1DcHAw1Go1unfvDhcXF7z44ot49913ERERAQDYvHkzdu/eLS2roKAABw4cwOnTpzFhwgQ4Ojri999/R3x8PNLT0wE8bCZv3bo15syZg/z8fGzYsAGvvPIK9uzZg1WrViEoKAjz58+Ht7c3xo8fj0mTJiEmJgYXL15EbGwstm3bBuDhqTCZTIZdu3bh9u3b0q/Ysnbt2oV79+4hNDQUjRs3xvTp05GQkIDS0lL069cPS5cuxdatW6Xymv1k1qxZAB6O9RMTE4OhQ4ciNjYWo0ePxk8//YQff/wRjRo1whdffIFVq1ZhyJAh+Prrr9GhQwfI5XLI5XI8/fTT2Lx5M3r37i29nzExMfD398egQYOwYMECjBs3DnFxcXjzzTfx+++/V/le2dra4oUXXkBkZCTS0tIwYcIEaZ6fnx/279+PuLg4nDt3DlOnTtX6Iq9I3759sWrVKpw4cQK///47XnnlFa1fxGPGjEHDhg2lVhrNezB9+nRcu3atyvoCkE5t5+fnY8+ePYiMjMQPP/yAK1euYMeOHSgpKZFOBS1evBienp5ITk5GYmIizp49i8zMTFhaWuL555/Hn3/+CVdXVwQHByMuLg4ZGRlITEzE2LFj0aRJE3zwwQewsrLCZ599htLSUhw6dAjTpk3DSy+9JL0G8NdwE2fPnkVAQADOnz+PFi1a4OTJk9Jn7cSJE7h27Rq++eYb9O7dW6/9sCIFBQVIT0/X+tN8/letWgUAWLRoEQ4fPoyrV69i3759yM3NhZubG95++23pR8CKFStw8OBBHD58GK+88opWa5zm/YuLi8OlS5cwfvx4fPnllxgyZAjmz5+PiRMn4s0330R8fDzy8/Px559/IiEhAU5OTvj666/x66+/Anh4Ov/OnTsoKCjAyy+/DAA4ceIEJk2ahDFjxsDKygrjx4/H6dOndW7zv8POzg5vvPEGZs6ciS+++AKpqan4448/sHLlSnzxxRcAHrZarV+/Hps3b0b//v0RERGB8ePH4969e1W+P5mZmVo/shISEnD16lW89tpruHv3LkaNGoXffvsNqamp+OGHHzBx4sRqB5/+/fvDx8cH48ePx8mTJ3H06FHpeKR5j1xdXaFWq7Fv3z7cvHkTmZmZem0vLdXuXfUYuXHjhnjttdeEl5eXUCqVokmTJmLo0KFSR9PqDo3wKFTQuVQIIfbt2ycCAgKEWq0W9vb2omvXrlpXCWRkZIiJEycKZ2dnYWVlJdq1ayd27dolzX/llVeEs7OzQBVDCGjWS6VSCYVCUW69Ro8eLezs7ISNjY2wtbUVLi4uwsrKSnh6eopVq1YJLy8v0bBhQ2l58fHxwt/fv9wQAseOHRP9+/cXVlZWQi6XC5lMJlQqldHWS6OioRGEEGLq1Kmibdu2oqSkRFy8eFH06dNHqNVqab0efZ5mvTRDBzy6XprLlP39/bUu6a9J3Srq5P7EE0+Id955R+Tn54u5c+eKTp06SR1S1Wq1mD9/vnjw4IEQQoizZ8+K7t27CwsLC6FUKoVCoRAARK9evcTJkye1hkZo27atUCqVomvXruLgwYMiJCRENG7cWFhYWEhXuWk6+h47dkx06tRJyOVyqZO7h4eH9L7l5eWJmTNnisaNGwulUil8fX3F+vXrpXVYtGiRcHNzEzKZrMJL0ocMGSKeffZZIYTQWk8HBwdpew8YMECrg6rmajH8f4f8Nm3aiKefflp4enoKpVIpXX2mGRpBcwUdADFnzhxpW2sulti3b5/W+2ljYyMsLCykTtmenp5izJgx0sUYuj7PGnFxcQKAtF4aGRkZIjg4WNja2gpXV1cxf/58MW7cOK1Oto/uF9evXxcDBgwQNjY2ws/PT+zZs6fc0AhpaWli3LhxomHDhkKlUonmzZuLyZMni8zMTJ111HjnnXfKXf6t+VOpVFLn5bL75v3794Wrq6twcnISSqVSNGrUSLRu3Vo4OzsLmUymNTSChYWFUKvVIigoSOp0XHZohEcv0/fy8hJPPvmkCA4OFm+//bZwdnYWtra2IiQkRPTr16/cZ+2ll14STk5O4v79+3rth48aP358hdukZcuWUhng4TAejRo1kq7us7CwEJ9++qlU5o033hBKpVIAD4cV2bNnj1AoFFoXSWjePwcHB2k5arVazJgxQ1y7dk36fMpkMmFnZyc6d+4s7OzsRIMGDcTs2bPF8OHDpYs2NPsFAOHp6Sm9xsmTJ6s1NEJZr7/+utZV6WX3mbL7f2lpqVixYoVo2bKlsLS0FC4uLiIoKEj8/PPP4tatW6JRo0ZaF7EUFhaKzp07i3/84x+Vvj+63oPQ0FAhxMMLHZ5//nnh6Ogo1Gq1aNWqlZgxY4bUQb+i7//g4GCt918zNIJSqRStWrUS33//vdZ3ghBCfPbZZ8LT01PI5fJyQyNUZ3vpIhPCDIe4JSK9/fDDDxg0aBDy8/NNNlAi0YQJE3D//v1q3Q6lX79+aNu2LT7++GPjV6wK77zzDj788EOd/Zn0FRgYiA4dOlR5NwYhBPz8/PDPf/4T4eHhBq/H4+zo0aPo1asXUlJSpAtYjKVedUAnqi9u3ryJnTt3ws/Pj0GKzN69e/cQGxuL2NhYrFmzxtTVAQAsXLgQ3t7e+OWXX9C1a9dyo7nXhtu3b2Pr1q1IT0/HxIkTa/3165rvvvsOtra28PPzQ0pKCl5//XX07NnT6EEKYJiqc44cOVLpPZqq23/A3Jjzeplz3XR59tlnkZ2dbdQD05UrVyq9KOPs2bN/+0pIqlhFl45r7N27F0899VQt1sZwOnbsiHv37uGDDz6ocSdqY+6Hpg4wrq6uaNiwIdauXYsGDRqYtC51QXZ2NubMmYMrV66gYcOGeOaZZ7Bs2bJaeW2e5qtj8vLycP36dZ3zK7r0ty4w5/Uy57qZUnFxcbkrJcvy9vbWeU84+ntSUlJ0zmvSpInRhiQxZ9wPyRwwTBERERHpod4NjUBERERkSAxTRERERHpgmCIiIiLSA8MUERERkR4YpoiIiIj0wDBFREREpAeGKSIiIiI9/B8LAZ2KFOxlowAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Re-Select best subset"
      ],
      "metadata": {
        "id": "cN8xK1vpe9az"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculate the combination of all columns\n",
        "combinations = []\n",
        "column_combinations = []\n",
        "num_columns = len(scaled_df.columns)\n",
        "\n",
        "for r in range(1, num_columns+1):\n",
        "    for cols in itertools.combinations(scaled_df.columns, r):\n",
        "        combination_sum = scaled_df[list(cols)].sum(axis=1)\n",
        "        combinations.append(combination_sum)\n",
        "        column_combinations.append(list(cols))  # Save column name combinations\n",
        "\n",
        "subset_df = pd.concat(combinations, axis=1)\n",
        "subset_df.columns = ['Combination_{}'.format(i) for i in range(len(combinations))]\n",
        "\n",
        "subset_rank_df = pd.DataFrame()\n",
        "subset_rank_df = subset_df.rank(method='first', ascending=False).astype('int16')\n",
        "\n",
        "subset_rank_df['QuestionId'] = subset_rank_df.index  # Add QuestionId Column\n",
        "subset_rank_df = subset_rank_df[['QuestionId'] + subset_rank_df.columns[:-1].tolist()]  # Move the QuestionId column to the front\n",
        "\n",
        "subset_templates = []\n",
        "for i in range(len(combinations)):\n",
        "  selected_columns = ['QuestionId', 'Combination_'+str(i)]\n",
        "  subset_templates.append(subset_rank_df[selected_columns])\n",
        "\n",
        "for template in subset_templates:\n",
        "  template.columns = ['QuestionId', 'ranking']"
      ],
      "metadata": {
        "id": "o1PNXjQnfP3x"
      },
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate Public\n",
        "eval_pub_subset_result = []\n",
        "for idx in range(len(subset_templates)):\n",
        "  eval_pub_subset_result.append(evaluate(eval_validation, subset_templates[idx]))\n",
        "\n",
        "score_dict = {}\n",
        "for idx in range(len(subset_templates)):\n",
        "  score, percent = getCompareDF(eval_validation, eval_pub_subset_result[idx])\n",
        "  score_dict[str(column_combinations[idx])] = (score, percent)\n",
        "  #print(f'{column_combinations[idx]}: {score}, {percent}')\n",
        "\n",
        "sorted_scores = sorted(score_dict.items(), key=lambda x: x[1][0], reverse=True)[:10]\n",
        "sorted_scores_rev = sorted(score_dict.items(), key=lambda x: x[1][0], reverse=False)[:10]\n",
        "\n",
        "print('Best top 10')\n",
        "for key, value in sorted_scores:\n",
        "  print(f'{key}: {value}')\n",
        "\n",
        "print('======')\n",
        "print('Worst top 10')\n",
        "for key, value in sorted_scores_rev:\n",
        "  print(f'{key}: {value}')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6UvfdTbffXWJ",
        "outputId": "a4c133b9-e419-4b80-9430-33e53e375302"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best top 10\n",
            "['Subject_Score', 'AnswerValue_Entropy']: (22, 0.88)\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy']: (22, 0.88)\n",
            "['Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: (22, 0.88)\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy']: (22, 0.88)\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: (22, 0.88)\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: (22, 0.88)\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: (22, 0.88)\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: (22, 0.88)\n",
            "['AnswerValue_Entropy']: (21, 0.84)\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy']: (21, 0.84)\n",
            "======\n",
            "Worst top 10\n",
            "['Correct_Rate']: (6, 0.24)\n",
            "['Correct_Rate', 'Subject_Score']: (6, 0.24)\n",
            "['Correct_Rate', 'IsCorrect_Entropy']: (7, 0.28)\n",
            "['Correct_Rate', 'Subject_Score', 'IsCorrect_Entropy']: (7, 0.28)\n",
            "['Correct_Rate', 'Correct_Rate_Inverse']: (9, 0.36)\n",
            "['Correct_Rate', 'Correct_Rate_Inverse', 'Subject_Score']: (10, 0.4)\n",
            "['Subject_Score']: (12, 0.48)\n",
            "['Correct_Rate', 'QuestionTextLength']: (12, 0.48)\n",
            "['Correct_Rate', 'Subject_Score', 'QuestionTextLength']: (13, 0.52)\n",
            "['QuestionTextLength']: (14, 0.56)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate Private\n",
        "eval_priv_subset_result = []\n",
        "for idx in range(len(subset_templates)):\n",
        "  eval_priv_subset_result.append(evaluate(eval_test, subset_templates[idx]))\n",
        "\n",
        "score_dict = {}\n",
        "for idx in range(len(subset_templates)):\n",
        "  score, percent = getCompareDF(eval_test, eval_priv_subset_result[idx])\n",
        "  score_dict[str(column_combinations[idx])] = (score, percent)\n",
        "  #print(f'{column_combinations[idx]}: {score}, {percent}')\n",
        "\n",
        "sorted_scores = sorted(score_dict.items(), key=lambda x: x[1][0], reverse=True)[:10]\n",
        "sorted_scores_rev = sorted(score_dict.items(), key=lambda x: x[1][0], reverse=False)[:10]\n",
        "\n",
        "print('Best top 10')\n",
        "for key, value in sorted_scores:\n",
        "  print(f'{key}: {value}')\n",
        "\n",
        "print('======')\n",
        "print('Worst top 10')\n",
        "for key, value in sorted_scores_rev:\n",
        "  print(f'{key}: {value}')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "y84AQbglff27",
        "outputId": "874aa56d-424c-49cc-83f7-6f5cc6884fe3"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best top 10\n",
            "['Correct_Rate_Inverse', 'QuestionTextLength']: (18, 0.72)\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'QuestionTextLength']: (18, 0.72)\n",
            "['Correct_Rate_Inverse']: (17, 0.68)\n",
            "['Correct_Rate_Inverse', 'Subject_Score']: (17, 0.68)\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy', 'QuestionTextLength']: (17, 0.68)\n",
            "['Correct_Rate_Inverse', 'IsCorrect_Entropy', 'QuestionTextLength']: (17, 0.68)\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'QuestionTextLength']: (17, 0.68)\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'IsCorrect_Entropy', 'QuestionTextLength']: (17, 0.68)\n",
            "['Correct_Rate_Inverse', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: (17, 0.68)\n",
            "['Correct_Rate_Inverse', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy', 'QuestionTextLength']: (17, 0.68)\n",
            "======\n",
            "Worst top 10\n",
            "['Correct_Rate']: (8, 0.32)\n",
            "['Correct_Rate', 'Subject_Score']: (8, 0.32)\n",
            "['Correct_Rate', 'IsCorrect_Entropy']: (10, 0.4)\n",
            "['Correct_Rate', 'Subject_Score', 'IsCorrect_Entropy']: (10, 0.4)\n",
            "['Subject_Score']: (11, 0.44)\n",
            "['Correct_Rate', 'AnswerValue_Entropy']: (11, 0.44)\n",
            "['Correct_Rate', 'Subject_Score', 'AnswerValue_Entropy']: (11, 0.44)\n",
            "['Correct_Rate', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: (11, 0.44)\n",
            "['Correct_Rate', 'Subject_Score', 'AnswerValue_Entropy', 'IsCorrect_Entropy']: (11, 0.44)\n",
            "['IsCorrect_Entropy']: (12, 0.48)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## PCA"
      ],
      "metadata": {
        "id": "dl2gtQv9Rmxg"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.decomposition import PCA\n",
        "\n",
        "pca = PCA(n_components=2)\n",
        "pca.fit(scaled_df)\n",
        "\n",
        "pca.components_.T\n",
        "pca.components_[1] *= -1\n",
        "\n",
        "df_pca = pca.transform(scaled_df)\n",
        "print(df_pca)\n",
        "\n",
        "loadings = pca.components_.T\n",
        "loadings\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "FSnzNOoHRpX5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "67522a2c-2686-4013-9fbb-712b9ec3c0ac"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[ -0.81494127   0.23293888]\n",
            " [  1.20245408   0.58995168]\n",
            " [ -2.21461632 -11.70095998]\n",
            " ...\n",
            " [ -1.04829387   0.34366849]\n",
            " [ -0.75421173   0.40352323]\n",
            " [  0.84721218   0.27720888]]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 0.53270367,  0.01169801],\n",
              "       [-0.53270367, -0.01169801],\n",
              "       [ 0.06004273,  0.87170423],\n",
              "       [-0.54241508,  0.09033928],\n",
              "       [-0.30052082,  0.31034138],\n",
              "       [-0.21052693, -0.36794713]])"
            ]
          },
          "metadata": {},
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# plot the loadings\n",
        "for i, feature in enumerate(loadings):\n",
        "    plt.arrow(0, 0, feature[0], feature[1], color='r', alpha=0.5)\n",
        "    plt.text(feature[0], feature[1], f\"{scaled_df.columns[i]}\", color='r')\n",
        "\n",
        "plt.xlabel('PC1')\n",
        "plt.ylabel('PC2')\n",
        "plt.axis('equal')\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 522
        },
        "id": "qkJWxg6Cz5em",
        "outputId": "e04640bb-d03d-41f7-c6fd-93e6095c5d26"
      },
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(-0.6010567591620027,\n",
              " 0.5914054147192408,\n",
              " -0.43425531514491317,\n",
              " 0.9385959220894285)"
            ]
          },
          "metadata": {},
          "execution_count": 32
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.13"
    },
    "colab": {
      "provenance": []
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}