baseline_models.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BxKOKMQPq-7Y"
      },
      "source": [
        "This Notebook will leverage existing, out-of-the-box models to create some \n",
        "baseline models for the project.\n",
        "\n",
        "The models will be evaluated using k-fold cross validation and the results \n",
        "will be compared to the results of the models created in the next notebook.\n",
        "\n",
        "Note that although the models are out-of-the-box, they are based on the feature \n",
        "engineering efforts in the first notebook, and they will be tuned to optimize \n",
        "the results => These baseline models are reasonably competitive"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "y9ASLFpJq-7c"
      },
      "outputs": [],
      "source": [
        "import sys\n",
        "import subprocess\n",
        "\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import torch"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "e6N57QnQq-7d"
      },
      "outputs": [],
      "source": [
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.model_selection import StratifiedKFold, train_test_split\n",
        "from sklearn.model_selection import GridSearchCV, RandomizedSearchCV\n",
        "from sklearn.metrics import make_scorer\n",
        "from sklearn.metrics import f1_score, classification_report"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "1PIB5K-wq-7d"
      },
      "outputs": [],
      "source": [
        "IN_COLAB = 'google.colab' in sys.modules\n",
        "# if IN_COLAB:\n",
        "#     subprocess.run('pip install -U lightgbm'.split(\" \"))\n",
        "\n",
        "# find gpu device\n",
        "GPU_AVAILABLE = torch.cuda.is_available()\n",
        "\n",
        "RANDOM_SEED = 42"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 458
        },
        "id": "kpB7ru1Xq-7d",
        "outputId": "63d1acf8-f0ce-4fa5-b54a-a0c36829c96f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n",
            "Colab environment\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        user_id  next_order_id  days_to_next_order  reorder_ratio  \\\n",
              "0             1        1187899                14.0       0.909091   \n",
              "1             2        1492625                30.0       0.387097   \n",
              "2             3        1402502                15.0       1.000000   \n",
              "3             4        2557754                 0.0       0.000000   \n",
              "4             5        2196797                 6.0       0.444444   \n",
              "...         ...            ...                 ...            ...   \n",
              "206204   206205        1716008                10.0       0.315789   \n",
              "206205   206206        1904200                11.0       1.000000   \n",
              "206206   206207        1005822                18.0       1.000000   \n",
              "206207   206208        1882108                 7.0       0.823529   \n",
              "206208   206209         272231                30.0       0.500000   \n",
              "\n",
              "        cls_days_to_next_order  cls_reorder_ratio  \n",
              "0                            1                  3  \n",
              "1                            3                  1  \n",
              "2                            2                  3  \n",
              "3                            0                  0  \n",
              "4                            0                  1  \n",
              "...                        ...                ...  \n",
              "206204                       1                  1  \n",
              "206205                       1                  3  \n",
              "206206                       2                  3  \n",
              "206207                       1                  2  \n",
              "206208                       3                  2  \n",
              "\n",
              "[206209 rows x 6 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-bce1a421-3410-4b66-98c0-df4ad157eb01\">\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>user_id</th>\n",
              "      <th>next_order_id</th>\n",
              "      <th>days_to_next_order</th>\n",
              "      <th>reorder_ratio</th>\n",
              "      <th>cls_days_to_next_order</th>\n",
              "      <th>cls_reorder_ratio</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>1187899</td>\n",
              "      <td>14.0</td>\n",
              "      <td>0.909091</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>1492625</td>\n",
              "      <td>30.0</td>\n",
              "      <td>0.387097</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>1402502</td>\n",
              "      <td>15.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>2557754</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>2196797</td>\n",
              "      <td>6.0</td>\n",
              "      <td>0.444444</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>206204</th>\n",
              "      <td>206205</td>\n",
              "      <td>1716008</td>\n",
              "      <td>10.0</td>\n",
              "      <td>0.315789</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>206205</th>\n",
              "      <td>206206</td>\n",
              "      <td>1904200</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>206206</th>\n",
              "      <td>206207</td>\n",
              "      <td>1005822</td>\n",
              "      <td>18.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>206207</th>\n",
              "      <td>206208</td>\n",
              "      <td>1882108</td>\n",
              "      <td>7.0</td>\n",
              "      <td>0.823529</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>206208</th>\n",
              "      <td>206209</td>\n",
              "      <td>272231</td>\n",
              "      <td>30.0</td>\n",
              "      <td>0.500000</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>206209 rows × 6 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-bce1a421-3410-4b66-98c0-df4ad157eb01')\"\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-bce1a421-3410-4b66-98c0-df4ad157eb01 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-bce1a421-3410-4b66-98c0-df4ad157eb01');\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
        }
      ],
      "source": [
        "if IN_COLAB:\n",
        "    from google.colab import drive\n",
        "    drive.mount('/content/drive')\n",
        "    print(\"Colab environment\")\n",
        "    d_path = \"./drive/MyDrive/data/sequence-embeddings/\"\n",
        "else:\n",
        "    print(\"Local environment\")\n",
        "    d_path = \"./data/\"\n",
        "\n",
        "# Load the data\n",
        "features = pd.read_csv(d_path + \"features.csv\")\n",
        "labels = pd.read_csv(d_path + \"labels.csv\")\n",
        "labels\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2WREkmBOq-7e",
        "outputId": "5882b895-2e5f-46a0-ac61-eabda033117e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Features with null values: \n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "user_days_since_std    8686\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ],
      "source": [
        "print(\"Features with null values: \")\n",
        "features.isna().sum()[features.isna().sum() > 0]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "zIQ2ZQPBq-7f"
      },
      "outputs": [],
      "source": [
        "# Only user_days_since_std has missing values - fill with 0\n",
        "features = features.fillna(0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AYdrcUG3q-7f",
        "outputId": "6b027441-ad48-42bc-d845-f38c456f4d81"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(206209, 55) (206209, 6)\n",
            "(206209, 54) (206209, 5)\n"
          ]
        }
      ],
      "source": [
        "features.sort_values(\"user_id\", inplace=True)\n",
        "labels.sort_values(\"user_id\", inplace=True)\n",
        "print(features.shape, labels.shape)\n",
        "features.drop(\"user_id\", axis=1, inplace=True)\n",
        "labels.drop(\"user_id\", axis=1, inplace=True)\n",
        "print(features.shape, labels.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SyM59SQgq-7g",
        "outputId": "4607926b-b1c4-4664-8a58-d1c379aad04c"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n",
              "       1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n",
              "       1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,\n",
              "       1., 1., 1.])"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ],
      "source": [
        "# As we will try a lot of different models, we will standardize the data\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "scaler = StandardScaler()\n",
        "features = scaler.fit_transform(features)\n",
        "np.std(features, axis=0)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "3__1Prugq-7g"
      },
      "outputs": [],
      "source": [
        "# Models to try:\n",
        "# 1. Logistic Regression one-vs-rest\n",
        "# 2. Random Forest\n",
        "# 3. Neural Network\n",
        "# 4. Gradient Boosting"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "sR8RApBJq-7h"
      },
      "outputs": [],
      "source": [
        "def ml_pipeline(model_class, init_params, X, y, param_distributions, n_folds=5, n_iter=10, n_jobs=-1, random_state=RANDOM_SEED):\n",
        "    \"\"\"\n",
        "    Run a machine learning pipeline with cross-validation.\n",
        "    Input:\n",
        "        model_class: class of model to use, must be sklearn compatible\n",
        "        init_params: dictionary of parameters to pass to model_class\n",
        "        X: features, shape (n_samples, n_features)\n",
        "        y: labels for one task, shape (n_samples,)\n",
        "        params: dictionary of parameters to use in randomized search\n",
        "        n_folds: number of folds to use for cross-validation\n",
        "    Return:\n",
        "        - best_model\n",
        "        - best_params\n",
        "        - best_score\n",
        "    \"\"\"\n",
        "    clf = model_class(**init_params)\n",
        "    cv = StratifiedKFold(n_splits=n_folds, shuffle=True, random_state=RANDOM_SEED)\n",
        "    scorer = make_scorer(f1_score, average=\"macro\")\n",
        "    searcher = RandomizedSearchCV(clf, param_distributions=param_distributions, \n",
        "        cv=cv, scoring=scorer, n_iter=n_iter, refit=False,\n",
        "        error_score='raise', n_jobs=n_jobs, verbose=1, random_state=random_state\n",
        "        )\n",
        "    cv_results = searcher.fit(X, y)\n",
        "    # best_model = searcher.best_estimator_\n",
        "    best_params = searcher.best_params_\n",
        "    best_score = searcher.best_score_\n",
        "    return cv_results, best_params, best_score"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "JNbvDziHq-7h"
      },
      "outputs": [],
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.neural_network import MLPClassifier\n",
        "from lightgbm import LGBMClassifier\n",
        "\n",
        "\n",
        "class LGBMClassifierEarlyStopping(LGBMClassifier):\n",
        "    def __init__(self, **kwargs) -> None:\n",
        "        super().__init__(**kwargs)\n",
        "    \n",
        "    def fit(self, X, y):\n",
        "        X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.1)\n",
        "        eval_set = [(X_val, y_val)]\n",
        "        super().fit(X_train, y_train, eval_set=eval_set, verbose=False)\n",
        "\n",
        "all_models = {\n",
        "    \"logistic_regression\": {\n",
        "        \"model_class\": LogisticRegression,\n",
        "        \"init_params\": {\n",
        "            \"multi_class\": \"auto\",\n",
        "            \"solver\": \"saga\",\n",
        "            \"random_state\": RANDOM_SEED,\n",
        "            \"max_iter\": 1000,\n",
        "        },\n",
        "        \"param_distributions\": {\n",
        "            \"C\": np.logspace(-3, 3, 7),\n",
        "            \"penalty\": [\"l1\", \"l2\", \"none\"],\n",
        "            \"class_weight\": [None, \"balanced\"],\n",
        "            \n",
        "        }\n",
        "    },\n",
        "    \"lightgbm\": {\n",
        "        \"model_class\": LGBMClassifierEarlyStopping,\n",
        "        \"init_params\": {\n",
        "            \"n_estimators\": 1000, # early stopping will stop training before this\n",
        "            \"n_jobs\": -1,\n",
        "            \"early_stopping_rounds\": 10,\n",
        "            \"random_state\": RANDOM_SEED\n",
        "        },\n",
        "        \"param_distributions\": {\n",
        "            \"max_depth\": [5, 10, 15, 20],\n",
        "            \"learning_rate\": np.logspace(-5, 0, 10),\n",
        "            \"num_leaves\": np.logspace(3, 8, 10, base=2, dtype=int),\n",
        "            \"min_child_samples\": [2, 5, 10, 20, 50, 100],\n",
        "            \"class_weight\": [None, \"balanced\"],\n",
        "            \"reg_alpha\": [0, 0.1, 0.01, 0.001, 0.0001],\n",
        "            \"reg_lambda\": [0, 0.1, 0.01, 0.001, 0.0001],\n",
        "        }\n",
        "    },\n",
        "    \"random_forest\": {\n",
        "        \"model_class\": RandomForestClassifier,\n",
        "        \"init_params\": {\n",
        "            \"n_estimators\": 100,\n",
        "            \"random_state\": RANDOM_SEED\n",
        "        },\n",
        "        \"param_distributions\": {\n",
        "            \"max_depth\": [5, 10, 15, 20],\n",
        "            \"min_samples_split\": [5, 10, 20, 50],\n",
        "            \"min_samples_leaf\": [1, 2, 5, 10],\n",
        "            \"max_features\": [\"sqrt\", \"log2\"],\n",
        "            \"class_weight\": [None, \"balanced\"],\n",
        "        }\n",
        "    },\n",
        "    \"mlp\": {\n",
        "        \"model_class\": MLPClassifier,\n",
        "        \"init_params\": {\n",
        "            \"random_state\": RANDOM_SEED,\n",
        "            \"solver\": \"adam\",\n",
        "            \"tol\": 1e-4,\n",
        "            \"early_stopping\": True,\n",
        "            \"max_iter\": 200\n",
        "        },\n",
        "        \"param_distributions\": {\n",
        "            \"hidden_layer_sizes\": [(100,), (100, 100), (100, 100, 100), (200,), (200, 200), (200, 200, 200)],\n",
        "            \"activation\": [\"logistic\", \"relu\"],\n",
        "            \"alpha\": np.logspace(-5, 0, 6),\n",
        "            \"learning_rate\": [\"constant\", \"invscaling\", \"adaptive\"],\n",
        "            \"learning_rate_init\": np.logspace(-4, -1, 4)\n",
        "        }\n",
        "    }\n",
        "}"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "id": "K0dLMZBSq-7j"
      },
      "outputs": [],
      "source": [
        "# Set aside 10% of the data for final evaluation\n",
        "# Use the remaining 90% for training and validation\n",
        "X_train, X_test, Y_train, Y_test = train_test_split(\n",
        "    features, \n",
        "    labels[[\"cls_days_to_next_order\",\"cls_reorder_ratio\"]], \n",
        "    test_size=0.1, \n",
        "    random_state=RANDOM_SEED\n",
        "    )"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "5HS0SRuSq-7j",
        "outputId": "9e30a50c-ea04-405b-90ad-cc5cc25733a5"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        cls_days_to_next_order  cls_reorder_ratio\n",
              "170013                       2                  1\n",
              "65476                        2                  2\n",
              "92306                        3                  1\n",
              "19297                        2                  2\n",
              "133164                       1                  2"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-967a9308-8c9e-4697-9e98-adf2b5c305d3\">\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>cls_days_to_next_order</th>\n",
              "      <th>cls_reorder_ratio</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>170013</th>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>65476</th>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>92306</th>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19297</th>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>133164</th>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-967a9308-8c9e-4697-9e98-adf2b5c305d3')\"\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-967a9308-8c9e-4697-9e98-adf2b5c305d3 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-967a9308-8c9e-4697-9e98-adf2b5c305d3');\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": 20
        }
      ],
      "source": [
        "Y_train.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "uvIb2gJgq-7j"
      },
      "outputs": [],
      "source": [
        "tuning_results_1 = {}"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CyGG1AdIq-7k",
        "outputId": "cee8df92-3d56-4943-8352-f39611052bbc"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training logistic_regression...\n",
            "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n",
            "Best params: {'penalty': 'l1', 'class_weight': 'balanced', 'C': 1000.0}\n",
            "Best score: 0.40093155931738467\n",
            "Training lightgbm...\n",
            "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n",
            "Best params: {'reg_lambda': 0.0001, 'reg_alpha': 0.001, 'num_leaves': 118, 'min_child_samples': 2, 'max_depth': 10, 'learning_rate': 0.07742636826811278, 'class_weight': 'balanced'}\n",
            "Best score: 0.42082584432693826\n",
            "Training random_forest...\n",
            "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/joblib/externals/loky/process_executor.py:703: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
            "  \"timeout or by a memory leak.\", UserWarning\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best params: {'min_samples_split': 20, 'min_samples_leaf': 10, 'max_features': 'sqrt', 'max_depth': 15, 'class_weight': 'balanced'}\n",
            "Best score: 0.4154807820164793\n",
            "Training mlp...\n",
            "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/joblib/externals/loky/process_executor.py:703: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
            "  \"timeout or by a memory leak.\", UserWarning\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best params: {'learning_rate_init': 0.001, 'learning_rate': 'invscaling', 'hidden_layer_sizes': (200, 200), 'alpha': 1e-05, 'activation': 'relu'}\n",
            "Best score: 0.39202614827296006\n"
          ]
        }
      ],
      "source": [
        "# Hyperparameter tuning\n",
        "# Apply ml_pipeline to task #1: cls_days_to_next_order\n",
        "\n",
        "for name, model_details in all_models.items():\n",
        "    print(f\"Training {name}...\")\n",
        "    cv_results, best_params, best_score = ml_pipeline(\n",
        "        model_class=model_details[\"model_class\"],\n",
        "        init_params=model_details[\"init_params\"],\n",
        "        X=X_train, \n",
        "        y=Y_train[\"cls_days_to_next_order\"],\n",
        "        param_distributions=model_details[\"param_distributions\"],\n",
        "        n_folds=5,\n",
        "        n_iter=10\n",
        "    )\n",
        "    print(f\"Best params: {best_params}\")\n",
        "    print(f\"Best score: {best_score}\")\n",
        "    tuning_results_1[name] = {\n",
        "        \"cv_results\": cv_results,\n",
        "        \"best_params\": best_params,\n",
        "        \"best_score\": best_score\n",
        "    }"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cPQ0rAHHq-7k",
        "outputId": "70df3e41-22ca-48fd-c9d7-8608867b71f9"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'logistic_regression': {'cv_results': RandomizedSearchCV(cv=StratifiedKFold(n_splits=5, random_state=42, shuffle=True),\n",
              "                     error_score='raise',\n",
              "                     estimator=LogisticRegression(max_iter=1000, random_state=42,\n",
              "                                                  solver='saga'),\n",
              "                     n_jobs=-1,\n",
              "                     param_distributions={'C': array([1.e-03, 1.e-02, 1.e-01, 1.e+00, 1.e+01, 1.e+02, 1.e+03]),\n",
              "                                          'class_weight': [None, 'balanced'],\n",
              "                                          'penalty': ['l1', 'l2', 'none']},\n",
              "                     random_state=42, refit=False,\n",
              "                     scoring=make_scorer(f1_score, average=macro), verbose=1),\n",
              "  'best_params': {'penalty': 'l1', 'class_weight': 'balanced', 'C': 1000.0},\n",
              "  'best_score': 0.40093155931738467},\n",
              " 'lightgbm': {'cv_results': RandomizedSearchCV(cv=StratifiedKFold(n_splits=5, random_state=42, shuffle=True),\n",
              "                     error_score='raise',\n",
              "                     estimator=LGBMClassifierEarlyStopping(early_stopping_rounds=10),\n",
              "                     n_jobs=-1,\n",
              "                     param_distributions={'class_weight': [None, 'balanced'],\n",
              "                                          'learning_rate': array([1.00000000e-05, 3.59381366e-05, 1.29154967e-04, 4.64158883e-04,\n",
              "         1.66810054e-03, 5.99484250e-03, 2.15443469e-02, 7.74263683e-02,\n",
              "         2.78255940e-01, 1.00000000e+00]),\n",
              "                                          'max_depth': [5, 10, 15, 20],\n",
              "                                          'min_child_samples': [2, 5, 10, 20, 50,\n",
              "                                                                100],\n",
              "                                          'num_leaves': array([  8,  11,  17,  25,  37,  54,  80, 118, 174, 256]),\n",
              "                                          'reg_alpha': [0, 0.1, 0.01, 0.001,\n",
              "                                                        0.0001],\n",
              "                                          'reg_lambda': [0, 0.1, 0.01, 0.001,\n",
              "                                                         0.0001]},\n",
              "                     random_state=42, refit=False,\n",
              "                     scoring=make_scorer(f1_score, average=macro), verbose=1),\n",
              "  'best_params': {'reg_lambda': 0.0001,\n",
              "   'reg_alpha': 0.001,\n",
              "   'num_leaves': 118,\n",
              "   'min_child_samples': 2,\n",
              "   'max_depth': 10,\n",
              "   'learning_rate': 0.07742636826811278,\n",
              "   'class_weight': 'balanced'},\n",
              "  'best_score': 0.42082584432693826},\n",
              " 'random_forest': {'cv_results': RandomizedSearchCV(cv=StratifiedKFold(n_splits=5, random_state=42, shuffle=True),\n",
              "                     error_score='raise',\n",
              "                     estimator=RandomForestClassifier(random_state=42), n_jobs=-1,\n",
              "                     param_distributions={'class_weight': [None, 'balanced'],\n",
              "                                          'max_depth': [5, 10, 15, 20],\n",
              "                                          'max_features': ['sqrt', 'log2'],\n",
              "                                          'min_samples_leaf': [1, 2, 5, 10],\n",
              "                                          'min_samples_split': [5, 10, 20, 50]},\n",
              "                     random_state=42, refit=False,\n",
              "                     scoring=make_scorer(f1_score, average=macro), verbose=1),\n",
              "  'best_params': {'min_samples_split': 20,\n",
              "   'min_samples_leaf': 10,\n",
              "   'max_features': 'sqrt',\n",
              "   'max_depth': 15,\n",
              "   'class_weight': 'balanced'},\n",
              "  'best_score': 0.4154807820164793},\n",
              " 'mlp': {'cv_results': RandomizedSearchCV(cv=StratifiedKFold(n_splits=5, random_state=42, shuffle=True),\n",
              "                     error_score='raise',\n",
              "                     estimator=MLPClassifier(early_stopping=True,\n",
              "                                             random_state=42),\n",
              "                     n_jobs=-1,\n",
              "                     param_distributions={'activation': ['logistic', 'relu'],\n",
              "                                          'alpha': array([1.e-05, 1.e-04, 1.e-03, 1.e-02, 1.e-01, 1.e+00]),\n",
              "                                          'hidden_layer_sizes': [(100,),\n",
              "                                                                 (100, 100),\n",
              "                                                                 (100, 100, 100),\n",
              "                                                                 (200,),\n",
              "                                                                 (200, 200),\n",
              "                                                                 (200, 200, 200)],\n",
              "                                          'learning_rate': ['constant',\n",
              "                                                            'invscaling',\n",
              "                                                            'adaptive'],\n",
              "                                          'learning_rate_init': array([0.0001, 0.001 , 0.01  , 0.1   ])},\n",
              "                     random_state=42, refit=False,\n",
              "                     scoring=make_scorer(f1_score, average=macro), verbose=1),\n",
              "  'best_params': {'learning_rate_init': 0.001,\n",
              "   'learning_rate': 'invscaling',\n",
              "   'hidden_layer_sizes': (200, 200),\n",
              "   'alpha': 1e-05,\n",
              "   'activation': 'relu'},\n",
              "  'best_score': 0.39202614827296006}}"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ],
      "source": [
        "tuning_results_1"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "id": "642kX-2nq-7l"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "source": [
        "import pickle\n",
        "with open(d_path + \"tuning_results_1.pkl\", \"wb\") as f:\n",
        "    pickle.dump(tuning_results_1, f)\n"
      ],
      "metadata": {
        "id": "FsNe6vlEwHeN"
      },
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Hyperparameter tuning\n",
        "# Apply ml_pipeline to task #2: cls_reorder_ratio\n",
        "tuning_results_2 = {}\n",
        "for name, model_details in all_models.items():\n",
        "    print(f\"Training {name}...\")\n",
        "    cv_results, best_params, best_score = ml_pipeline(\n",
        "        model_class=model_details[\"model_class\"],\n",
        "        init_params=model_details[\"init_params\"],\n",
        "        X=X_train, \n",
        "        y=Y_train[\"cls_reorder_ratio\"],\n",
        "        param_distributions=model_details[\"param_distributions\"],\n",
        "        n_folds=5,\n",
        "        n_iter=10\n",
        "    )\n",
        "    print(f\"Best params: {best_params}\")\n",
        "    print(f\"Best score: {best_score}\")\n",
        "    tuning_results_2[name] = {\n",
        "        \"cv_results\": cv_results,\n",
        "        \"best_params\": best_params,\n",
        "        \"best_score\": best_score\n",
        "    }"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sS8gFCe-wakc",
        "outputId": "e4b35ae7-cae0-4fff-f42c-93ea38480e11"
      },
      "execution_count": 25,
      "outputs": [
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Training logistic_regression...\n",
            "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n",
            "Best params: {'penalty': 'l1', 'class_weight': 'balanced', 'C': 1000.0}\n",
            "Best score: 0.42818345854469475\n",
            "Training lightgbm...\n",
            "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n",
            "Best params: {'reg_lambda': 0.0001, 'reg_alpha': 0.001, 'num_leaves': 118, 'min_child_samples': 2, 'max_depth': 10, 'learning_rate': 0.07742636826811278, 'class_weight': 'balanced'}\n",
            "Best score: 0.45836202728999637\n",
            "Training random_forest...\n",
            "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/joblib/externals/loky/process_executor.py:703: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
            "  \"timeout or by a memory leak.\", UserWarning\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best params: {'min_samples_split': 20, 'min_samples_leaf': 1, 'max_features': 'log2', 'max_depth': 20, 'class_weight': 'balanced'}\n",
            "Best score: 0.4707900262724459\n",
            "Training mlp...\n",
            "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n",
            "Best params: {'learning_rate_init': 0.001, 'learning_rate': 'adaptive', 'hidden_layer_sizes': (100, 100), 'alpha': 0.1, 'activation': 'logistic'}\n",
            "Best score: 0.44935031908268297\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pickle\n",
        "with open(d_path + \"tuning_results_2.pkl\", \"wb\") as f:\n",
        "    pickle.dump(tuning_results_2, f)\n"
      ],
      "metadata": {
        "id": "WoB00KeqxBSR"
      },
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tuning_results_2"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "muf7ArcqxE2A",
        "outputId": "ad7ef206-78e8-40f3-9fba-2e89c8f30332"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'logistic_regression': {'cv_results': RandomizedSearchCV(cv=StratifiedKFold(n_splits=5, random_state=42, shuffle=True),\n",
              "                     error_score='raise',\n",
              "                     estimator=LogisticRegression(max_iter=1000, random_state=42,\n",
              "                                                  solver='saga'),\n",
              "                     n_jobs=-1,\n",
              "                     param_distributions={'C': array([1.e-03, 1.e-02, 1.e-01, 1.e+00, 1.e+01, 1.e+02, 1.e+03]),\n",
              "                                          'class_weight': [None, 'balanced'],\n",
              "                                          'penalty': ['l1', 'l2', 'none']},\n",
              "                     random_state=42, refit=False,\n",
              "                     scoring=make_scorer(f1_score, average=macro), verbose=1),\n",
              "  'best_params': {'penalty': 'l1', 'class_weight': 'balanced', 'C': 1000.0},\n",
              "  'best_score': 0.42818345854469475},\n",
              " 'lightgbm': {'cv_results': RandomizedSearchCV(cv=StratifiedKFold(n_splits=5, random_state=42, shuffle=True),\n",
              "                     error_score='raise',\n",
              "                     estimator=LGBMClassifierEarlyStopping(early_stopping_rounds=10),\n",
              "                     n_jobs=-1,\n",
              "                     param_distributions={'class_weight': [None, 'balanced'],\n",
              "                                          'learning_rate': array([1.00000000e-05, 3.59381366e-05, 1.29154967e-04, 4.64158883e-04,\n",
              "         1.66810054e-03, 5.99484250e-03, 2.15443469e-02, 7.74263683e-02,\n",
              "         2.78255940e-01, 1.00000000e+00]),\n",
              "                                          'max_depth': [5, 10, 15, 20],\n",
              "                                          'min_child_samples': [2, 5, 10, 20, 50,\n",
              "                                                                100],\n",
              "                                          'num_leaves': array([  8,  11,  17,  25,  37,  54,  80, 118, 174, 256]),\n",
              "                                          'reg_alpha': [0, 0.1, 0.01, 0.001,\n",
              "                                                        0.0001],\n",
              "                                          'reg_lambda': [0, 0.1, 0.01, 0.001,\n",
              "                                                         0.0001]},\n",
              "                     random_state=42, refit=False,\n",
              "                     scoring=make_scorer(f1_score, average=macro), verbose=1),\n",
              "  'best_params': {'reg_lambda': 0.0001,\n",
              "   'reg_alpha': 0.001,\n",
              "   'num_leaves': 118,\n",
              "   'min_child_samples': 2,\n",
              "   'max_depth': 10,\n",
              "   'learning_rate': 0.07742636826811278,\n",
              "   'class_weight': 'balanced'},\n",
              "  'best_score': 0.45836202728999637},\n",
              " 'random_forest': {'cv_results': RandomizedSearchCV(cv=StratifiedKFold(n_splits=5, random_state=42, shuffle=True),\n",
              "                     error_score='raise',\n",
              "                     estimator=RandomForestClassifier(random_state=42), n_jobs=-1,\n",
              "                     param_distributions={'class_weight': [None, 'balanced'],\n",
              "                                          'max_depth': [5, 10, 15, 20],\n",
              "                                          'max_features': ['sqrt', 'log2'],\n",
              "                                          'min_samples_leaf': [1, 2, 5, 10],\n",
              "                                          'min_samples_split': [5, 10, 20, 50]},\n",
              "                     random_state=42, refit=False,\n",
              "                     scoring=make_scorer(f1_score, average=macro), verbose=1),\n",
              "  'best_params': {'min_samples_split': 20,\n",
              "   'min_samples_leaf': 1,\n",
              "   'max_features': 'log2',\n",
              "   'max_depth': 20,\n",
              "   'class_weight': 'balanced'},\n",
              "  'best_score': 0.4707900262724459},\n",
              " 'mlp': {'cv_results': RandomizedSearchCV(cv=StratifiedKFold(n_splits=5, random_state=42, shuffle=True),\n",
              "                     error_score='raise',\n",
              "                     estimator=MLPClassifier(early_stopping=True,\n",
              "                                             random_state=42),\n",
              "                     n_jobs=-1,\n",
              "                     param_distributions={'activation': ['logistic', 'relu'],\n",
              "                                          'alpha': array([1.e-05, 1.e-04, 1.e-03, 1.e-02, 1.e-01, 1.e+00]),\n",
              "                                          'hidden_layer_sizes': [(100,),\n",
              "                                                                 (100, 100),\n",
              "                                                                 (100, 100, 100),\n",
              "                                                                 (200,),\n",
              "                                                                 (200, 200),\n",
              "                                                                 (200, 200, 200)],\n",
              "                                          'learning_rate': ['constant',\n",
              "                                                            'invscaling',\n",
              "                                                            'adaptive'],\n",
              "                                          'learning_rate_init': array([0.0001, 0.001 , 0.01  , 0.1   ])},\n",
              "                     random_state=42, refit=False,\n",
              "                     scoring=make_scorer(f1_score, average=macro), verbose=1),\n",
              "  'best_params': {'learning_rate_init': 0.001,\n",
              "   'learning_rate': 'adaptive',\n",
              "   'hidden_layer_sizes': (100, 100),\n",
              "   'alpha': 0.1,\n",
              "   'activation': 'logistic'},\n",
              "  'best_score': 0.44935031908268297}}"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Evaluation"
      ],
      "metadata": {
        "id": "oWyOeu6q41wL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# For cls_days_to_next_order task\n",
        "{model:{\"best_params\": result[\"best_params\"],\n",
        "        \"best_score\": result[\"best_score\"]} for model, result in tuning_results_1.items()}"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hQ4eQ2qN40mD",
        "outputId": "48afcbcb-e905-40c7-b42f-a2f124d3d3b1"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'logistic_regression': {'best_params': {'penalty': 'l1',\n",
              "   'class_weight': 'balanced',\n",
              "   'C': 1000.0},\n",
              "  'best_score': 0.40093155931738467},\n",
              " 'lightgbm': {'best_params': {'reg_lambda': 0.0001,\n",
              "   'reg_alpha': 0.001,\n",
              "   'num_leaves': 118,\n",
              "   'min_child_samples': 2,\n",
              "   'max_depth': 10,\n",
              "   'learning_rate': 0.07742636826811278,\n",
              "   'class_weight': 'balanced'},\n",
              "  'best_score': 0.42082584432693826},\n",
              " 'random_forest': {'best_params': {'min_samples_split': 20,\n",
              "   'min_samples_leaf': 10,\n",
              "   'max_features': 'sqrt',\n",
              "   'max_depth': 15,\n",
              "   'class_weight': 'balanced'},\n",
              "  'best_score': 0.4154807820164793},\n",
              " 'mlp': {'best_params': {'learning_rate_init': 0.001,\n",
              "   'learning_rate': 'invscaling',\n",
              "   'hidden_layer_sizes': (200, 200),\n",
              "   'alpha': 1e-05,\n",
              "   'activation': 'relu'},\n",
              "  'best_score': 0.39202614827296006}}"
            ]
          },
          "metadata": {},
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "task1_cv = pd.DataFrame({model:{\"best_params\": result[\"best_params\"],\n",
        "        \"best_score\": result[\"best_score\"]} for model, result in tuning_results_1.items()}).T\n",
        "task1_cv.reset_index(inplace=True)\n",
        "task1_cv.rename(columns={\"index\":\"model_type\"}, inplace=True)\n",
        "task1_cv"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 175
        },
        "id": "dhUr8tBXB098",
        "outputId": "72b645ff-046f-4345-e3d9-243fc7eab5ac"
      },
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            model_type                                        best_params  \\\n",
              "0  logistic_regression  {'penalty': 'l1', 'class_weight': 'balanced', ...   \n",
              "1             lightgbm  {'reg_lambda': 0.0001, 'reg_alpha': 0.001, 'nu...   \n",
              "2        random_forest  {'min_samples_split': 20, 'min_samples_leaf': ...   \n",
              "3                  mlp  {'learning_rate_init': 0.001, 'learning_rate':...   \n",
              "\n",
              "  best_score  \n",
              "0   0.400932  \n",
              "1   0.420826  \n",
              "2   0.415481  \n",
              "3   0.392026  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-f1d9f0d2-fbdb-45da-8093-dd0c70d4630f\">\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>model_type</th>\n",
              "      <th>best_params</th>\n",
              "      <th>best_score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>logistic_regression</td>\n",
              "      <td>{'penalty': 'l1', 'class_weight': 'balanced', ...</td>\n",
              "      <td>0.400932</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>lightgbm</td>\n",
              "      <td>{'reg_lambda': 0.0001, 'reg_alpha': 0.001, 'nu...</td>\n",
              "      <td>0.420826</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>random_forest</td>\n",
              "      <td>{'min_samples_split': 20, 'min_samples_leaf': ...</td>\n",
              "      <td>0.415481</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>mlp</td>\n",
              "      <td>{'learning_rate_init': 0.001, 'learning_rate':...</td>\n",
              "      <td>0.392026</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f1d9f0d2-fbdb-45da-8093-dd0c70d4630f')\"\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-f1d9f0d2-fbdb-45da-8093-dd0c70d4630f 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-f1d9f0d2-fbdb-45da-8093-dd0c70d4630f');\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": 38
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sns.barplot(x=\"model_type\", y=\"best_score\", data=task1_cv)\n",
        "plt.title(\"f1-score (macro) for cls_days_to_next_order\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        },
        "id": "TRPL40bg4V25",
        "outputId": "0f1c1dde-767a-49e1-b55b-22f4d7de7c0f"
      },
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "task2_cv = pd.DataFrame({model:{\"best_params\": result[\"best_params\"],\n",
        "        \"best_score\": result[\"best_score\"]} for model, result in tuning_results_2.items()}).T\n",
        "task2_cv.reset_index(inplace=True)\n",
        "task2_cv.rename(columns={\"index\":\"model_type\"}, inplace=True)\n",
        "task2_cv"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 175
        },
        "id": "Er_3hzIM6MJI",
        "outputId": "05e960cc-0429-4db2-fd9a-f2a7c0001ed7"
      },
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            model_type                                        best_params  \\\n",
              "0  logistic_regression  {'penalty': 'l1', 'class_weight': 'balanced', ...   \n",
              "1             lightgbm  {'reg_lambda': 0.0001, 'reg_alpha': 0.001, 'nu...   \n",
              "2        random_forest  {'min_samples_split': 20, 'min_samples_leaf': ...   \n",
              "3                  mlp  {'learning_rate_init': 0.001, 'learning_rate':...   \n",
              "\n",
              "  best_score  \n",
              "0   0.428183  \n",
              "1   0.458362  \n",
              "2    0.47079  \n",
              "3    0.44935  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1e297d3f-f0f8-4f37-bd82-84dcb786b718\">\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>model_type</th>\n",
              "      <th>best_params</th>\n",
              "      <th>best_score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>logistic_regression</td>\n",
              "      <td>{'penalty': 'l1', 'class_weight': 'balanced', ...</td>\n",
              "      <td>0.428183</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>lightgbm</td>\n",
              "      <td>{'reg_lambda': 0.0001, 'reg_alpha': 0.001, 'nu...</td>\n",
              "      <td>0.458362</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>random_forest</td>\n",
              "      <td>{'min_samples_split': 20, 'min_samples_leaf': ...</td>\n",
              "      <td>0.47079</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>mlp</td>\n",
              "      <td>{'learning_rate_init': 0.001, 'learning_rate':...</td>\n",
              "      <td>0.44935</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1e297d3f-f0f8-4f37-bd82-84dcb786b718')\"\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-1e297d3f-f0f8-4f37-bd82-84dcb786b718 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-1e297d3f-f0f8-4f37-bd82-84dcb786b718');\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": 40
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sns.barplot(x=\"model_type\", y=\"best_score\", data=task2_cv)\n",
        "plt.title(\"f1-score (macro) for cls_reorder_ratio\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        },
        "id": "akYGf1I_6VLo",
        "outputId": "747e0bca-ef4b-4b38-8c8d-37b27a6f038c"
      },
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from utils import auc_report, evaluate"
      ],
      "metadata": {
        "id": "QRprznLk6gbN"
      },
      "execution_count": 48,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import utils"
      ],
      "metadata": {
        "id": "YG2d_NgoAG5P"
      },
      "execution_count": 53,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import importlib\n",
        "importlib.reload(utils)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K6nJRPxu_3fJ",
        "outputId": "8313165d-274a-4324-d39d-c0db715e360d"
      },
      "execution_count": 55,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<module 'utils' from '/content/utils.py'>"
            ]
          },
          "metadata": {},
          "execution_count": 55
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate auc\n",
        "# For cls_days_to_next_order task\n",
        "for name, details in all_models.items():\n",
        "    model = details[\"model_class\"](**details[\"init_params\"], **tuning_results_1[name][\"best_params\"])\n",
        "    model.fit(X_train, Y_train[\"cls_days_to_next_order\"])\n",
        "    utils.evaluate(model, X_test, Y_test[\"cls_days_to_next_order\"])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "RlhqWsYB-IBM",
        "outputId": "fb7fdf81-ef2d-4e01-a2e8-fa264cd1fedd"
      },
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluating model\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.43      0.55      0.48      4587\n",
            "           1       0.41      0.28      0.33      5762\n",
            "           2       0.29      0.20      0.24      4380\n",
            "           3       0.47      0.62      0.54      5892\n",
            "\n",
            "    accuracy                           0.42     20621\n",
            "   macro avg       0.40      0.41      0.40     20621\n",
            "weighted avg       0.41      0.42      0.40     20621\n",
            "\n",
            "0: 0.7488951630716927\n",
            "1: 0.6447946226170104\n",
            "2: 0.609174536849211\n",
            "3: 0.7490585281946285\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/lightgbm/engine.py:123: UserWarning: Found `early_stopping_rounds` in params. Will use it instead of argument\n",
            "  warnings.warn(\"Found `{}` in params. Will use it instead of argument\".format(alias))\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluating model\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.49      0.51      0.50      4587\n",
            "           1       0.43      0.32      0.36      5762\n",
            "           2       0.30      0.26      0.27      4380\n",
            "           3       0.48      0.62      0.54      5892\n",
            "\n",
            "    accuracy                           0.44     20621\n",
            "   macro avg       0.42      0.43      0.42     20621\n",
            "weighted avg       0.43      0.44      0.43     20621\n",
            "\n",
            "0: 0.7649349013877449\n",
            "1: 0.6599115686060562\n",
            "2: 0.6300073746499291\n",
            "3: 0.7589258795831473\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluating model\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.50      0.48      0.49      4587\n",
            "           1       0.41      0.33      0.37      5762\n",
            "           2       0.29      0.23      0.25      4380\n",
            "           3       0.46      0.65      0.54      5892\n",
            "\n",
            "    accuracy                           0.43     20621\n",
            "   macro avg       0.42      0.42      0.41     20621\n",
            "weighted avg       0.42      0.43      0.42     20621\n",
            "\n",
            "0: 0.7573070077622005\n",
            "1: 0.6537248819920792\n",
            "2: 0.6230309501939817\n",
            "3: 0.7530673885201005\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluating model\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.52      0.43      0.47      4587\n",
            "           1       0.41      0.39      0.40      5762\n",
            "           2       0.34      0.07      0.11      4380\n",
            "           3       0.43      0.77      0.56      5892\n",
            "\n",
            "    accuracy                           0.44     20621\n",
            "   macro avg       0.43      0.42      0.39     20621\n",
            "weighted avg       0.43      0.44      0.40     20621\n",
            "\n",
            "0: 0.7574714310899019\n",
            "1: 0.6521413166210603\n",
            "2: 0.6108430127370861\n",
            "3: 0.7515082515675717\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate auc\n",
        "# For cls_reorder_ratio task\n",
        "for name, details in all_models.items():\n",
        "    model = details[\"model_class\"](**details[\"init_params\"], **tuning_results_2[name][\"best_params\"])\n",
        "    model.fit(X_train, Y_train[\"cls_reorder_ratio\"])\n",
        "    utils.evaluate(model, X_test, Y_test[\"cls_reorder_ratio\"])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "dUtmDPiW_Cvx",
        "outputId": "12867d83-e31f-41b5-b064-5b30e1772ce6"
      },
      "execution_count": 57,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluating model\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.22      0.61      0.32      1554\n",
            "           1       0.47      0.43      0.45      5380\n",
            "           2       0.63      0.38      0.47      9041\n",
            "           3       0.43      0.55      0.48      4646\n",
            "\n",
            "    accuracy                           0.45     20621\n",
            "   macro avg       0.44      0.49      0.43     20621\n",
            "weighted avg       0.51      0.45      0.46     20621\n",
            "\n",
            "0: 0.8167574965445632\n",
            "1: 0.7441378408709242\n",
            "2: 0.6809540265522311\n",
            "3: 0.7394854745731769\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/lightgbm/engine.py:123: UserWarning: Found `early_stopping_rounds` in params. Will use it instead of argument\n",
            "  warnings.warn(\"Found `{}` in params. Will use it instead of argument\".format(alias))\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluating model\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.26      0.54      0.35      1554\n",
            "           1       0.46      0.49      0.47      5380\n",
            "           2       0.62      0.44      0.51      9041\n",
            "           3       0.46      0.53      0.49      4646\n",
            "\n",
            "    accuracy                           0.48     20621\n",
            "   macro avg       0.45      0.50      0.46     20621\n",
            "weighted avg       0.52      0.48      0.49     20621\n",
            "\n",
            "0: 0.8292921749417266\n",
            "1: 0.7456848444166817\n",
            "2: 0.6957617944275731\n",
            "3: 0.7585339770964237\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluating model\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.34      0.35      0.34      1554\n",
            "           1       0.47      0.58      0.52      5380\n",
            "           2       0.60      0.50      0.55      9041\n",
            "           3       0.48      0.51      0.49      4646\n",
            "\n",
            "    accuracy                           0.51     20621\n",
            "   macro avg       0.47      0.48      0.48     20621\n",
            "weighted avg       0.52      0.51      0.51     20621\n",
            "\n",
            "0: 0.8253191904264439\n",
            "1: 0.7516220554564592\n",
            "2: 0.6916736536434768\n",
            "3: 0.761338415531694\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Evaluating model\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.41      0.21      0.27      1554\n",
            "           1       0.49      0.48      0.49      5380\n",
            "           2       0.56      0.68      0.62      9041\n",
            "           3       0.53      0.41      0.46      4646\n",
            "\n",
            "    accuracy                           0.53     20621\n",
            "   macro avg       0.50      0.45      0.46     20621\n",
            "weighted avg       0.53      0.53      0.52     20621\n",
            "\n",
            "0: 0.8253651571687971\n",
            "1: 0.7554786065467609\n",
            "2: 0.6936198156202247\n",
            "3: 0.7646930706542792\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "y9S76pH9A79n"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3.10.4 ('cs7641project')",
      "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": "3.10.4"
    },
    "orig_nbformat": 4,
    "vscode": {
      "interpreter": {
        "hash": "0347152bab7395cf6d4d53b744fcb402ec7f725dba3cd9225265b4a9dcbbfed9"
      }
    },
    "colab": {
      "provenance": [],
      "machine_shape": "hm"
    },
    "gpuClass": "standard"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}