{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "39d8b9c2",
   "metadata": {},
   "source": [
    "# tmva103_Application\n",
    "This tutorial illustrates how you can conveniently apply BDTs in C++ using\n",
    "the fast tree inference engine offered by TMVA. Supported workflows are\n",
    "event-by-event inference, batch inference and pipelines with RDataFrame.\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "**Author:** Stefan Wunsch  \n",
    "<i><small>This notebook tutorial was automatically generated with <a href= \"https://github.com/root-project/root/blob/master/documentation/doxygen/converttonotebook.py\">ROOTBOOK-izer</a> from the macro found in the ROOT repository  on Tuesday, May 19, 2026 at 08:22 PM.</small></i>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d4661a77",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "using namespace TMVA::Experimental;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "536a7d5a",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "const char* model_filename = \"tmva101.root\";\n",
    "\n",
    "if (gSystem->AccessPathName(model_filename)) {\n",
    "   Info(\"tmva103_Application.C\", \"%s does not exist\", model_filename);\n",
    "   return;\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "045347fa",
   "metadata": {},
   "source": [
    "Load BDT model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "042e47c7",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "RBDT bdt(\"myBDT\", model_filename);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2713893f",
   "metadata": {},
   "source": [
    "Apply model on a single input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "84fe481f",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "auto y1 = bdt.Compute({1.0, 2.0, 3.0, 4.0});\n",
    "\n",
    "std::cout << \"Apply model on a single input vector: \" << y1[0] << std::endl;"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "727b430f",
   "metadata": {},
   "source": [
    "Apply model on a batch of inputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "936e4bca",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "float data[8] = {1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0};\n",
    "RTensor<float> x(data, {2, 4});\n",
    "auto y2 = bdt.Compute(x);\n",
    "\n",
    "std::cout << \"Apply model on an input tensor: \" << y2 << std::endl;"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e9957c36",
   "metadata": {},
   "source": [
    "Apply model as part of an RDataFrame workflow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c641b006",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "ROOT::RDataFrame df(\"Events\", \"root://eospublic.cern.ch//eos/root-eos/cms_opendata_2012_nanoaod/SMHiggsToZZTo4L.root\");\n",
    "auto df2 = df.Filter(\"nMuon >= 2\")\n",
    "             .Filter(\"nElectron >= 2\")\n",
    "             .Define(\"Muon_pt_1\", \"Muon_pt[0]\")\n",
    "             .Define(\"Muon_pt_2\", \"Muon_pt[1]\")\n",
    "             .Define(\"Electron_pt_1\", \"Electron_pt[0]\")\n",
    "             .Define(\"Electron_pt_2\", \"Electron_pt[1]\")\n",
    "             .Define(\"y\",\n",
    "                     Compute<4, float>(bdt),\n",
    "                     {\"Muon_pt_1\", \"Muon_pt_2\", \"Electron_pt_1\", \"Electron_pt_2\"});\n",
    "\n",
    "std::cout << \"Mean response on the signal sample: \" << *df2.Mean(\"y\") << std::endl;"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "ROOT C++",
   "language": "c++",
   "name": "root"
  },
  "language_info": {
   "codemirror_mode": "text/x-c++src",
   "file_extension": ".C",
   "mimetype": " text/x-c++src",
   "name": "c++"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}