{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "cf3f752e",
   "metadata": {},
   "source": [
    "# TMVA_SOFIE_RSofieReader\n",
    "This macro provides an example of using a trained model with Keras\n",
    "and make inference using SOFIE with the RSofieReader class\n",
    "This macro uses as input a Keras model generated with the\n",
    "TMVA_Higgs_Classification.C tutorial\n",
    "You need to run that macro before to generate the trained Keras model\n",
    "\n",
    "\n",
    "Execute in this order:\n",
    "```\n",
    "root TMVA_Higgs_Classification.C\n",
    "root TMVA_SOFIE_RSofieReader.C\n",
    "```\n",
    "\n",
    "\n",
    "\n",
    "**Author:** Lorenzo Moneta  \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 08:23 PM.</small></i>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4f33b656",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "using namespace TMVA::Experimental;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3d6181e7",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "RSofieReader model(\"HiggsModel.keras\", {}, true );"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9193917d",
   "metadata": {},
   "source": [
    "for debugging\n",
    "RSofieReader model(\"Higgs_trained_model.keras\", {}, true);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a77f7cc0",
   "metadata": {},
   "source": [
    "the input shape for this model is a tensor with shape (1,7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5e250257",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "std::vector<float> input = {0.1,0.2,0.3,0.4,0.5,0.6,0.7};"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3d70f2ca",
   "metadata": {},
   "source": [
    "predict model on a single event (takes a std::vector<float>)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "482f6871",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "auto output = model.Compute(input);\n",
    "\n",
    "std::cout << \"Event prediction = \" << output[0] << std::endl;"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7dc8059",
   "metadata": {},
   "source": [
    "predict model now on a input file using RDataFrame"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5df4fdf7",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "std::string inputFileName = \"Higgs_data.root\";\n",
    "std::string inputFile = std::string{gROOT->GetTutorialDir()} + \"/machine_learning/data/\" + inputFileName;\n",
    "\n",
    "\n",
    "ROOT::RDataFrame df1(\"sig_tree\", inputFile);\n",
    "\n",
    "auto h1 = df1.Define(\"DNN_Values\", Compute<7, float>(model),\n",
    "                         {\"m_jj\", \"m_jjj\", \"m_lv\", \"m_jlv\", \"m_bb\", \"m_wbb\", \"m_wwbb\"})\n",
    "               .Define(\"y\",\"DNN_Values[0]\")\n",
    "               .Histo1D({\"h_sig\", \"\", 100, 0, 1}, \"y\");\n",
    "\n",
    "ROOT::RDataFrame df2(\"bkg_tree\", inputFile);\n",
    "auto h2 = df2.Define(\"DNN_Values\", Compute<7, float>(model),\n",
    "                         {\"m_jj\", \"m_jjj\", \"m_lv\", \"m_jlv\", \"m_bb\", \"m_wbb\", \"m_wwbb\"})\n",
    "               .Define(\"y\",\"DNN_Values[0]\")\n",
    "               .Histo1D({\"h_bkg\", \"\", 100, 0, 1}, \"y\");\n",
    "\n",
    "h1->SetLineColor(kRed);\n",
    "h2->SetLineColor(kBlue);\n",
    "\n",
    "auto c1 = new TCanvas();\n",
    "gStyle->SetOptStat(0);\n",
    "\n",
    "h2->DrawClone();\n",
    "h1->DrawClone(\"SAME\");\n",
    "c1->BuildLegend();"
   ]
  }
 ],
 "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
}