{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "efb09f11",
   "metadata": {},
   "source": [
    "# TMVA_SOFIE_ONNX\n",
    "This macro provides a simple example for the parsing of ONNX files into\n",
    "RModel object and further generating the .hxx header files for inference.\n",
    "\n",
    "\n",
    "\n",
    "**Author:** Sanjiban Sengupta  \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": "d0b18cd7",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "using namespace TMVA::Experimental;"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "06a290c9",
   "metadata": {},
   "source": [
    " Arguments are defined. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b0c96d4d",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "std::string inputFile = \"\";"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dc4693c0",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "if (inputFile.empty() )\n",
    "   inputFile = std::string(gROOT->GetTutorialsDir()) + \"/machine_learning/Linear_16.onnx\";"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c57ea89d",
   "metadata": {},
   "source": [
    "Creating parser object to parse ONNX files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "84b8b5de",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    " SOFIE::RModelParser_ONNX parser;\n",
    " SOFIE::RModel model = parser.Parse(inputFile, true);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79ac7848",
   "metadata": {},
   "source": [
    "Generating inference code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "93ee2422",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    " model.Generate();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "54a5f430",
   "metadata": {},
   "source": [
    "write the code in a file (by default Linear_16.hxx and Linear_16.dat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0b7c329c",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    " model.OutputGenerated();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "53ef8148",
   "metadata": {},
   "source": [
    "Printing required input tensors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a84c54ed",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    " model.PrintRequiredInputTensors();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c2e8f1c",
   "metadata": {},
   "source": [
    "Printing initialized tensors (weights)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aef6102e",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    " std::cout<<\"\\n\\n\";\n",
    " model.PrintInitializedTensors();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0321524a",
   "metadata": {},
   "source": [
    "Printing intermediate tensors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2ff643d8",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    " std::cout<<\"\\n\\n\";\n",
    " model.PrintIntermediateTensors();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6058fee7",
   "metadata": {},
   "source": [
    "Checking if tensor already exist in model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "08a75c2f",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    " std::cout<<\"\\n\\nTensor \\\"16weight\\\" already exist: \"<<std::boolalpha<<model.CheckIfTensorAlreadyExist(\"16weight\")<<\"\\n\\n\";\n",
    " std::vector<size_t> tensorShape = model.GetTensorShape(\"16weight\");\n",
    " std::cout<<\"Shape of tensor \\\"16weight\\\": \";\n",
    " for(auto& it:tensorShape){\n",
    "     std::cout<<it<<\",\";\n",
    " }\n",
    " std::cout<<\"\\n\\nData type of tensor \\\"16weight\\\": \";\n",
    " SOFIE::ETensorType tensorType = model.GetTensorType(\"16weight\");\n",
    " std::cout<<SOFIE::ConvertTypeToString(tensorType);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d774b80",
   "metadata": {},
   "source": [
    "Printing generated inference code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7642604b",
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    " std::cout<<\"\\n\\n\";\n",
    " model.PrintGenerated();"
   ]
  }
 ],
 "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
}