tmva_comparison_study.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebooks aims at a quick study comparing a couple of standard choices of BDT architectures from the ROOT TMVA framework to the HistGradientBoostingClassifier from scikit-learn. Three learning tasks are tested: the four default input features only, adding two guassian noise features and adding ten gaussian noise features.\n",
    "\n",
    "The BDT settings are inspired from example code, such as https://root.cern/doc/master/TMVAClassification_8C.html and their configuration files are stored in `TMVA_configs/`. Any new TMVA BDT architecture can be added by creating another such YAML config file using the file name (without the `.yaml` extension) as the `model_identifier` argument during training.\n",
    "\n",
    "Note that the training of the models can take quite long and thus one is encouraged to export the correspondoing cells into individual scripts and submit them to a HCP cluster. Each model is loaded again for evaluation and thus the training cells can be skipped when rerunning the notebook.\n",
    "\n",
    "In order to run the study, both `ROOT` and `uproot` are required to be installed. This can be done via conda:\n",
    "```bash\n",
    "conda install -c conda-forge root\n",
    "conda install -c conda-forge uproot\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from os.path import join\n",
    "from utils import (load_lhco_rd, add_gaussian_features, train_model_multirun,\n",
    "                   load_models_allruns, eval_ensemble, multi_roc_sigeffs)\n",
    "from tmva_utils import train_tmva_multi, eval_tmva_multi\n",
    "from plot_utils import plot_sic_curves"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Some global settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# base directory to store all models\n",
    "base_dir = \"TMVA_comparison_study/\"\n",
    "\n",
    "# How often to re-run the entire ensemble training procedure\n",
    "num_runs = 10\n",
    "\n",
    "# How many models constitute a single ensemble\n",
    "ensembles_per_model = 10\n",
    "\n",
    "# cross-validation mode\n",
    "cv_mode = \"random\"\n",
    "\n",
    "# HGB-specific parameters\n",
    "max_iters = 100\n",
    "compute_val_weights = True"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Loading the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# loading data and adding gaussian features\n",
    "data = load_lhco_rd(\"./lhco_rd\")\n",
    "data_2g = add_gaussian_features(data, n_gaussians=2)\n",
    "data_10g = add_gaussian_features(data, n_gaussians=10)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training the trees"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model training can also be done separately in a script, e.g. running on a HCP. In that case, one can skip the following training cells when evaluating the models."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training the TMVA BDTs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDT\", root_file_dir_base=join(base_dir, \"BDT_0G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data_2g,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDT\", root_file_dir_base=join(base_dir, \"BDT_2G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data_10g,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDT\", root_file_dir_base=join(base_dir, \"BDT_10G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDTB\", root_file_dir_base=join(base_dir, \"BDTB_0G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data_2g,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDTB\", root_file_dir_base=join(base_dir, \"BDTB_2G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data_10g,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDTB\", root_file_dir_base=join(base_dir, \"BDTB_10G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDTD\", root_file_dir_base=join(base_dir, \"BDTD_0G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data_2g,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDTD\", root_file_dir_base=join(base_dir, \"BDTD_2G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data_10g,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDTD\", root_file_dir_base=join(base_dir, \"BDTD_10G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDTG\", root_file_dir_base=join(base_dir, \"BDTG_0G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data_2g,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDTG\", root_file_dir_base=join(base_dir, \"BDTG_2G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_ = train_tmva_multi(\n",
    "    data_10g,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, cv_mode=cv_mode,\n",
    "    model_identifier=\"BDTG\", root_file_dir_base=join(base_dir, \"BDTG_10G\"))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training the HGB benchmark classifiers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_, models_hgb_0g = train_model_multirun(\n",
    "    data,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model,\n",
    "    max_iters=max_iters, model_type=\"HGB\", compute_val_weights=True,\n",
    "    save_model_dir=join(base_dir, \"HGB_0G\"),\n",
    "    cv_mode=cv_mode, early_stopping=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_, models_hgb_2g = train_model_multirun(\n",
    "    data_2g,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model,\n",
    "    max_iters=max_iters, model_type=\"HGB\", compute_val_weights=True,\n",
    "    save_model_dir=join(base_dir, \"HGB_2G\"),\n",
    "    cv_mode=cv_mode, early_stopping=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "_, models_hgb_10g = train_model_multirun(\n",
    "    data_10g,\n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model,\n",
    "    max_iters=max_iters, model_type=\"HGB\", compute_val_weights=True,\n",
    "    save_model_dir=join(base_dir, \"HGB_10G\"),\n",
    "    cv_mode=cv_mode, early_stopping=True)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Computing predictions"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate TMVA BDTs"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the case of the TMVA models, the evaluation can be somewhat lengthy as the pyROOT implementation here includes a python for loop over events. Once done, these cells store the test set predictions on disk and can be skipped when rerunning the notebook. There are cells loading the predictions from disk further below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_bdt_0g = eval_tmva_multi(\n",
    "    data, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDT\",\n",
    "    root_file_dir_base=join(base_dir, \"BDT_0G\"), save_ensemble_preds=True)\n",
    "\n",
    "predictions_bdt_2g = eval_tmva_multi(\n",
    "    data_2g, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDT\",\n",
    "    root_file_dir_base=join(base_dir, \"BDT_2G\"), save_ensemble_preds=True)\n",
    "\n",
    "predictions_bdt_10g = eval_tmva_multi(\n",
    "    data_10g, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDT\",\n",
    "    root_file_dir_base=join(base_dir, \"BDT_10G\"), save_ensemble_preds=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_bdtb_0g = eval_tmva_multi(\n",
    "    data, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDTB\",\n",
    "    root_file_dir_base=join(base_dir, \"BDTB_0G\"), save_ensemble_preds=True)\n",
    "\n",
    "predictions_bdtb_2g = eval_tmva_multi(\n",
    "    data_2g, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDTB\",\n",
    "    root_file_dir_base=join(base_dir, \"BDTB_2G\"), save_ensemble_preds=True)\n",
    "\n",
    "predictions_bdtb_10g = eval_tmva_multi(\n",
    "    data_10g, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDTB\",\n",
    "    root_file_dir_base=join(base_dir, \"BDTB_10G\"), save_ensemble_preds=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_bdtd_0g = eval_tmva_multi(\n",
    "    data, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDTD\",\n",
    "    root_file_dir_base=join(base_dir, \"BDTD_0G\"), save_ensemble_preds=True)\n",
    "\n",
    "predictions_bdtd_2g = eval_tmva_multi(\n",
    "    data_2g, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDTD\",\n",
    "    root_file_dir_base=join(base_dir, \"BDTD_2G\"), save_ensemble_preds=True)\n",
    "\n",
    "predictions_bdtd_10g = eval_tmva_multi(\n",
    "    data_10g, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDTD\",\n",
    "    root_file_dir_base=join(base_dir, \"BDTD_10G\"), save_ensemble_preds=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_bdtg_0g = eval_tmva_multi(\n",
    "    data, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDTG\",\n",
    "    root_file_dir_base=join(base_dir, \"BDTG_0G\"), save_ensemble_preds=True)\n",
    "\n",
    "predictions_bdtg_2g = eval_tmva_multi(\n",
    "    data_2g, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDTG\",\n",
    "    root_file_dir_base=join(base_dir, \"BDTG_2G\"), save_ensemble_preds=True)\n",
    "\n",
    "predictions_bdtg_10g = eval_tmva_multi(\n",
    "    data_10g, \n",
    "    num_runs=num_runs, ensembles_per_model=ensembles_per_model, model_identifier=\"BDTG\",\n",
    "    root_file_dir_base=join(base_dir, \"BDTG_10G\"), save_ensemble_preds=True)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate HGB benchmark"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Loading the models from disk in case the notebook was restarted.\n",
    "# Does not need to be run in case the notebook is run from top to bottom.\n",
    "\n",
    "models_hgb_0g = load_models_allruns(join(base_dir, \"HGB_0G\"))\n",
    "models_hgb_2g = load_models_allruns(join(base_dir, \"HGB_2G\"))\n",
    "models_hgb_10g = load_models_allruns(join(base_dir, \"HGB_10G\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_hgb_0g = eval_ensemble(\n",
    "    models_hgb_0g, data, model_type=\"HGB\")\n",
    "\n",
    "predictions_hgb_2g = eval_ensemble(\n",
    "    models_hgb_2g, data_2g, model_type=\"HGB\")\n",
    "\n",
    "predictions_hgb_10g = eval_ensemble(\n",
    "    models_hgb_10g, data_10g, model_type=\"HGB\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Shortcut to skip costly live evaluation on TMVA BDTs"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following two cells are just needed in case the TMVA tree evaluation has been run already and the notebook was restarted in the meanwhile. In that case the test set predictions are simply read from disk."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_predictions_tmva(num_runs, root_file_dir_base):\n",
    "    predictions = []\n",
    "    for i in range(num_runs):\n",
    "        predictions.append(np.load(join(f\"{root_file_dir_base}_{i}\",\n",
    "                                        f\"ensemble_preds_run{i}.npy\")))\n",
    "    return np.stack(np.mean(predictions, axis=0), axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions_bdt_0g = load_predictions_tmva(num_runs, join(base_dir, \"BDT_0G\"))\n",
    "predictions_bdt_2g = load_predictions_tmva(num_runs, join(base_dir, \"BDT_2G\"))\n",
    "predictions_bdt_10g = load_predictions_tmva(num_runs, join(base_dir, \"BDT_10G\"))\n",
    "\n",
    "predictions_bdtb_0g = load_predictions_tmva(num_runs, join(base_dir, \"BDTB_0G\"))\n",
    "predictions_bdtb_2g = load_predictions_tmva(num_runs, join(base_dir, \"BDTB_2G\"))\n",
    "predictions_bdtb_10g = load_predictions_tmva(num_runs, join(base_dir, \"BDTB_10G\"))\n",
    "\n",
    "predictions_bdtd_0g = load_predictions_tmva(num_runs, join(base_dir, \"BDTD_0G\"))\n",
    "predictions_bdtd_2g = load_predictions_tmva(num_runs, join(base_dir, \"BDTD_2G\"))\n",
    "predictions_bdtd_10g = load_predictions_tmva(num_runs, join(base_dir, \"BDTD_10G\"))\n",
    "\n",
    "predictions_bdtg_0g = load_predictions_tmva(num_runs, join(base_dir, \"BDTG_0G\"))\n",
    "predictions_bdtg_2g = load_predictions_tmva(num_runs, join(base_dir, \"BDTG_2G\"))\n",
    "predictions_bdtg_10g = load_predictions_tmva(num_runs, join(base_dir, \"BDTG_10G\"))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Performance comparison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(\"working on BDT...\")\n",
    "tpr_bdt_0g, fpr_bdt_0g = multi_roc_sigeffs(predictions_bdt_0g, data[\"y_test\"])\n",
    "tpr_bdt_2g, fpr_bdt_2g = multi_roc_sigeffs(predictions_bdt_2g, data[\"y_test\"])\n",
    "tpr_bdt_10g, fpr_bdt_10g = multi_roc_sigeffs(predictions_bdt_10g, data[\"y_test\"])\n",
    "\n",
    "print(\"working on BDTB...\")\n",
    "tpr_bdtb_0g, fpr_bdtb_0g = multi_roc_sigeffs(predictions_bdtb_0g, data[\"y_test\"])\n",
    "tpr_bdtb_2g, fpr_bdtb_2g = multi_roc_sigeffs(predictions_bdtb_2g, data[\"y_test\"])\n",
    "tpr_bdtb_10g, fpr_bdtb_10g = multi_roc_sigeffs(predictions_bdtb_10g, data[\"y_test\"])\n",
    "\n",
    "print(\"working on BDTD...\")\n",
    "tpr_bdtd_0g, fpr_bdtd_0g = multi_roc_sigeffs(predictions_bdtd_0g, data[\"y_test\"])\n",
    "tpr_bdtd_2g, fpr_bdtd_2g = multi_roc_sigeffs(predictions_bdtd_2g, data[\"y_test\"])\n",
    "tpr_bdtd_10g, fpr_bdtd_10g = multi_roc_sigeffs(predictions_bdtd_10g, data[\"y_test\"])\n",
    "\n",
    "print(\"working on BDTG...\")\n",
    "tpr_bdtg_0g, fpr_bdtg_0g = multi_roc_sigeffs(predictions_bdtg_0g, data[\"y_test\"])\n",
    "tpr_bdtg_2g, fpr_bdtg_2g = multi_roc_sigeffs(predictions_bdtg_2g, data[\"y_test\"])\n",
    "tpr_bdtg_10g, fpr_bdtg_10g = multi_roc_sigeffs(predictions_bdtg_10g, data[\"y_test\"])\n",
    "\n",
    "print(\"working on HGB...\")\n",
    "tpr_hgb_0g, fpr_hgb_0g = multi_roc_sigeffs(predictions_hgb_0g, data[\"y_test\"])\n",
    "tpr_hgb_2g, fpr_hgb_2g = multi_roc_sigeffs(predictions_hgb_2g, data[\"y_test\"])\n",
    "tpr_hgb_10g, fpr_hgb_10g = multi_roc_sigeffs(predictions_hgb_10g, data[\"y_test\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "with np.errstate(divide='ignore', invalid='ignore'):  # ignoring annoying warnings\n",
    "        plot_sic_curves([tpr_bdt_0g, tpr_bdtb_0g, tpr_bdtd_0g, tpr_bdtg_0g, tpr_hgb_0g],\n",
    "                        [fpr_bdt_0g, fpr_bdtb_0g, fpr_bdtd_0g, fpr_bdtg_0g, fpr_hgb_0g],\n",
    "                        5*[data[\"y_test\"]],\n",
    "                        out_filename=join(base_dir, \"TMVA_BDT_vs_HGB_0G.pdf\"),\n",
    "                        labels=[\"TMVA BDT, 0G\", \"TMVA BDTB, 0G\", \"TMVA BDTD, 0G\",\n",
    "                                \"TMVA BDTG, 0G\", \"HGB, 0G\"],\n",
    "                        max_y=17)\n",
    "        plot_sic_curves([tpr_bdt_2g, tpr_bdtb_2g, tpr_bdtd_2g, tpr_bdtg_2g, tpr_hgb_2g],\n",
    "                        [fpr_bdt_2g, fpr_bdtb_2g, fpr_bdtd_2g, fpr_bdtg_2g, fpr_hgb_2g],\n",
    "                        5*[data[\"y_test\"]],\n",
    "                        out_filename=join(base_dir, \"TMVA_BDT_vs_HGB_2G.pdf\"),\n",
    "                        labels=[\"TMVA BDT, 2G\", \"TMVA BDTB, 2G\", \"TMVA BDTD, 2G\",\n",
    "                                \"TMVA BDTG, 2G\", \"HGB, 2G\"],\n",
    "                        max_y=17)\n",
    "        plot_sic_curves([tpr_bdt_10g, tpr_bdtb_10g, tpr_bdtd_10g, tpr_bdtg_10g, tpr_hgb_10g],\n",
    "                        [fpr_bdt_10g, fpr_bdtb_10g, fpr_bdtd_10g, fpr_bdtg_10g, fpr_hgb_10g],\n",
    "                        5*[data[\"y_test\"]],\n",
    "                        out_filename=join(base_dir, \"TMVA_BDT_vs_HGB_10G.pdf\"),\n",
    "                        labels=[\"TMVA BDT, 10G\", \"TMVA BDTB, 10G\", \"TMVA BDTD, 10G\",\n",
    "                                \"TMVA BDTG, 10G\", \"HGB, 10G\"],\n",
    "                        max_y=17)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "TreeEnv",
   "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": "3.8.13"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}