multi_scenario.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "3d2eee11-c637-46c0-84a4-ee9e169d8a92",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams['figure.figsize'] = (12, 12)\n",
    "plt.style.use('tableau-colorblind10')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "027660d3-8ef3-4e15-a3e6-137fe247a645",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/markdown": [
       "# Best possible performance\n",
       "\n",
       "This is what would be obtainable if we had no data loss.\n"
      ],
      "text/plain": [
       "<IPython.core.display.Markdown object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>scenario</th>\n",
       "      <th>train_rows</th>\n",
       "      <th>test_rows</th>\n",
       "      <th>roc_auc_score</th>\n",
       "      <th>brier_score_loss</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>santander</td>\n",
       "      <td>28507</td>\n",
       "      <td>9503</td>\n",
       "      <td>0.816830</td>\n",
       "      <td>0.032872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>car_insurance</td>\n",
       "      <td>43944</td>\n",
       "      <td>14648</td>\n",
       "      <td>0.649671</td>\n",
       "      <td>0.059832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>tabular_playground</td>\n",
       "      <td>19927</td>\n",
       "      <td>6643</td>\n",
       "      <td>0.573514</td>\n",
       "      <td>0.170877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>synthetic_1</td>\n",
       "      <td>75000</td>\n",
       "      <td>25000</td>\n",
       "      <td>0.932896</td>\n",
       "      <td>0.025190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>synthetic_2_dataset_shift</td>\n",
       "      <td>75000</td>\n",
       "      <td>25000</td>\n",
       "      <td>0.886555</td>\n",
       "      <td>0.033067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>cover_type_dataset_shift</td>\n",
       "      <td>435759</td>\n",
       "      <td>145253</td>\n",
       "      <td>0.990271</td>\n",
       "      <td>0.025968</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    scenario  train_rows  test_rows  roc_auc_score  \\\n",
       "0                  santander       28507       9503       0.816830   \n",
       "1              car_insurance       43944      14648       0.649671   \n",
       "2         tabular_playground       19927       6643       0.573514   \n",
       "3                synthetic_1       75000      25000       0.932896   \n",
       "4  synthetic_2_dataset_shift       75000      25000       0.886555   \n",
       "5   cover_type_dataset_shift      435759     145253       0.990271   \n",
       "\n",
       "   brier_score_loss  \n",
       "0          0.032872  \n",
       "1          0.059832  \n",
       "2          0.170877  \n",
       "3          0.025190  \n",
       "4          0.033067  \n",
       "5          0.025968  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import display, Markdown\n",
    "from databoost.scenario import SCENARIOS\n",
    "from sklearn.metrics import roc_auc_score, brier_score_loss\n",
    "from sklearn.inspection import permutation_importance\n",
    "import warnings\n",
    "\n",
    "METRICS = [roc_auc_score, brier_score_loss]\n",
    "\n",
    "best_result = []\n",
    "importances = {}\n",
    "for scenario in SCENARIOS:\n",
    "    s = SCENARIOS[scenario]\n",
    "    train_X, test_X, train_y, test_y = s.train_test(legacy_size=0.0, legacy_drop_frac=0.0, legacy_missing_label=0.0)\n",
    "    base_pipe = s.build_pipeline(train_X)\n",
    "\n",
    "    with warnings.catch_warnings():\n",
    "        warnings.simplefilter(\"ignore\")\n",
    "        base_pipe.fit(train_X, train_y)\n",
    "        base_scored = base_pipe.predict_proba(test_X)[:,1]\n",
    "        imp = permutation_importance(base_pipe, train_X, train_y, max_samples=1000)\n",
    "        importances[scenario] = pd.DataFrame({\n",
    "            'column': list(test_X.columns),\n",
    "            'importance': np.abs(imp['importances_mean']),\n",
    "        })\n",
    "        \n",
    "    row_result = {\n",
    "        'scenario': scenario,\n",
    "        'train_rows': len(train_y),\n",
    "        'test_rows': len(test_y),\n",
    "    }\n",
    "    for metric in METRICS:\n",
    "        row_result[str(metric.__name__)] = metric(test_y, base_scored)\n",
    "\n",
    "    best_result.append(row_result)\n",
    "    \n",
    "\n",
    "\n",
    "display(Markdown(\"\"\"# Best possible performance\n",
    "\n",
    "This is what would be obtainable if we had no data loss.\n",
    "\"\"\"))\n",
    "\n",
    "best_result = pd.DataFrame(best_result)\n",
    "best_result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "28cd1784-11fd-442d-8143-39b7e1b35eff",
   "metadata": {},
   "outputs": [],
   "source": [
    "from databoost.utils import SummedPredictors"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4464787-25f4-4434-8594-66b87c4e189d",
   "metadata": {},
   "source": [
    "# Data boosting\n",
    "\n",
    "This model uses adaboost-style weights to build a second model that corrects the first one. The final model is merely the sum of the two."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "65ee1a38-63c7-4699-9234-154cf66e3003",
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.special import xlogy\n",
    "\n",
    "def adaboost_weights(estimator, X, y, learning_rate=0.5, sample_weight=None):\n",
    "    \"\"\"Implement weights for a single boost using the SAMME.R real algorithm.\"\"\"\n",
    "\n",
    "    with warnings.catch_warnings():\n",
    "        warnings.simplefilter(\"ignore\")\n",
    "        y_predict_proba = estimator.predict_proba(X)\n",
    "    if sample_weight is None:\n",
    "        sample_weight = np.ones(shape=y.shape) / len(y)\n",
    "\n",
    "    n_classes = 2\n",
    "    classes = np.array([0,1])\n",
    "    y_codes = np.array([-1.0 / (n_classes - 1), 1.0])\n",
    "    y_coding = y_codes.take(classes == y[:, np.newaxis])\n",
    "\n",
    "    # Displace zero probabilities so the log is defined.\n",
    "    # Also fix negative elements which may occur with\n",
    "    # negative sample weights.\n",
    "    proba = y_predict_proba  # alias for readability\n",
    "    np.clip(proba, np.finfo(proba.dtype).eps, None, out=proba)\n",
    "\n",
    "    # Boost weight using multi-class AdaBoost SAMME.R alg\n",
    "    estimator_weight = (\n",
    "        -1.0\n",
    "        * learning_rate\n",
    "        * ((n_classes - 1.0) / n_classes)\n",
    "        * xlogy(y_coding, y_predict_proba).sum(axis=1)\n",
    "    )\n",
    "\n",
    "    # Only boost positive weights\n",
    "    sample_weight *= np.exp(\n",
    "        estimator_weight * ((sample_weight > 0) | (estimator_weight < 0))\n",
    "    )\n",
    "\n",
    "    return sample_weight\n",
    "\n",
    "\n",
    "def databoosted_pipe(train_X, train_y, scenario):\n",
    "    old_idx = (train_X['legacy'] == 1)\n",
    "\n",
    "    pipe1 = scenario.build_pipeline(train_X[old_idx])\n",
    "\n",
    "    with warnings.catch_warnings():\n",
    "        warnings.simplefilter(\"ignore\")\n",
    "        pipe1.fit(\n",
    "            train_X[old_idx], \n",
    "            train_y[old_idx], \n",
    "        )\n",
    "\n",
    "    sample_weight = adaboost_weights(\n",
    "        pipe1, train_X[~old_idx].copy(), \n",
    "        train_y[~old_idx].copy(), \n",
    "        learning_rate=0.25,\n",
    "        sample_weight=scenario.base_sample_weight(train_X[~old_idx], train_y[~old_idx]),\n",
    "    )\n",
    "\n",
    "    pipe2 = scenario.build_pipeline(train_X[~old_idx])\n",
    "    with warnings.catch_warnings():\n",
    "        warnings.simplefilter(\"ignore\")\n",
    "        pipe2.fit(train_X[~old_idx], train_y[~old_idx], final_estimator__sample_weight=sample_weight)\n",
    "\n",
    "    boosted_pipe = SummedPredictors([pipe1, pipe2], [1.0, 1.0])\n",
    "    return boosted_pipe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0e05aba5-6ef4-4790-926d-bf76d30fa1b8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'fpr')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x1200 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_auc_score, roc_curve\n",
    "\n",
    "n_scenarios = len(SCENARIOS)\n",
    "plot_cols = 3\n",
    "plot_rows = int(np.ceil(n_scenarios/plot_cols))\n",
    "i = 1\n",
    "\n",
    "for scenario in SCENARIOS:\n",
    "    plt.subplot(plot_rows*100+plot_cols*10+i)\n",
    "    i += 1\n",
    "    \n",
    "    s = SCENARIOS[scenario]\n",
    "    train_X, test_X, train_y, test_y = s.train_test(legacy_size=0.85, legacy_drop_frac=0.4, legacy_missing_label=0.0)\n",
    "    base_pipe = s.build_pipeline(train_X)\n",
    "    new_idx = ~(train_X['legacy'].astype(bool))\n",
    "    new_only_pipe = s.build_pipeline(train_X[new_idx])\n",
    "    \n",
    "    with warnings.catch_warnings():\n",
    "        warnings.simplefilter(\"ignore\")\n",
    "        base_pipe.fit(train_X, train_y)\n",
    "        new_only_pipe.fit(train_X[new_idx], train_y[new_idx])\n",
    "        bp = databoosted_pipe(train_X, train_y, s)\n",
    "    \n",
    "    with warnings.catch_warnings():\n",
    "        warnings.simplefilter(\"ignore\")\n",
    "        base_scored = base_pipe.predict_proba(test_X)[:,1]\n",
    "        new_only_scored = new_only_pipe.predict_proba(test_X)[:,1]\n",
    "        boosted_scored = bp.predict_proba(test_X)[:,1]\n",
    "\n",
    "    fpr, recall, _ = roc_curve(test_y, base_scored)\n",
    "    plt.plot(fpr, recall, label='base, auc={:.3f}'.format(\n",
    "        roc_auc_score(test_y, base_scored)\n",
    "    ))\n",
    "    fpr, recall, _ = roc_curve(test_y, new_only_scored)\n",
    "    plt.plot(fpr, recall, label='new only, auc={:.3f}'.format(\n",
    "        roc_auc_score(test_y, new_only_scored)\n",
    "    ))\n",
    "    fpr, recall, _ = roc_curve(test_y, boosted_scored)\n",
    "    plt.plot(fpr, recall, label='boosted, auc={:.3f}'.format(\n",
    "        roc_auc_score(test_y, boosted_scored)\n",
    "    ))\n",
    "\n",
    "    plt.title(scenario)\n",
    "    plt.ylabel('recall')\n",
    "    plt.legend()\n",
    "\n",
    "plt.xlabel('fpr')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19a725d3-d07f-4225-817c-3a0a882db4b9",
   "metadata": {},
   "source": [
    "# Aggregate results\n",
    "\n",
    "Let us now measure the benefit, across all datasets, and multiple parameter choices.\n",
    "\n",
    "In what is seen below, `legacy_size` represents the fraction of the dataset that is missing features. This ranges from 10% to 95%,\n",
    "and for each value of the parameter I do 5 iterations. Each iteration represents a different (randomly chosen) set of features\n",
    "which get dropped on the legacy data set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8eac40a8-8952-4680-b808-d79b82511b19",
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "Found unknown categories ['C20'] in column 0 during transform",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[6], line 25\u001b[0m\n\u001b[1;32m     23\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m warnings\u001b[38;5;241m.\u001b[39mcatch_warnings():\n\u001b[1;32m     24\u001b[0m     warnings\u001b[38;5;241m.\u001b[39msimplefilter(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mignore\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m---> 25\u001b[0m     base_scored \u001b[38;5;241m=\u001b[39m \u001b[43mbase_pipe\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mpredict_proba\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtest_X\u001b[49m\u001b[43m)\u001b[49m[:,\u001b[38;5;241m1\u001b[39m]\n\u001b[1;32m     26\u001b[0m     new_only_scored \u001b[38;5;241m=\u001b[39m new_only_pipe\u001b[38;5;241m.\u001b[39mpredict_proba(test_X)[:,\u001b[38;5;241m1\u001b[39m]\n\u001b[1;32m     27\u001b[0m     boosted_scored \u001b[38;5;241m=\u001b[39m bp\u001b[38;5;241m.\u001b[39mpredict_proba(test_X)[:,\u001b[38;5;241m1\u001b[39m]\n",
      "File \u001b[0;32m~/virtualenvs/databoost/lib/python3.8/site-packages/sklearn/pipeline.py:523\u001b[0m, in \u001b[0;36mPipeline.predict_proba\u001b[0;34m(self, X, **predict_proba_params)\u001b[0m\n\u001b[1;32m    521\u001b[0m Xt \u001b[38;5;241m=\u001b[39m X\n\u001b[1;32m    522\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m _, name, transform \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_iter(with_final\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mFalse\u001b[39;00m):\n\u001b[0;32m--> 523\u001b[0m     Xt \u001b[38;5;241m=\u001b[39m \u001b[43mtransform\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtransform\u001b[49m\u001b[43m(\u001b[49m\u001b[43mXt\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    524\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msteps[\u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m][\u001b[38;5;241m1\u001b[39m]\u001b[38;5;241m.\u001b[39mpredict_proba(Xt, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mpredict_proba_params)\n",
      "File \u001b[0;32m~/virtualenvs/databoost/lib/python3.8/site-packages/sklearn_pandas/dataframe_mapper.py:260\u001b[0m, in \u001b[0;36mDataFrameMapper.transform\u001b[0;34m(self, X)\u001b[0m\n\u001b[1;32m    258\u001b[0m Xt \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_get_col_subset(X, columns, input_df)\n\u001b[1;32m    259\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m transformers \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m--> 260\u001b[0m     Xt \u001b[38;5;241m=\u001b[39m \u001b[43mtransformers\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtransform\u001b[49m\u001b[43m(\u001b[49m\u001b[43mXt\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    261\u001b[0m extracted\u001b[38;5;241m.\u001b[39mappend(_handle_feature(Xt))\n\u001b[1;32m    263\u001b[0m alias \u001b[38;5;241m=\u001b[39m options\u001b[38;5;241m.\u001b[39mget(\u001b[38;5;124m'\u001b[39m\u001b[38;5;124malias\u001b[39m\u001b[38;5;124m'\u001b[39m)\n",
      "File \u001b[0;32m~/virtualenvs/databoost/lib/python3.8/site-packages/sklearn/preprocessing/_encoders.py:882\u001b[0m, in \u001b[0;36mOneHotEncoder.transform\u001b[0;34m(self, X)\u001b[0m\n\u001b[1;32m    877\u001b[0m \u001b[38;5;66;03m# validation of X happens in _check_X called by _transform\u001b[39;00m\n\u001b[1;32m    878\u001b[0m warn_on_unknown \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdrop \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mhandle_unknown \u001b[38;5;129;01min\u001b[39;00m {\n\u001b[1;32m    879\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mignore\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[1;32m    880\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124minfrequent_if_exist\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[1;32m    881\u001b[0m }\n\u001b[0;32m--> 882\u001b[0m X_int, X_mask \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_transform\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    883\u001b[0m \u001b[43m    \u001b[49m\u001b[43mX\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    884\u001b[0m \u001b[43m    \u001b[49m\u001b[43mhandle_unknown\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mhandle_unknown\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    885\u001b[0m \u001b[43m    \u001b[49m\u001b[43mforce_all_finite\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mallow-nan\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\n\u001b[1;32m    886\u001b[0m \u001b[43m    \u001b[49m\u001b[43mwarn_on_unknown\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mwarn_on_unknown\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    887\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    888\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_map_infrequent_categories(X_int, X_mask)\n\u001b[1;32m    890\u001b[0m n_samples, n_features \u001b[38;5;241m=\u001b[39m X_int\u001b[38;5;241m.\u001b[39mshape\n",
      "File \u001b[0;32m~/virtualenvs/databoost/lib/python3.8/site-packages/sklearn/preprocessing/_encoders.py:160\u001b[0m, in \u001b[0;36m_BaseEncoder._transform\u001b[0;34m(self, X, handle_unknown, force_all_finite, warn_on_unknown)\u001b[0m\n\u001b[1;32m    155\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m handle_unknown \u001b[38;5;241m==\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124merror\u001b[39m\u001b[38;5;124m\"\u001b[39m:\n\u001b[1;32m    156\u001b[0m     msg \u001b[38;5;241m=\u001b[39m (\n\u001b[1;32m    157\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mFound unknown categories \u001b[39m\u001b[38;5;132;01m{0}\u001b[39;00m\u001b[38;5;124m in column \u001b[39m\u001b[38;5;132;01m{1}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    158\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m during transform\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;241m.\u001b[39mformat(diff, i)\n\u001b[1;32m    159\u001b[0m     )\n\u001b[0;32m--> 160\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(msg)\n\u001b[1;32m    161\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    162\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m warn_on_unknown:\n",
      "\u001b[0;31mValueError\u001b[0m: Found unknown categories ['C20'] in column 0 during transform"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import roc_auc_score, roc_curve\n",
    "import itertools\n",
    "\n",
    "result = []\n",
    "iterations_per_scenario = 5\n",
    "\n",
    "for scenario in SCENARIOS:\n",
    "    for legacy_size, _ in itertools.product(np.arange(0.1, 1.0, 0.05), np.arange(iterations_per_scenario)):\n",
    "\n",
    "        s = SCENARIOS[scenario]\n",
    "        train_X, test_X, train_y, test_y = s.train_test(legacy_size=legacy_size, legacy_drop_frac=0.3, legacy_missing_label=0.0)\n",
    "        base_pipe = s.build_pipeline(train_X)\n",
    "        \n",
    "        new_idx = ~(train_X['legacy'].astype(bool))\n",
    "        new_only_pipe = s.build_pipeline(train_X[new_idx])\n",
    "\n",
    "        with warnings.catch_warnings():\n",
    "            warnings.simplefilter(\"ignore\")\n",
    "            base_pipe.fit(train_X, train_y)\n",
    "            new_only_pipe.fit(train_X[new_idx], train_y[new_idx])\n",
    "            bp = databoosted_pipe(train_X, train_y, s)\n",
    "\n",
    "        with warnings.catch_warnings():\n",
    "            warnings.simplefilter(\"ignore\")\n",
    "            base_scored = base_pipe.predict_proba(test_X)[:,1]\n",
    "            new_only_scored = new_only_pipe.predict_proba(test_X)[:,1]\n",
    "            boosted_scored = bp.predict_proba(test_X)[:,1]\n",
    "            \n",
    "        \n",
    "        row_result = {'scenario': scenario, 'legacy_size': legacy_size}\n",
    "        for max_fpr in [None]:\n",
    "            if max_fpr is None:\n",
    "                suffix = ''\n",
    "            else:\n",
    "                suffix = '_{}'.format(max_fpr)\n",
    "            row_result['base_roc_auc{}'.format(suffix)] = roc_auc_score(test_y, base_scored, max_fpr=max_fpr)\n",
    "            row_result['boosted_roc_auc{}'.format(suffix)] = roc_auc_score(test_y, boosted_scored, max_fpr=max_fpr)\n",
    "            row_result['new_only_roc_auc{}'.format(suffix)] = roc_auc_score(test_y, new_only_scored, max_fpr=max_fpr)\n",
    "        \n",
    "        row_result['deleted_importance'] = (\n",
    "            pd.isnull(train_X[~new_idx]).all() @ importances[scenario].set_index('column')['importance']\n",
    "        )\n",
    "        result.append(row_result)\n",
    "\n",
    "result = pd.DataFrame(result)\n",
    "\n",
    "\n",
    "result = pd.merge(\n",
    "    result,\n",
    "    best_result[['scenario', 'roc_auc_score']].rename(columns={'roc_auc_score': 'full_data_auc_score'}),\n",
    "    on='scenario',\n",
    "    how='left'\n",
    ")\n",
    "\n",
    "result['boosted_regret'] = result['full_data_auc_score'] - result['boosted_roc_auc']\n",
    "result['base_regret'] = result['full_data_auc_score'] - result['base_roc_auc']\n",
    "result['new_only_regret'] = result['full_data_auc_score'] - result['new_only_roc_auc']\n",
    "result['win_base'] = result['boosted_roc_auc'] > result['base_roc_auc']\n",
    "result['win_new_only'] = result['boosted_roc_auc'] > result['new_only_roc_auc']\n",
    "result['delta_base'] = result['boosted_roc_auc'] - result['base_roc_auc']\n",
    "result['delta_new_only'] = result['boosted_roc_auc'] - result['new_only_roc_auc']\n",
    "result['win_delta'] = (result['boosted_roc_auc'] - result['base_roc_auc'])*result['win_base']\n",
    "result.groupby('scenario').agg({\n",
    "    'win_base': 'mean',\n",
    "    'win_new_only': 'mean',\n",
    "    'delta_base': 'mean',\n",
    "    'delta_new_only': 'mean',\n",
    "    'win_delta': 'mean',\n",
    "    'boosted_regret': 'mean',\n",
    "    'base_regret': 'mean',\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7d167fe0-0203-47d6-ad16-62daee67aef3",
   "metadata": {},
   "outputs": [],
   "source": [
    "result['new_only_regret'] = result['full_data_auc_score'] - result['new_only_roc_auc']\n",
    "result.groupby('scenario').agg({\n",
    "    'win_base': 'mean',\n",
    "    'win_new_only': 'mean',\n",
    "    'delta_base': 'mean',\n",
    "    'delta_new_only': 'mean',\n",
    "    'win_delta': 'mean',\n",
    "    'boosted_regret': 'mean',\n",
    "    'base_regret': 'mean',\n",
    "    'new_only_regret': 'mean',\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39ab6aab-962f-44ad-aba9-5e378cd4801c",
   "metadata": {},
   "source": [
    "## Dataset shift\n",
    "\n",
    "Let me repeat the above experiment, but I'll impose an additional constraint on the data: on the legacy data, 20% of labels will be (wrongly)\n",
    "set to zero."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "86cac913-d8c1-4fb5-982e-8629b2fe6f56",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(result.groupby('scenario').agg({\n",
    "    'win_base': 'mean',\n",
    "    'win_new_only': 'mean',\n",
    "    'delta_base': 'mean',\n",
    "    'delta_new_only': 'mean',\n",
    "#    'win_delta': 'mean',\n",
    "    'boosted_regret': 'mean',\n",
    "    'base_regret': 'mean',\n",
    "    'new_only_regret': 'mean',\n",
    "}).to_markdown(floatfmt=\".3f\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f3a3bd78-41a1-4451-83f7-7d785db5bdc0",
   "metadata": {},
   "outputs": [],
   "source": [
    "result.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8145e8e0-4447-4829-8d6a-98527f004c87",
   "metadata": {},
   "outputs": [],
   "source": [
    "scenario = 'car_insurance'\n",
    "x = result[result['scenario'] == scenario]\n",
    "\n",
    "plt.title(\"scenario={}\".format(scenario))\n",
    "plt.scatter(x['deleted_importance'], x['delta_base'], label='boosted_vs_base', alpha=0.25)\n",
    "plt.scatter(x['deleted_importance'], x['delta_new_only'], label='boosted_vs_new_only', alpha=0.25)\n",
    "plt.axhline(0, color='gray', alpha=0.5)\n",
    "plt.legend()\n",
    "plt.xlabel(\"Importance of censored features\")\n",
    "plt.ylabel(\"roc_auc(boosted) - roc_auc(base)\")\n",
    "\n",
    "plt.show()\n",
    "x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "eee3b25b-3789-4ce5-af27-1d4b6d754ca7",
   "metadata": {},
   "outputs": [],
   "source": [
    "n_scenarios = len(SCENARIOS)\n",
    "plot_cols = 3\n",
    "plot_rows = int(np.ceil(n_scenarios/plot_cols))\n",
    "i = 1\n",
    "\n",
    "for scenario in SCENARIOS:\n",
    "    plt.subplot(plot_rows*100+plot_cols*10+i)\n",
    "    i += 1\n",
    "    x = result[result['scenario'] == scenario]\n",
    "\n",
    "    plt.title(\"scenario={}\".format(scenario))\n",
    "    plt.scatter(x['legacy_size'], x['delta_base'], label='boosted_vs_base', alpha=0.25)\n",
    "    plt.scatter(x['legacy_size'], x['delta_new_only'], label='boosted_vs_new_only', alpha=0.25)\n",
    "    plt.axhline(0, color='gray', alpha=0.5)\n",
    "    plt.legend()\n",
    "    plt.xlabel(\"Fraction of old data\")\n",
    "    if (i % 3 == 2):\n",
    "        plt.ylabel(\"roc_auc(boosted) - roc_auc(base)\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8339178b-da93-4e4e-8ba6-51d28b711789",
   "metadata": {},
   "outputs": [],
   "source": [
    "n_scenarios = len(SCENARIOS)\n",
    "plot_cols = 3\n",
    "plot_rows = int(np.ceil(n_scenarios/plot_cols))\n",
    "i = 1\n",
    "\n",
    "for scenario in SCENARIOS:\n",
    "    plt.subplot(plot_rows*100+plot_cols*10+i)\n",
    "    i += 1\n",
    "    x = result[result['scenario'] == scenario]\n",
    "\n",
    "    plt.title(\"scenario={}\".format(scenario))\n",
    "    plt.scatter(x['deleted_importance'], x['delta_base'], label='boosted_vs_base', alpha=0.25)\n",
    "    plt.scatter(x['deleted_importance'], x['delta_new_only'], label='boosted_vs_new_only', alpha=0.25)\n",
    "    plt.axhline(0, color='gray', alpha=0.5)\n",
    "    plt.legend()\n",
    "    plt.xlabel(\"Importance of deleted features\")\n",
    "    if (i % 3 == 2):\n",
    "        plt.ylabel(\"roc_auc(boosted) - roc_auc(base)\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a9058c7c-8747-4321-96ed-0449339a77c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_auc_score, roc_curve\n",
    "import itertools\n",
    "\n",
    "result = []\n",
    "iterations_per_scenario = 5\n",
    "\n",
    "for scenario in SCENARIOS:\n",
    "    for legacy_drop_frac, _ in itertools.product(np.arange(0.1, 0.95, 0.05), np.arange(iterations_per_scenario)):\n",
    "\n",
    "        s = SCENARIOS[scenario]\n",
    "        train_X, test_X, train_y, test_y = s.train_test(legacy_size=0.85, legacy_drop_frac=legacy_drop_frac, legacy_missing_label=0.0)\n",
    "        base_pipe = s.build_pipeline(train_X)\n",
    "        \n",
    "        new_idx = ~(train_X['legacy'].astype(bool))\n",
    "        new_only_pipe = s.build_pipeline(train_X[new_idx])\n",
    "\n",
    "        with warnings.catch_warnings():\n",
    "            warnings.simplefilter(\"ignore\")\n",
    "            base_pipe.fit(train_X, train_y)\n",
    "            new_only_pipe.fit(train_X[new_idx], train_y[new_idx])\n",
    "            bp = databoosted_pipe(train_X, train_y, s)\n",
    "\n",
    "        with warnings.catch_warnings():\n",
    "            warnings.simplefilter(\"ignore\")\n",
    "            base_scored = base_pipe.predict_proba(test_X)[:,1]\n",
    "            new_only_scored = new_only_pipe.predict_proba(test_X)[:,1]\n",
    "            boosted_scored = bp.predict_proba(test_X)[:,1]\n",
    "            \n",
    "        \n",
    "        row_result = {'scenario': scenario, 'legacy_drop_frac': legacy_drop_frac}\n",
    "        for max_fpr in [None]:\n",
    "            if max_fpr is None:\n",
    "                suffix = ''\n",
    "            else:\n",
    "                suffix = '_{}'.format(max_fpr)\n",
    "            row_result['base_roc_auc{}'.format(suffix)] = roc_auc_score(test_y, base_scored, max_fpr=max_fpr)\n",
    "            row_result['boosted_roc_auc{}'.format(suffix)] = roc_auc_score(test_y, boosted_scored, max_fpr=max_fpr)\n",
    "            row_result['new_only_roc_auc{}'.format(suffix)] = roc_auc_score(test_y, new_only_scored, max_fpr=max_fpr)\n",
    "        \n",
    "        row_result['deleted_importance'] = (\n",
    "            pd.isnull(train_X[~new_idx]).all() @ importances[scenario].set_index('column')['importance']\n",
    "        )\n",
    "        result.append(row_result)\n",
    "\n",
    "result = pd.DataFrame(result)\n",
    "\n",
    "\n",
    "result = pd.merge(\n",
    "    result,\n",
    "    best_result[['scenario', 'roc_auc_score']].rename(columns={'roc_auc_score': 'full_data_auc_score'}),\n",
    "    on='scenario',\n",
    "    how='left'\n",
    ")\n",
    "\n",
    "result['boosted_regret'] = result['full_data_auc_score'] - result['boosted_roc_auc']\n",
    "result['base_regret'] = result['full_data_auc_score'] - result['base_roc_auc']\n",
    "result['new_only_regret'] = result['full_data_auc_score'] - result['new_only_roc_auc']\n",
    "result['win_base'] = result['boosted_roc_auc'] > result['base_roc_auc']\n",
    "result['win_new_only'] = result['boosted_roc_auc'] > result['new_only_roc_auc']\n",
    "result['delta_base'] = result['boosted_roc_auc'] - result['base_roc_auc']\n",
    "result['delta_new_only'] = result['boosted_roc_auc'] - result['new_only_roc_auc']\n",
    "result['win_delta'] = (result['boosted_roc_auc'] - result['base_roc_auc'])*result['win_base']\n",
    "result.groupby('scenario').agg({\n",
    "    'win_base': 'mean',\n",
    "    'win_new_only': 'mean',\n",
    "    'delta_base': 'mean',\n",
    "    'delta_new_only': 'mean',\n",
    "    'win_delta': 'mean',\n",
    "    'boosted_regret': 'mean',\n",
    "    'base_regret': 'mean',\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2abf39a6-2937-4935-82f7-17b6edec6e40",
   "metadata": {},
   "outputs": [],
   "source": [
    "n_scenarios = len(SCENARIOS)\n",
    "plot_cols = 3\n",
    "plot_rows = int(np.ceil(n_scenarios/plot_cols))\n",
    "i = 1\n",
    "\n",
    "for scenario in SCENARIOS:\n",
    "    plt.subplot(plot_rows*100+plot_cols*10+i)\n",
    "    i += 1\n",
    "    x = result[result['scenario'] == scenario]\n",
    "\n",
    "    plt.title(\"scenario={}\".format(scenario))\n",
    "    plt.scatter(x['deleted_importance'], x['delta_base'], label='boosted_vs_base', alpha=0.25)\n",
    "    plt.scatter(x['deleted_importance'], x['delta_new_only'], label='boosted_vs_new_only', alpha=0.25)\n",
    "    plt.axhline(0, color='gray', alpha=0.5)\n",
    "    plt.legend()\n",
    "    plt.xlabel(\"Importance of deleted features\")\n",
    "    if (i % 3 == 2):\n",
    "        plt.ylabel(\"roc_auc(boosted) - roc_auc(base)\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9e0ab359-8e76-4366-a6b2-82f0e2985a46",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "databoost",
   "language": "python",
   "name": "databoost"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}