notebooks/cohort/cohort_manager.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Managing cohorts\n",
    "\n",
    "This library allows users to apply certain mitigations over a specific cohort, instead of ​applying it to the entire dataset. This is useful when, for example, two cohorts have a very different label distribution and we want to rebalance each cohort individually, instead of rebalancing the entire dataset. In some scenarios, applying a mitigation over each cohort separately is more advantageous than applying it over the whole dataset.\n",
    "\n",
    "In this notebook, we'll show how to apply different pipelines for each cohort separately. To accomplish this, we'll use the `CohortManager` class. This class allows us to create multiple pipelines (using transformers from `scikit-learn` or from the current library) and apply them to each cohort separately. All this is done following the ``fit()`` and ``transform()`` paradigms, ​similar  to the other mitigations from the `dataprocessing` module.\n",
    "\n",
    "First of all, let's import the ``CohortManager`` class, which is found in the `cohort` module."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import xgboost as xgb\n",
    "from sklearn.pipeline import Pipeline\n",
    "\n",
    "from raimitigations.utils import create_dummy_dataset\n",
    "import raimitigations.dataprocessing as dp\n",
    "from raimitigations.cohort import CohortDefinition, CohortManager, plot_value_counts_cohort\n",
    "\n",
    "SEED = 42\n",
    "\n",
    "np.random.seed(SEED)\n",
    "random.seed(SEED)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Toy dataset\n",
    "\n",
    "Next, we'll create a toy dataset so we can showcase how the ``CohortManager`` operates."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_df(with_null: bool = True):\n",
    "    np.random.seed(SEED)\n",
    "    random.seed(SEED)\n",
    "    def add_nan(vec, pct):\n",
    "        vec = list(vec)\n",
    "        nan_index = random.sample(range(len(vec)), int(pct * len(vec)))\n",
    "        for index in nan_index:\n",
    "            vec[index] = np.nan\n",
    "        return vec\n",
    "\n",
    "    df = create_dummy_dataset(\n",
    "        samples=1000,\n",
    "        n_features=2,\n",
    "        n_num_num=0,\n",
    "        n_cat_num=2,\n",
    "        n_cat_cat=0,\n",
    "        num_num_noise=[0.01, 0.05],\n",
    "        pct_change=[0.05, 0.1],\n",
    "    )\n",
    "    if with_null:\n",
    "        col_with_nan = [\"num_0\", \"num_1\", ]\n",
    "        for col in col_with_nan:\n",
    "            if col != \"label\":\n",
    "                df[col] = add_nan(df[col], 0.1)\n",
    "\n",
    "    X = df.drop(columns=[\"label\"])\n",
    "    y = df[[\"label\"]]\n",
    "\n",
    "    return X, y\n",
    "\n",
    "# -----------------------------------\n",
    "def get_model():\n",
    "    model = xgb.XGBClassifier(\n",
    "            objective=\"binary:logistic\",\n",
    "            learning_rate=0.1,\n",
    "            n_estimators=30,\n",
    "            max_depth=10,\n",
    "            colsample_bytree=0.7,\n",
    "            alpha=0.0,\n",
    "            reg_lambda=10.0,\n",
    "            nthreads=4,\n",
    "            verbosity=0,\n",
    "            use_label_encoder=False,\n",
    "        )\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_0</th>\n",
       "      <th>num_1</th>\n",
       "      <th>CN_0_num_0</th>\n",
       "      <th>CN_1_num_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.601034</td>\n",
       "      <td>2.535353</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.755945</td>\n",
       "      <td>-2.172352</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.354479</td>\n",
       "      <td>NaN</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.103090</td>\n",
       "      <td>-0.766515</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.753178</td>\n",
       "      <td>1.787514</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>2.713939</td>\n",
       "      <td>1.451639</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>1.420812</td>\n",
       "      <td>2.535627</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>3.844315</td>\n",
       "      <td>1.211294</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>0.974124</td>\n",
       "      <td>2.619258</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>4.641478</td>\n",
       "      <td>-0.243075</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        num_0     num_1 CN_0_num_0 CN_1_num_1\n",
       "0    1.601034  2.535353     val0_0     val1_3\n",
       "1    1.755945 -2.172352     val0_1     val1_0\n",
       "2    2.354479       NaN     val0_0     val1_0\n",
       "3    4.103090 -0.766515     val0_1     val1_2\n",
       "4    1.753178  1.787514     val0_0     val1_1\n",
       "..        ...       ...        ...        ...\n",
       "995  2.713939  1.451639     val0_1     val1_2\n",
       "996  1.420812  2.535627     val0_0     val1_3\n",
       "997  3.844315  1.211294     val0_1     val1_2\n",
       "998  0.974124  2.619258     val0_0     val1_3\n",
       "999  4.641478 -0.243075     val0_1     val1_1\n",
       "\n",
       "[1000 rows x 4 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X, y = create_df()\n",
    "X"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Basic scenario\n",
    "\n",
    "Next, we'll explore a basic scenario where the ``CohortManager`` can prove its usefulness. In this scenario, we want to apply an imputation, followed by a min max scaler over each cohort separately. Once again: this is useful when each cohort has different behaviors or distributions for a particular column, so applying an imputation based on the mean value of the column over each cohort separately will make more sense than using the mean of the column of the entire dataset, for example.\n",
    "\n",
    "The `CohortManager` requires two main parameters: a transformation pipeline and a list of cohort definitions. The latter is already explored and explained in the notebook showcasing the `CohortDefinition` class (which is used internally by the `CohortManager`). The former is simply a list of objects that implement the `fit()`, `transform()`, `predict()` paradigm, or the `fit_resample()` method (for rebalancing classes, such as the `Rebalance` class from the `dataprocessing` module, or the rebalance classes from the `imblearn` library).\n",
    "\n",
    "There are two constraints when defining the cohorts:\n",
    "\n",
    "1. The definitions must not allow any instance to belong to more than 1 cohort. If that happens, an error is raised;\n",
    "2. The definitions must include all instances (without repetition due to the previous constraint). Since in some cases it is difficult to define a set of conditions for the last cohort such that it includes all instances not belonging to any of the previous cohorts, we created an easier way to accomplish this: the condition list for the last cohort is allowed to be `None`. When this happens, this cohort is understood as being the one with all the remaining instances that don't belong to any of the other cohorts.\n",
    "\n",
    "Once we instantiate the `CohortManager` object, we can treat it similarly to other classes from the `dataprocessing` module: call the `fit()` and `transform()` methods. Internally, what will happen is the following: \n",
    "\n",
    "1. the cohort pipeline will be duplicated, such that we'll have one pipeline for each cohort, with different objects instantiation;\n",
    "2. during the `fit()` method, we'll cycle through all cohorts, and for each cohort:\n",
    "    * we'll first filter the dataset provided (in this case, `X` and `y`) so that it includes only instances that belong to the current cohort;\n",
    "    * with the subset of the current cohort, we'll call the `fit()` method for all the transformations in the cohort's pipeline (in the order in which they appear) using the subset dataset as input;\n",
    "3. during the ``transform()`` method, we'll use a similar approach, but instead of calling the `fit()` method for the transformers in each cohort's pipeline, this time we'll call the `transform()` method;\n",
    "    * if all cohorts are compatible with each other, that is, have the same columns after applying all transforms, then all subsets are concatenated together. The full dataset is then returned. If there are any inconsistencies between the cohorts, then we'll just add each subset into a list of subsets, and the final variable returned in this case is a list of subsets (we'll explore some examples ahead)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_0</th>\n",
       "      <th>num_1</th>\n",
       "      <th>CN_0_num_0</th>\n",
       "      <th>CN_1_num_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.616278</td>\n",
       "      <td>0.790075</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.340936</td>\n",
       "      <td>0.040024</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.910331</td>\n",
       "      <td>0.660560</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.816398</td>\n",
       "      <td>0.284064</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.675657</td>\n",
       "      <td>0.688320</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>0.534997</td>\n",
       "      <td>0.669115</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>0.545942</td>\n",
       "      <td>0.790112</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>0.763978</td>\n",
       "      <td>0.627393</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>0.371610</td>\n",
       "      <td>0.801491</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>0.925459</td>\n",
       "      <td>0.374928</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        num_0     num_1 CN_0_num_0 CN_1_num_1\n",
       "0    0.616278  0.790075     val0_0     val1_3\n",
       "1    0.340936  0.040024     val0_1     val1_0\n",
       "2    0.910331  0.660560     val0_0     val1_0\n",
       "3    0.816398  0.284064     val0_1     val1_2\n",
       "4    0.675657  0.688320     val0_0     val1_1\n",
       "..        ...       ...        ...        ...\n",
       "995  0.534997  0.669115     val0_1     val1_2\n",
       "996  0.545942  0.790112     val0_0     val1_3\n",
       "997  0.763978  0.627393     val0_1     val1_2\n",
       "998  0.371610  0.801491     val0_0     val1_3\n",
       "999  0.925459  0.374928     val0_1     val1_1\n",
       "\n",
       "[1000 rows x 4 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohort_pipeline = [\n",
    "    dp.BasicImputer(verbose=False),\n",
    "    dp.DataMinMaxScaler(verbose=False),\n",
    "]\n",
    "\n",
    "c1 = [ ['CN_0_num_0', '==', 'val0_1'], 'and', ['num_0', '>', 0.0] ]\n",
    "c2 = [ ['CN_0_num_0', '==', 'val0_0'], 'and', ['num_0', '>', 0.0] ]\n",
    "c3 = None\n",
    "\n",
    "cohort_set = CohortManager(\n",
    "    transform_pipe=cohort_pipeline,\n",
    "    cohort_def=[c1, c2, c3]\n",
    ")\n",
    "cohort_set.fit(X=X, y=y)\n",
    "new_X = cohort_set.transform(X)\n",
    "new_X"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pipelines that result in incompatible cohorts\n",
    "\n",
    "As previously mentioned, in some cases, the pipeline associated with one or more cohorts may result in a cohort incompatible with at least one of the other cohorts. For example, consider the case where we add a one-hot encoding transformation in the pipeline. This will probably result in each cohort having different columns, which are the columns created by the one-hot encoder (for example, in case one of the cohorts doesn't have all existing values in a categorical column that is being encoded). \n",
    "\n",
    "This is also valid for any other encoder method since categorical encoders require a list of all valid categorical values in the column before creating the encoding. However, for the ordinal encoder, for example, it is not easy to automatically identify an incompatibility between the cohorts, since the ordinal encoder doesn't change the number of columns. To remedy this (at least for the mitigations from the `dataprocessing` module), we implemented a way for the `CohortManager` to query if a given class results in incompatible cohorts or not. Therefore, if we use the `EncoderOrdinal` from the ``dataprocessing`` module in the pipeline, we'll be able to automatically identify that each cohort will be incompatible between each other, and thus, the ``transform()`` method will return a dictionary of subsets instead of a single dataset. This dictionary will have a key assigned to each of the cohorts' names."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING: the transformations used over the cohorts resulted in each cohort having different columns. The transform() method will return a list of transformed subsets (one for each cohort).\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "dict_keys(['cohort_0', 'cohort_1', 'cohort_2'])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohort_pipeline = [\n",
    "    dp.BasicImputer(verbose=False),\n",
    "    dp.DataMinMaxScaler(verbose=False),\n",
    "    dp.EncoderOrdinal(verbose=False)\n",
    "]\n",
    "\n",
    "cohort_set = CohortManager(\n",
    "    transform_pipe=cohort_pipeline,\n",
    "    cohort_def=[c1, c2, c3]\n",
    ")\n",
    "cohort_set.fit(X=X, y=y)\n",
    "subset_dict = cohort_set.transform(X)\n",
    "subset_dict.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_0</th>\n",
       "      <th>num_1</th>\n",
       "      <th>CN_0_num_0</th>\n",
       "      <th>CN_1_num_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.340936</td>\n",
       "      <td>0.040024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.816398</td>\n",
       "      <td>0.284064</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.402537</td>\n",
       "      <td>0.692783</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.299020</td>\n",
       "      <td>0.843083</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.626370</td>\n",
       "      <td>0.180212</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>990</th>\n",
       "      <td>0.371002</td>\n",
       "      <td>0.696937</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>991</th>\n",
       "      <td>0.358171</td>\n",
       "      <td>0.089578</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>0.534997</td>\n",
       "      <td>0.669115</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>0.763978</td>\n",
       "      <td>0.627393</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>0.925459</td>\n",
       "      <td>0.374928</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>542 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        num_0     num_1  CN_0_num_0  CN_1_num_1\n",
       "1    0.340936  0.040024           0           0\n",
       "3    0.816398  0.284064           0           2\n",
       "5    0.402537  0.692783           0           3\n",
       "8    0.299020  0.843083           0           1\n",
       "10   0.626370  0.180212           0           2\n",
       "..        ...       ...         ...         ...\n",
       "990  0.371002  0.696937           0           2\n",
       "991  0.358171  0.089578           0           0\n",
       "995  0.534997  0.669115           0           2\n",
       "997  0.763978  0.627393           0           2\n",
       "999  0.925459  0.374928           0           1\n",
       "\n",
       "[542 rows x 4 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subset_dict['cohort_0']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Adding estimators to the pipeline\n",
    "\n",
    "The pipelines are not limited to having only transformers, as we can also add estimators to them. Some considerations when adding an estimator into the pipeline:\n",
    "\n",
    "1. The estimator must be the last object in the pipeline (this implies that we can only have 1 estimator in each pipeline);\n",
    "2. The estimator must implement at least the `predict()` method. It can also implement the `predict_proba()` method, but the latter is not mandatory;\n",
    "3. If the pipeline is comprised of a set of transformers (that implement the `transform()` method) and the estimator (that is, the pipeline doesn't contain *only* the estimator), then the cohort manager object will be allowed to call the `transform()` and `predict()` methods (and the `predict_proba()` as well, but only if the estimator implements this method)\n",
    "    * in this case, the `transform()` method will call only the `transform()` methods of the transformers up until the estimator, since this latter doesn't have a `transform()` method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mmendonca/anaconda3/envs/raipub/lib/python3.9/site-packages/xgboost/sklearn.py:1421: UserWarning: `use_label_encoder` is deprecated in 1.7.0.\n",
      "  warnings.warn(\"`use_label_encoder` is deprecated in 1.7.0.\")\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<raimitigations.cohort.cohort_manager.CohortManager at 0x7fc074d77d00>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohort_pipeline = [\n",
    "    dp.BasicImputer(verbose=False),\n",
    "    dp.DataMinMaxScaler(verbose=False),\n",
    "    dp.EncoderOrdinal(verbose=False),\n",
    "    get_model()\n",
    "]\n",
    "\n",
    "cohort_set = CohortManager(\n",
    "    transform_pipe=cohort_pipeline,\n",
    "    cohort_def=[c1, c2, c3]\n",
    ")\n",
    "cohort_set.fit(X=X, y=y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When calling the ``transform()`` method, it will return a list of subsets, since the ``EncoderOrdinal`` results in incompatible cohorts (as previously mentioned). The ``transform()`` method will run the ``transform()`` of all transformers up until we get to the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING: the transformations used over the cohorts resulted in each cohort having different columns. The transform() method will return a list of transformed subsets (one for each cohort).\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_0</th>\n",
       "      <th>num_1</th>\n",
       "      <th>CN_0_num_0</th>\n",
       "      <th>CN_1_num_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.340936</td>\n",
       "      <td>0.040024</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.816398</td>\n",
       "      <td>0.284064</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.402537</td>\n",
       "      <td>0.692783</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.299020</td>\n",
       "      <td>0.843083</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.626370</td>\n",
       "      <td>0.180212</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>990</th>\n",
       "      <td>0.371002</td>\n",
       "      <td>0.696937</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>991</th>\n",
       "      <td>0.358171</td>\n",
       "      <td>0.089578</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>0.534997</td>\n",
       "      <td>0.669115</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>0.763978</td>\n",
       "      <td>0.627393</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>0.925459</td>\n",
       "      <td>0.374928</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>542 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        num_0     num_1  CN_0_num_0  CN_1_num_1\n",
       "1    0.340936  0.040024           0           0\n",
       "3    0.816398  0.284064           0           2\n",
       "5    0.402537  0.692783           0           3\n",
       "8    0.299020  0.843083           0           1\n",
       "10   0.626370  0.180212           0           2\n",
       "..        ...       ...         ...         ...\n",
       "990  0.371002  0.696937           0           2\n",
       "991  0.358171  0.089578           0           0\n",
       "995  0.534997  0.669115           0           2\n",
       "997  0.763978  0.627393           0           2\n",
       "999  0.925459  0.374928           0           1\n",
       "\n",
       "[542 rows x 4 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subset_dict = cohort_set.transform(X)\n",
    "subset_dict['cohort_0']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can then call the ``predict()`` method. Since the predictions will have the same format regardless if the cohorts are compatible or not, then the ``predict()`` and ``predict_proba()`` will always return a single prediction array by default. However, this behavior can be changed through the ``split_pred`` parameter for both of these methods: if ``split_pred = False``, then a single prediction array is returned (this is the default behavior), but if ``split_pred = True``, then a dictionary of predictions is returned instead."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1000,)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_arr = cohort_set.predict(X)\n",
    "pred_arr.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cohort_0: (542,)\n",
      "cohort_1: (341,)\n",
      "cohort_2: (117,)\n"
     ]
    }
   ],
   "source": [
    "pred_dict = cohort_set.predict(X, split_pred=True)\n",
    "for cohort_name in pred_dict.keys():\n",
    "    print(f\"{cohort_name}: {pred_dict[cohort_name].shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Similarly, we can also call the ``predict_proba()`` method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1000, 2)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_arr = cohort_set.predict_proba(X)\n",
    "pred_arr.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that if none of the transformers in the pipeline result in incompatible cohorts, then the `predict()` method will return a single array of predictions, just like a regular estimator. To demonstrate this, let's remove the categorical columns of the dataset (so that the estimator works without an encoder). Let's also redefine our cohorts so that we don't reference any of the categorical columns removed.\n",
    "\n",
    "When calling the ``transform()`` method in this scenario, it will return a single dataset, as previously shown."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mmendonca/anaconda3/envs/raipub/lib/python3.9/site-packages/xgboost/sklearn.py:1421: UserWarning: `use_label_encoder` is deprecated in 1.7.0.\n",
      "  warnings.warn(\"`use_label_encoder` is deprecated in 1.7.0.\")\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_0</th>\n",
       "      <th>num_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.028385</td>\n",
       "      <td>0.891836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.072545</td>\n",
       "      <td>0.048710</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.243165</td>\n",
       "      <td>0.623450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.741629</td>\n",
       "      <td>0.300489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.071756</td>\n",
       "      <td>0.757902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>0.345633</td>\n",
       "      <td>0.697749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>0.971967</td>\n",
       "      <td>0.781923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>0.667862</td>\n",
       "      <td>0.654704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>0.806418</td>\n",
       "      <td>0.792162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>0.895103</td>\n",
       "      <td>0.394234</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        num_0     num_1\n",
       "0    0.028385  0.891836\n",
       "1    0.072545  0.048710\n",
       "2    0.243165  0.623450\n",
       "3    0.741629  0.300489\n",
       "4    0.071756  0.757902\n",
       "..        ...       ...\n",
       "995  0.345633  0.697749\n",
       "996  0.971967  0.781923\n",
       "997  0.667862  0.654704\n",
       "998  0.806418  0.792162\n",
       "999  0.895103  0.394234\n",
       "\n",
       "[1000 rows x 2 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_num = X.drop(columns=['CN_0_num_0', 'CN_1_num_1'])\n",
    "\n",
    "cohort_pipeline = [\n",
    "    dp.BasicImputer(verbose=False),\n",
    "    dp.DataMinMaxScaler(verbose=False),\n",
    "    get_model()\n",
    "]\n",
    "\n",
    "c1 = [ ['num_0', '>', 1.5] ]\n",
    "c2 = None\n",
    "\n",
    "cohort_set = CohortManager(\n",
    "    transform_pipe=cohort_pipeline,\n",
    "    cohort_def=[c1, c2]\n",
    ")\n",
    "cohort_set.fit(X=X_num, y=y)\n",
    "new_X = cohort_set.transform(X_num)\n",
    "new_X"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The same applies to the ``predict()`` method, where it will return a single array of predictions this time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1000,)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred = cohort_set.predict(X_num)\n",
    "pred.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Getting a list of subsets\n",
    "\n",
    "Sometimes we just want to break a new dataset into subsets, where we have one subset for each of the cohorts being used. We can do this using the ``get_subsets()`` method from the ``CohortManager`` class. This method accepts three parameters:\n",
    "\n",
    "1. A dataset that has at least the columns used by the cohorts' filters (this means that the dataset may also have other columns not used by the filters);\n",
    "2. A dataset containing only the label column (`y` dataset). This parameter is optional, and it is useful when we want to filter a feature dataset (`X`) and a label dataset (`y`), and get a list of subsets from ``X`` and ``y``;\n",
    "3. A boolean value indicating if we want to apply the transformations pipeline used for each cohort or not. If set to True, this method will behave similarly to the `transform()` method, with the main difference being that this method will always return a list of subsets, even if the cohorts are compatible with each other.\n",
    "\n",
    "The returned variable is a dictionary where the primary keys are the name of the cohorts, and the secondary keys are:\n",
    "\n",
    "* `X`: the subset of the features dataset;\n",
    "* `y`: the subset of the label dataset. This key will only be returned if the `y` dataset is passed in the method's call.\n",
    "\n",
    "The following cell shows an example of how to call this method without applying the transformations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_0</th>\n",
       "      <th>num_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.601034</td>\n",
       "      <td>2.535353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.755945</td>\n",
       "      <td>-2.172352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.354479</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.103090</td>\n",
       "      <td>-0.766515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.753178</td>\n",
       "      <td>1.787514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>990</th>\n",
       "      <td>1.904364</td>\n",
       "      <td>1.611914</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>991</th>\n",
       "      <td>1.841023</td>\n",
       "      <td>-1.886888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>2.713939</td>\n",
       "      <td>1.451639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>3.844315</td>\n",
       "      <td>1.211294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>4.641478</td>\n",
       "      <td>-0.243075</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>637 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        num_0     num_1\n",
       "0    1.601034  2.535353\n",
       "1    1.755945 -2.172352\n",
       "2    2.354479       NaN\n",
       "3    4.103090 -0.766515\n",
       "4    1.753178  1.787514\n",
       "..        ...       ...\n",
       "990  1.904364  1.611914\n",
       "991  1.841023 -1.886888\n",
       "995  2.713939  1.451639\n",
       "997  3.844315  1.211294\n",
       "999  4.641478 -0.243075\n",
       "\n",
       "[637 rows x 2 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subsets = cohort_set.get_subsets(X_num, apply_transform=False)\n",
    "subsets['cohort_0']['X']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['cohort_0', 'cohort_1'])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subsets.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's now compare the resulting subsets when setting the ``apply_transform`` to True:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_0</th>\n",
       "      <th>num_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.028385</td>\n",
       "      <td>0.891836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.072545</td>\n",
       "      <td>0.048710</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.243165</td>\n",
       "      <td>0.623450</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.741629</td>\n",
       "      <td>0.300489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.071756</td>\n",
       "      <td>0.757902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>990</th>\n",
       "      <td>0.114854</td>\n",
       "      <td>0.726453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>991</th>\n",
       "      <td>0.096797</td>\n",
       "      <td>0.099836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>0.345633</td>\n",
       "      <td>0.697749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>0.667862</td>\n",
       "      <td>0.654704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>0.895103</td>\n",
       "      <td>0.394234</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>637 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        num_0     num_1\n",
       "0    0.028385  0.891836\n",
       "1    0.072545  0.048710\n",
       "2    0.243165  0.623450\n",
       "3    0.741629  0.300489\n",
       "4    0.071756  0.757902\n",
       "..        ...       ...\n",
       "990  0.114854  0.726453\n",
       "991  0.096797  0.099836\n",
       "995  0.345633  0.697749\n",
       "997  0.667862  0.654704\n",
       "999  0.895103  0.394234\n",
       "\n",
       "[637 rows x 2 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subsets = cohort_set.get_subsets(X_num, apply_transform=True)\n",
    "subsets['cohort_0']['X']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['X'])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subsets['cohort_0'].keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the previous cell, we can see that we only have the `X` key for each cohort, since we didn't pass the `y` dataset as a parameter to the method. Let's now look at an example where we want the subsets of both the features and labels (X and y):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['X', 'y'])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subsets = cohort_set.get_subsets(X_num, y, apply_transform=True)\n",
    "subsets['cohort_0'].keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "cohort_0\n",
      "(637, 2)\n",
      "(637, 1)\n",
      "\n",
      "cohort_1\n",
      "(363, 2)\n",
      "(363, 1)\n"
     ]
    }
   ],
   "source": [
    "for key in subsets.keys():\n",
    "    print(f\"\\n{key}\")\n",
    "    print(subsets[key][\"X\"].shape)\n",
    "    print(subsets[key][\"y\"].shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Naming the cohorts\n",
    "\n",
    "So far, we created a list of conditions for each cohort and passed that to the ``CohortManager`` without specifying the cohort's name. In that case, the cohorts will be named automatically. This name is later used in ​the subsets dictionary returned by the ``get_subsets()`` method, as we've shown in the previous subsection. We can provide a specific name to each cohort by passing a dictionary of condition lists to the ``CohortManager`` instead of passing a list of conditions. The following cell demonstrates this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['Large num_0', 'Remaining'])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohort_set = CohortManager(\n",
    "    transform_pipe=cohort_pipeline,\n",
    "    cohort_def={\"Large num_0\":c1, \"Remaining\":c2}\n",
    ")\n",
    "cohort_set.fit(X_num, y)\n",
    "subsets = cohort_set.get_subsets(X_num, y, apply_transform=True)\n",
    "subsets.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Handling datasets without column names\n",
    "\n",
    "When handling a dataset without column names, we can use the column indices as the column identifier in the conditions list. However, the indices must be passed as a string, not a number. In the following cell, we demonstrate an example where we create a set of conditions using the column indices. Note that this approach will only work if the dataset doesn't have any column names. If the dataset has valid column names, then these names should be used instead of the indices.\n",
    "\n",
    "Note that in the following example, we are not passing a list of transformations to the ``transform_pipe`` parameter. Instead, we are passing a single transformer. This is also allowed, and what will happen in the background is that an empty list will be created for each cohort, and this transformer will be copied and added to each of these lists."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.616278</td>\n",
       "      <td>0.790075</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.340936</td>\n",
       "      <td>0.040024</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.910331</td>\n",
       "      <td>NaN</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.816398</td>\n",
       "      <td>0.284064</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.675657</td>\n",
       "      <td>0.688320</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>0.534997</td>\n",
       "      <td>0.669115</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>0.545942</td>\n",
       "      <td>0.790112</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>0.763978</td>\n",
       "      <td>0.627393</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>0.371610</td>\n",
       "      <td>0.801491</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>0.925459</td>\n",
       "      <td>0.374928</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            0         1       2       3\n",
       "0    0.616278  0.790075  val0_0  val1_3\n",
       "1    0.340936  0.040024  val0_1  val1_0\n",
       "2    0.910331       NaN  val0_0  val1_0\n",
       "3    0.816398  0.284064  val0_1  val1_2\n",
       "4    0.675657  0.688320  val0_0  val1_1\n",
       "..        ...       ...     ...     ...\n",
       "995  0.534997  0.669115  val0_1  val1_2\n",
       "996  0.545942  0.790112  val0_0  val1_3\n",
       "997  0.763978  0.627393  val0_1  val1_2\n",
       "998  0.371610  0.801491  val0_0  val1_3\n",
       "999  0.925459  0.374928  val0_1  val1_1\n",
       "\n",
       "[1000 rows x 4 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_X = X.copy()\n",
    "new_y = y.copy()\n",
    "new_X.columns = [i for i in range(new_X.shape[1])]\n",
    "new_y.columns = [i for i in range(new_y.shape[1])]\n",
    "\n",
    "c1 = [ ['2', '==', 'val0_1'], 'and', ['0', '>', 0.0] ]\n",
    "c2 = [ ['2', '==', 'val0_0'], 'and', ['0', '>', 0.0] ]\n",
    "c3 = None\n",
    "\n",
    "cohort_set = CohortManager(\n",
    "    transform_pipe=dp.DataMinMaxScaler(verbose=False),\n",
    "    cohort_def=[c1, c2, c3]\n",
    ")\n",
    "cohort_set.fit(X=new_X, y=new_y)\n",
    "new_X = cohort_set.transform(new_X)\n",
    "new_X"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using CohortManager with Sci-Kit Learn's Pipeline\n",
    "\n",
    "`scikit-learn`'s `Pipeline` class is a well-established approach for creating pipelines. Therefore, the ``CohortManager`` class was designed to work with ``Pipeline``, allowing users to create even more complex data processing pipelines.\n",
    "\n",
    "The following cell shows an example of how to create a ``Pipeline`` using an instance of the ``CohortManager`` class. The pipeline created in this example will do the following:\n",
    "\n",
    "1. Run the ``BasicImputer`` and ``DataMinMaxScaler`` (in that order) over each cohort separately. Notice that each cohort will have a separate object from these classes. Since these transformations result in compatible cohorts, the ``transform()`` method will return a single dataset;\n",
    "2. Run the ``EncoderOrdinal`` over the dataset returned by the ``CohortManager`` (that is, the encoding is done over the entire dataset, not over each cohort individually);\n",
    "3. Fit the model (an XGBoost in this case) over the dataset processed by the ``CohortManager`` and ``EncoderOrdinal``."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mmendonca/anaconda3/envs/raipub/lib/python3.9/site-packages/xgboost/sklearn.py:1421: UserWarning: `use_label_encoder` is deprecated in 1.7.0.\n",
      "  warnings.warn(\"`use_label_encoder` is deprecated in 1.7.0.\")\n"
     ]
    }
   ],
   "source": [
    "cohort_pipeline = [\n",
    "    dp.BasicImputer(verbose=False),\n",
    "    dp.DataMinMaxScaler(verbose=False),\n",
    "]\n",
    "\n",
    "c1 = [ ['CN_0_num_0', '==', 'val0_1'], 'and', ['num_0', '>', 0.0] ]\n",
    "c2 = [ ['CN_0_num_0', '==', 'val0_0'], 'and', ['num_0', '>', 0.0] ]\n",
    "c3 = None\n",
    "\n",
    "cohort_set = CohortManager(\n",
    "    transform_pipe=cohort_pipeline,\n",
    "    cohort_def=[c1, c2, c3]\n",
    ")\n",
    "\n",
    "skpipe = Pipeline([\n",
    "    (\"cohort_preprocess\", cohort_set),\n",
    "    (\"encoder\", dp.EncoderOrdinal(verbose=False)),\n",
    "    (\"model\", get_model())\n",
    "])\n",
    "skpipe.fit(X, y)\n",
    "pred = skpipe.predict_proba(X)\n",
    "pred = skpipe.predict(X)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fetching the queries used for each cohort\n",
    "\n",
    "The ``get_queries()`` method returns a dictionary with the queries used for each cohort."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cohort_0': '(`CN_0_num_0` == \"val0_1\") and (`num_0` > 0.0)',\n",
       " 'cohort_1': '(`CN_0_num_0` == \"val0_0\") and (`num_0` > 0.0)',\n",
       " 'cohort_2': '((`CN_0_num_0` != \"val0_1\") or ((`num_0` <= 0.0) or (`num_0`.isnull()))) and ((`CN_0_num_0` != \"val0_0\") or ((`num_0` <= 0.0) or (`num_0`.isnull())))'}"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "queries = cohort_set.get_queries()\n",
    "queries"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating cohorts based on a set of columns\n",
    "\n",
    "In some cases, users might be interested in creating one cohort for each possible value in a given column (for example, if a dataset has a **gender** column, we might be interested in creating one cohort for each gender). To avoid the cumbersome task of creating one condition list for each value in a given column, we ``CohortManager`` use the ``cohort_col`` parameter instead of the ``cohort_def`` during the instantiation of the object. Here are more details about these two parameters:\n",
    "\n",
    "* ``cohort_def``: already explored in the previous subsections. Accepts a list of condition lists (one for each cohort), or a dictionary of condition lists;\n",
    "* ``cohort_col``: accepts a list of column names or indices, from which one cohort is created for each unique combination of values for these columns.\n",
    "\n",
    "In the following example, we show how to create a set of cohorts based on the ``CN_0_num_0`` column. We then use the ``get_queries()`` method to look at the cohorts created."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cohort_0': '(`CN_0_num_0` == \"val0_0\")',\n",
       " 'cohort_1': '(`CN_0_num_0` == \"val0_1\")'}"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohort_set = CohortManager(\n",
    "    transform_pipe=dp.DataMinMaxScaler(verbose=False),\n",
    "    cohort_col=[\"CN_0_num_0\"]\n",
    ")\n",
    "cohort_set.fit(X, y)\n",
    "queries = cohort_set.get_queries()\n",
    "queries"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also use the column indices instead of their names:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cohort_0': '(`CN_0_num_0` == \"val0_0\")',\n",
       " 'cohort_1': '(`CN_0_num_0` == \"val0_1\")'}"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohort_set = CohortManager(\n",
    "    transform_pipe=dp.DataMinMaxScaler(verbose=False),\n",
    "    cohort_col=[2]\n",
    ")\n",
    "cohort_set.fit(X, y)\n",
    "queries = cohort_set.get_queries()\n",
    "queries"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that in both examples we had to call the ``fit()`` method before calling the ``get_queries()`` method. This is necessary because, when using the ``cohort_col`` parameter instead of the ``cohort_def``, the cohorts can only be created when a valid dataset is provided. Since we didn't provide a dataset in the constructor method, then the cohorts are only built when the fit method is called. In the following cell we provide the dataset in the constructor method just to show this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cohort_0': '(`CN_0_num_0` == \"val0_0\")',\n",
       " 'cohort_1': '(`CN_0_num_0` == \"val0_1\")'}"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohort_set = CohortManager(\n",
    "    transform_pipe=dp.DataMinMaxScaler(verbose=False),\n",
    "    cohort_col=[2],\n",
    "    X=X, y=y\n",
    ")\n",
    "queries = cohort_set.get_queries()\n",
    "queries"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Just to better demonstrate this, let's now use the ``cohort_def`` parameter, while not providing the datasets in the constructor method. In this case, we can also call the ``get_queries()`` method before calling the ``fit()`` method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cohort_0': '(`CN_0_num_0` == \"val0_1\") and (`num_0` > 0.0)',\n",
       " 'cohort_1': '(`CN_0_num_0` == \"val0_0\") and (`num_0` > 0.0)',\n",
       " 'cohort_2': '((`CN_0_num_0` != \"val0_1\") or ((`num_0` <= 0.0) or (`num_0`.isnull()))) and ((`CN_0_num_0` != \"val0_0\") or ((`num_0` <= 0.0) or (`num_0`.isnull())))'}"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohort_set = CohortManager(\n",
    "    transform_pipe=dp.DataMinMaxScaler(verbose=False),\n",
    "    cohort_def=[c1, c2, c3]\n",
    ")\n",
    "queries = cohort_set.get_queries()\n",
    "queries"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As previously mentioned, the ``cohort_col`` parameter accepts a list of column names, not a single column. When more than one column is provided, we create one cohort for each combination of unique values in these columns:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cohort_0': '(`CN_0_num_0` == \"val0_0\") and (`CN_1_num_1` == \"val1_0\")',\n",
       " 'cohort_1': '(`CN_0_num_0` == \"val0_0\") and (`CN_1_num_1` == \"val1_1\")',\n",
       " 'cohort_2': '(`CN_0_num_0` == \"val0_0\") and (`CN_1_num_1` == \"val1_2\")',\n",
       " 'cohort_3': '(`CN_0_num_0` == \"val0_0\") and (`CN_1_num_1` == \"val1_3\")',\n",
       " 'cohort_4': '(`CN_0_num_0` == \"val0_1\") and (`CN_1_num_1` == \"val1_0\")',\n",
       " 'cohort_5': '(`CN_0_num_0` == \"val0_1\") and (`CN_1_num_1` == \"val1_1\")',\n",
       " 'cohort_6': '(`CN_0_num_0` == \"val0_1\") and (`CN_1_num_1` == \"val1_2\")',\n",
       " 'cohort_7': '(`CN_0_num_0` == \"val0_1\") and (`CN_1_num_1` == \"val1_3\")'}"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohort_set = CohortManager(\n",
    "    transform_pipe=dp.DataMinMaxScaler(verbose=False),\n",
    "    cohort_col=[\"CN_0_num_0\", \"CN_1_num_1\"]\n",
    ")\n",
    "cohort_set.fit(X=X, y=y)\n",
    "queries = cohort_set.get_queries()\n",
    "queries"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Saving and loading the conditions used for the cohorts\n",
    "\n",
    "The ``save_cohorts()`` method saves the list of conditions used by all cohorts into a JSON file. This way, users are able to then reuse these condition lists in another ``CohortManager`` object. The ``save_cohorts()`` has the same conditions as the ``get_queries()`` method: if using ``cohort_col``, then we must either pass the datasets in the constructor method, or call the ``fit()`` method prior to calling the ``save_cohorts()`` method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "cohort_set = CohortManager(\n",
    "    transform_pipe=dp.DataMinMaxScaler(verbose=False),\n",
    "    cohort_def=[c1, c2, c3]\n",
    ")\n",
    "cohort_set.fit(X=X, y=y)\n",
    "\n",
    "json_files = [\n",
    "    \"json_files/CohortManager_tutorial/cht_0.json\",\n",
    "    \"json_files/CohortManager_tutorial/cht_1.json\",\n",
    "    \"json_files/CohortManager_tutorial/cht_2.json\",\n",
    "]\n",
    "cohort_set.save_cohorts(json_files)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The JSON file saved has the following structure (note that it follows the same JSON structure used by the [raiwidgets library](https://github.com/microsoft/responsible-ai-toolbox/blob/main/notebooks/responsibleaidashboard/responsibleaidashboard-census-classification-model-debugging.ipynb)):\n",
    "\n",
    "![cohort](./imgs/cohort.png)\n",
    "\n",
    "We can then create a new ``CohortManager`` object and pass the list of paths to the JSON files containing the cohort definitions to the ``cohort_json_files`` parameter. After loading the cohort files, we can verify that the same conditions are being used."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cohort_0': \"((`CN_0_num_0` in ['val0_1']) and (`num_0` > 0.0))\",\n",
       " 'cohort_1': \"((`CN_0_num_0` in ['val0_0']) and (`num_0` > 0.0))\",\n",
       " 'cohort_2': \"(((`CN_0_num_0` not in ['val0_1']) or ((`num_0` <= 0.0) or (`num_0`.isnull()))) and ((`CN_0_num_0` not in ['val0_0']) or ((`num_0` <= 0.0) or (`num_0`.isnull()))))\"}"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cohort_set = CohortManager(\n",
    "    transform_pipe=dp.BasicImputer(verbose=False),\n",
    "    cohort_json_files=json_files\n",
    ")\n",
    "queries = cohort_set.get_queries()\n",
    "queries"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you have the file for only a few of the cohorts that needs to be created, but don't want to create a file for the cohort with the remaining instances, you can proceed similarly to what we do with the ``cohort_def`` parameter: the last json file passed through the ``cohort_json_files`` can be a ``None`` value, which indicates that that cohort should include all instances that doesn't belong to any of the other cohorts.\n",
    "\n",
    "To demonstrate this, suppose that we only have the json files for ``cohort_0`` and ``cohort_1``. To load these files using the ``CohortManager``, we should do the following:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cohort_0': \"((`CN_0_num_0` in ['val0_1']) and (`num_0` > 0.0))\",\n",
       " 'cohort_1': \"((`CN_0_num_0` in ['val0_0']) and (`num_0` > 0.0))\",\n",
       " 'Remaining Instances': \"(((`CN_0_num_0` not in ['val0_1']) or ((`num_0` <= 0.0) or (`num_0`.isnull())))) and (((`CN_0_num_0` not in ['val0_0']) or ((`num_0` <= 0.0) or (`num_0`.isnull()))))\"}"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "json_files = [\n",
    "    \"json_files/CohortManager_tutorial/cht_0.json\",\n",
    "    \"json_files/CohortManager_tutorial/cht_1.json\",\n",
    "    None\n",
    "]\n",
    "cohort_set = CohortManager(\n",
    "    transform_pipe=dp.BasicImputer(verbose=False),\n",
    "    cohort_json_files=json_files\n",
    ")\n",
    "queries = cohort_set.get_queries()\n",
    "queries"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using different pipelines for each cohort\n",
    "\n",
    "So far, every time we created an instance of the ``CohortMManager`` class, the same pipeline was used over all cohorts (different instances of the same pipeline, given that the ``fit()`` must be called using only the subset associated to a given cohort). However, there might be some scenarios where we might be interested in using different transformations for each cohort. We won't cover here these scenarios. Instead, here we'll show how to create different pipelines for each cohort.\n",
    "\n",
    "There are four scenarios associated with the ``transform_pipe`` parameter:\n",
    "\n",
    "1. **An empty list or ``None`` is provided:** in this case, the ``CohortManager`` won't apply any transformations over the dataset. The ``transform()`` method will simply return the dataset provided;\n",
    "2. **A single transformer is provided:** in this case, this single transformer is placed in a list (a list with a single transformer), which is then replicated such that each cohort has its own list of transformations (pipeline);\n",
    "3. **A list of transformers is provided:** in this case, this pipeline is replicated for each cohort;\n",
    "4. **A list of pipelines is provided:** a list of pipelines is basically a list of lists of transformations. In this case, the list of pipelines should have one pipeline for each cohort created, that is, the length of the ``transform_pipe`` parameter should be the same as the number of cohorts created. The pipelines will be assigned to each cohort following the same order as the ``cohort_def`` parameter (depicted in the following example).\n",
    "\n",
    "In the following example, we'll create 3 cohorts and 3 pipelines: the first pipeline will contain only a ``DataMinMaxScaler`` transformer, the second pipeline contains only a ``DataPowerTransformer`` transformer, and the third pipeline is empty, that is, it doesn't apply any transformations over the data. These pipelines will be associated with cohorts ``c1``, ``c2``, and ``c3`` respectively."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_0</th>\n",
       "      <th>num_1</th>\n",
       "      <th>CN_0_num_0</th>\n",
       "      <th>CN_1_num_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.601034</td>\n",
       "      <td>2.535353</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.755945</td>\n",
       "      <td>-2.172352</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.354479</td>\n",
       "      <td>NaN</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.103090</td>\n",
       "      <td>-0.766515</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.753178</td>\n",
       "      <td>1.787514</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>2.713939</td>\n",
       "      <td>1.451639</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>1.420812</td>\n",
       "      <td>2.535627</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>3.844315</td>\n",
       "      <td>1.211294</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>0.974124</td>\n",
       "      <td>2.619258</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>4.641478</td>\n",
       "      <td>-0.243075</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        num_0     num_1 CN_0_num_0 CN_1_num_1\n",
       "0    1.601034  2.535353     val0_0     val1_3\n",
       "1    1.755945 -2.172352     val0_1     val1_0\n",
       "2    2.354479       NaN     val0_0     val1_0\n",
       "3    4.103090 -0.766515     val0_1     val1_2\n",
       "4    1.753178  1.787514     val0_0     val1_1\n",
       "..        ...       ...        ...        ...\n",
       "995  2.713939  1.451639     val0_1     val1_2\n",
       "996  1.420812  2.535627     val0_0     val1_3\n",
       "997  3.844315  1.211294     val0_1     val1_2\n",
       "998  0.974124  2.619258     val0_0     val1_3\n",
       "999  4.641478 -0.243075     val0_1     val1_1\n",
       "\n",
       "[1000 rows x 4 columns]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>num_0</th>\n",
       "      <th>num_1</th>\n",
       "      <th>CN_0_num_0</th>\n",
       "      <th>CN_1_num_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.560649</td>\n",
       "      <td>0.730360</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.369841</td>\n",
       "      <td>0.231139</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.089526</td>\n",
       "      <td>NaN</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.103090</td>\n",
       "      <td>-0.766515</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.753178</td>\n",
       "      <td>1.787514</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>2.713939</td>\n",
       "      <td>1.451639</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>0.524491</td>\n",
       "      <td>0.730402</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>3.844315</td>\n",
       "      <td>1.211294</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>0.434872</td>\n",
       "      <td>0.743060</td>\n",
       "      <td>val0_0</td>\n",
       "      <td>val1_3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>4.641478</td>\n",
       "      <td>-0.243075</td>\n",
       "      <td>val0_1</td>\n",
       "      <td>val1_1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        num_0     num_1 CN_0_num_0 CN_1_num_1\n",
       "0    0.560649  0.730360     val0_0     val1_3\n",
       "1    0.369841  0.231139     val0_1     val1_0\n",
       "2    1.089526       NaN     val0_0     val1_0\n",
       "3    4.103090 -0.766515     val0_1     val1_2\n",
       "4    1.753178  1.787514     val0_0     val1_1\n",
       "..        ...       ...        ...        ...\n",
       "995  2.713939  1.451639     val0_1     val1_2\n",
       "996  0.524491  0.730402     val0_0     val1_3\n",
       "997  3.844315  1.211294     val0_1     val1_2\n",
       "998  0.434872  0.743060     val0_0     val1_3\n",
       "999  4.641478 -0.243075     val0_1     val1_1\n",
       "\n",
       "[1000 rows x 4 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X, y = create_df()\n",
    "\n",
    "c1 = [ ['CN_1_num_1', '==', 'val1_3'] ]\n",
    "c2 = [ ['CN_1_num_1', '==', 'val1_0'] ]\n",
    "c3 = None\n",
    "\n",
    "c1_pipe = [dp.DataMinMaxScaler(verbose=False)]\n",
    "c2_pipe = [dp.DataPowerTransformer(verbose=False)]\n",
    "c3_pipe = []\n",
    "\n",
    "cohort_set = CohortManager(\n",
    "    transform_pipe=[c1_pipe, c2_pipe, c3_pipe],\n",
    "    cohort_def=[c1, c2, c3]\n",
    ")\n",
    "cohort_set.fit(X=X, y=y)\n",
    "new_X = cohort_set.transform(X)\n",
    "new_X"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Use Rebalance over each cohort separately\n",
    "\n",
    "Rebalancing a dataset is not always a trivial task. In some cases, we might have a class imbalance in the whole dataset, but different cohorts of this dataset depict different imbalance distributions. In such a scenario, rebalancing the full dataset might result in an even greater imbalance for a set of cohorts, even though the full dataset is now balanced. Therefore, we might want to rebalance each cohort separately, in order to obtain a balanced dataset and balanced cohorts.\n",
    "\n",
    "In this subsection, we'll explore how the ``CohortManager`` can easily help us achieve this goal."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll create a ``CohortManager`` just to obtain the cohorts of interest and plot their distributions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "label\n",
      "1        795\n",
      "0        205\n",
      "dtype: int64\n",
      "label\n",
      "0        119\n",
      "1         11\n",
      "dtype: int64\n",
      "label\n",
      "0        74\n",
      "1        31\n",
      "dtype: int64\n",
      "label\n",
      "1        495\n",
      "0          6\n",
      "dtype: int64\n",
      "label\n",
      "1        258\n",
      "0          6\n",
      "dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X, y = create_df(with_null=False)\n",
    "\n",
    "cohort_set = CohortManager(\n",
    "    cohort_col=[\"CN_1_num_1\"]\n",
    ")\n",
    "cohort_set.fit(X=X, y=y)\n",
    "subsets = cohort_set.get_subsets(X, y, apply_transform=False)\n",
    "\n",
    "print(y.value_counts())\n",
    "for key in subsets.keys():\n",
    "    print(subsets[key][\"y\"].value_counts())\n",
    "\n",
    "plot_value_counts_cohort(y, subsets, normalize=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the full dataset has a great imbalance, where we have a considerably larger number of occurrences of the ``1`` class in relation to the ``0`` class. This imbalance is even greater for cohorts ``cohort_2`` and ``cohort_3``, while for cohorts ``cohort_0`` and ``cohort_1`` this imbalance is inverted: there are a lot more instances from the ``0`` class when compared to the ``1`` class.\n",
    "\n",
    "### Balancing the full dataset\n",
    "\n",
    "Let's now rebalance the full dataset. Here, We use the ``Rebalance`` class from the ``dataprocessing`` module. Afterward, we plot the label distributions once more to see if our problem is solved."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mmendonca/anaconda3/envs/raipub/lib/python3.9/site-packages/sklearn/preprocessing/_encoders.py:808: FutureWarning: `sparse` was renamed to `sparse_output` in version 1.2 and will be removed in 1.4. `sparse_output` is ignored unless you leave `sparse` to its default value.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1    795\n",
      "0    795\n",
      "Name: label, dtype: int64\n",
      "0    463\n",
      "1     11\n",
      "Name: label, dtype: int64\n",
      "0    313\n",
      "1     31\n",
      "Name: label, dtype: int64\n",
      "1    495\n",
      "0      8\n",
      "Name: label, dtype: int64\n",
      "1    258\n",
      "0     11\n",
      "Name: label, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rebalance = dp.Rebalance(verbose=False)\n",
    "new_X, new_y = rebalance.fit_resample(X, y)\n",
    "\n",
    "cohort_set.fit(X=new_X, y=new_y)\n",
    "subsets = cohort_set.get_subsets(new_X, new_y, apply_transform=False)\n",
    "\n",
    "print(new_y.value_counts())\n",
    "for key in subsets.keys():\n",
    "    print(subsets[key][\"y\"].value_counts())\n",
    "\n",
    "plot_value_counts_cohort(new_y, subsets, normalize=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see, the full dataset is now balanced. However, looking at each cohort separately, we noticed that we only made matters worse: we now have an even greater label imbalance inside each cohort. If we think of these cohorts as sensitive groups, that is, groups separated by sensitive features, *e.g.* gender, race, and nationality, and that we don't want our models to perform differently in one group compared to another, then this imbalance between cohorts is a scenario that needs to be fixed.\n",
    "\n",
    "### Balancing each cohort separately\n",
    "\n",
    "We'll now use the ``Rebalance`` class over each cohort separately, in order to obtain a set of balanced cohorts. This might aid us in obtaining fair models, that is, models that perform equally through cohorts. To do this, we'll use the ``CohortManager``, and we'll pass the ``Rebalance`` object to the ``transform_pipe`` parameter. We'll then call the ``fit_resample()`` method, implemented by the ``Rebalance`` class. Note that when using a pipeline with an object that implements the ``fit_resample()`` class, there can't be any transformers that implement the ``transform()`` and ``predict()`` methods, since these transformations serve different purposes: the former is used for rebalancing, that is, creating new instances in the ``X`` and ``y`` datasets, while the latter require that the number of instances of the dataset is kept fixed between each transformation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mmendonca/anaconda3/envs/raipub/lib/python3.9/site-packages/sklearn/preprocessing/_encoders.py:808: FutureWarning: `sparse` was renamed to `sparse_output` in version 1.2 and will be removed in 1.4. `sparse_output` is ignored unless you leave `sparse` to its default value.\n",
      "  warnings.warn(\n",
      "/home/mmendonca/anaconda3/envs/raipub/lib/python3.9/site-packages/sklearn/preprocessing/_encoders.py:808: FutureWarning: `sparse` was renamed to `sparse_output` in version 1.2 and will be removed in 1.4. `sparse_output` is ignored unless you leave `sparse` to its default value.\n",
      "  warnings.warn(\n",
      "/home/mmendonca/anaconda3/envs/raipub/lib/python3.9/site-packages/sklearn/preprocessing/_encoders.py:808: FutureWarning: `sparse` was renamed to `sparse_output` in version 1.2 and will be removed in 1.4. `sparse_output` is ignored unless you leave `sparse` to its default value.\n",
      "  warnings.warn(\n",
      "/home/mmendonca/anaconda3/envs/raipub/lib/python3.9/site-packages/sklearn/preprocessing/_encoders.py:808: FutureWarning: `sparse` was renamed to `sparse_output` in version 1.2 and will be removed in 1.4. `sparse_output` is ignored unless you leave `sparse` to its default value.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rebalance_cohort = CohortManager(\n",
    "    transform_pipe=dp.Rebalance(verbose=False),\n",
    "    cohort_col=[\"CN_1_num_1\"]\n",
    ")\n",
    "new_X, new_y = rebalance_cohort.fit_resample(X, y)\n",
    "\n",
    "subsets = rebalance_cohort.get_subsets(new_X, new_y, apply_transform=False)\n",
    "\n",
    "plot_value_counts_cohort(new_y, subsets, normalize=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    946\n",
       "1    946\n",
       "Name: label, dtype: int64"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_y.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see, we now have a balanced full dataset, while also having balanced cohorts.\n",
    "\n",
    "## Balancing only a set of cohorts\n",
    "\n",
    "Suppose that we trained a model that is under-performing only for cohorts ``cohort_2`` and ``cohort_3``. In this case, we can keep the remaining cohorts unchanged, and balance only cohorts ``cohort_2`` and ``cohort_3``. To do this, we can use a separate pipeline for each cohort, where cohorts ``cohort_2`` and ``cohort_3`` use a pipeline comprised with only a ``Rebalance`` object, while the other two cohorts use an empty pipeline."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/mmendonca/anaconda3/envs/raipub/lib/python3.9/site-packages/sklearn/preprocessing/_encoders.py:808: FutureWarning: `sparse` was renamed to `sparse_output` in version 1.2 and will be removed in 1.4. `sparse_output` is ignored unless you leave `sparse` to its default value.\n",
      "  warnings.warn(\n",
      "/home/mmendonca/anaconda3/envs/raipub/lib/python3.9/site-packages/sklearn/preprocessing/_encoders.py:808: FutureWarning: `sparse` was renamed to `sparse_output` in version 1.2 and will be removed in 1.4. `sparse_output` is ignored unless you leave `sparse` to its default value.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1800x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rebalance_cohort = CohortManager(\n",
    "    transform_pipe=[\n",
    "        [],\n",
    "        [],\n",
    "        [dp.Rebalance(verbose=False)],\n",
    "        [dp.Rebalance(verbose=False)],\n",
    "    ],\n",
    "    cohort_col=[\"CN_1_num_1\"]\n",
    ")\n",
    "new_X, new_y = rebalance_cohort.fit_resample(X, y)\n",
    "\n",
    "subsets = rebalance_cohort.get_subsets(new_X, new_y, apply_transform=False)\n",
    "\n",
    "plot_value_counts_cohort(new_y, subsets, normalize=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "rai",
   "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"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "239972f3e2cce5deeefd34fe5d6dd927057b08952d864776048f8db51d1a831a"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}