TMVA DataLoader.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"http://oproject.org/tiki-download_file.php?fileId=8&display&x=450&y=128\">\n",
    "<img src=\"http://gfif.udea.edu.co/root/tmva/img/tmva_logo.gif\">\n",
    "\n",
    "## Required header files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#include <iostream>\n",
    "#include \"TChain.h\"\n",
    "#include \"TFile.h\"\n",
    "#include \"TTree.h\"\n",
    "#include \"TString.h\"\n",
    "#include \"TSystem.h\"\n",
    "#include \"TROOT.h\"\n",
    "\n",
    "#include \"TMVA/Factory.h\"\n",
    "#include \"TMVA/Tools.h\"\n",
    "#include \"TMVA/DataLoader.h\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Declare Factory and DataLoader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- Factory                  : You are running ROOT Version: 6.05/03, Sep 14, 2015\n",
      "--- Factory                  : \n",
      "--- Factory                  : _/_/_/_/_/ _|      _|  _|      _|    _|_|   \n",
      "--- Factory                  :    _/      _|_|  _|_|  _|      _|  _|    _| \n",
      "--- Factory                  :   _/       _|  _|  _|  _|      _|  _|_|_|_| \n",
      "--- Factory                  :  _/        _|      _|    _|  _|    _|    _| \n",
      "--- Factory                  : _/         _|      _|      _|      _|    _| \n",
      "--- Factory                  : \n",
      "--- Factory                  : ___________TMVA Version 5.0.0 alpha, Oct 30, 2015\n",
      "--- Factory                  : \n"
     ]
    }
   ],
   "source": [
    "TMVA::Tools::Instance();\n",
    "\n",
    "TString outfileName( \"TMVAOutputDL.root\" );\n",
    "TFile* outputFile = TFile::Open( outfileName, \"RECREATE\" );\n",
    "\n",
    "TMVA::Factory *factory = new TMVA::Factory( \"TMVAClassification\", outputFile,                                       \"!V:ROC:!Correlations:!Silent:Color:!DrawProgressBar:AnalysisType=Classification\" );\n",
    "   \n",
    "TMVA::DataLoader *loader1=new TMVA::DataLoader(\"mymc-dataset1\");\n",
    "TMVA::DataLoader *loader2=new TMVA::DataLoader(\"mymc-dataset2\");\n",
    "TFile *input = TFile::Open( \"MyMCdataset.root\");                "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Adding variables to DataLoader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "//adding variables to dataset 1\n",
    "loader1->AddVariable( \"var0\", 'F' );\n",
    "loader1->AddVariable( \"var1\", 'F' );\n",
    "loader1->AddVariable( \"var2\", 'F' );\n",
    "\n",
    "//adding variables to dataset 2\n",
    "loader2->AddVariable( \"var0\", 'F' );\n",
    "loader2->AddVariable( \"var1\", 'F' );\n",
    "loader2->AddVariable( \"var2\", 'F' );\n",
    "loader2->AddVariable( \"var3 := var0-var1\", 'F' );\n",
    "loader2->AddVariable( \"var4 := var0*var2\", 'F' );\n",
    "loader2->AddVariable( \"var5 := var1+var2\", 'F' );"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setting up dataset from Trees"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- TMVAClassification       : Using input file: MyMCdataset.root\n",
      "--- DataSetInfo              : Dataset[mymc-dataset1] : Added class \"Signal\"\t with internal class number 0\n",
      "--- mymc-dataset1            : Add Tree MyMCSig of type Signal with 5449 events\n",
      "--- DataSetInfo              : Dataset[mymc-dataset1] : Added class \"Background\"\t with internal class number 1\n",
      "--- mymc-dataset1            : Add Tree MyMCBkg of type Background with 5449 events\n",
      "--- mymc-dataset1            : Preparing trees for training and testing...\n",
      "--- DataSetInfo              : Dataset[mymc-dataset2] : Added class \"Signal\"\t with internal class number 0\n",
      "--- mymc-dataset2            : Add Tree MyMCSig of type Signal with 5449 events\n",
      "--- DataSetInfo              : Dataset[mymc-dataset2] : Added class \"Background\"\t with internal class number 1\n",
      "--- mymc-dataset2            : Add Tree MyMCBkg of type Background with 5449 events\n",
      "--- mymc-dataset2            : Preparing trees for training and testing...\n"
     ]
    }
   ],
   "source": [
    "std::cout << \"--- TMVAClassification       : Using input file: \" << input->GetName() << std::endl;\n",
    "   \n",
    "// --- Register the training and test trees\n",
    "\n",
    "TTree *tsignal     = (TTree*)input->Get(\"MyMCSig\");\n",
    "TTree *tbackground = (TTree*)input->Get(\"MyMCBkg\");\n",
    "   \n",
    "Double_t signalWeight     = 1.0;\n",
    "Double_t backgroundWeight = 1.0;\n",
    "\n",
    "TCut mycuts = \"\";\n",
    "TCut mycutb = \"\";\n",
    "   \n",
    "loader1->AddSignalTree    ( tsignal,     signalWeight     );\n",
    "loader1->AddBackgroundTree( tbackground, backgroundWeight );       \n",
    "loader1->fSignalWeight=signalWeight;\n",
    "loader1->fBackgroundWeight=backgroundWeight;\n",
    "loader1->fTreeS=tsignal;\n",
    "loader1->fTreeB=tbackground;\n",
    "loader1->PrepareTrainingAndTestTree( mycuts, mycutb,  \"nTrain_Signal=3000:nTrain_Background=3000:nTest_Signal=1449:nTest_Background=1449:SplitMode=Random:NormMode=NumEvents:!V\" ); \n",
    "\n",
    "\n",
    "loader2->AddSignalTree    ( tsignal,     signalWeight     );\n",
    "loader2->AddBackgroundTree( tbackground, backgroundWeight );       \n",
    "loader2->fSignalWeight=signalWeight;\n",
    "loader2->fBackgroundWeight=backgroundWeight;\n",
    "loader2->fTreeS=tsignal;\n",
    "loader2->fTreeB=tbackground;\n",
    "loader2->PrepareTrainingAndTestTree( mycuts, mycutb,  \"nTrain_Signal=3000:nTrain_Background=3000:nTest_Signal=1449:nTest_Background=1449:SplitMode=Random:NormMode=NumEvents:!V\" );   \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "// TMVA ANN: MLP (recommended ANN) -- all ANNs in TMVA are Multilayer Perceptrons\n",
    "factory->BookMethod( loader2,TMVA::Types::kMLP, \"MLP\", \"!V:NeuronType=tanh:VarTransform=N:NCycles=50:HiddenLayers=N+5:TestRate=5:!UseRegulator\" );## Booking Methods using different DataLoaders"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- Factory                  : Booking method: \u001b[1mMLP\u001b[0m DataSet Name: \u001b[1mmymc-dataset1\u001b[0m\n",
      "--- MLP                      : Dataset[mymc-dataset1] : Create Transformation \"N\" with events from all classes.\n",
      "--- Norm                     : Transformation, Variable selection : \n",
      "--- Norm                     : Input : variable 'var0' (index=0).   <---> Output : variable 'var0' (index=0).\n",
      "--- Norm                     : Input : variable 'var1' (index=1).   <---> Output : variable 'var1' (index=1).\n",
      "--- Norm                     : Input : variable 'var2' (index=2).   <---> Output : variable 'var2' (index=2).\n",
      "--- MLP                      : Building Network\n",
      "--- MLP                      : Initializing weights\n",
      "--- Factory                  : Booking method: \u001b[1mBDT\u001b[0m DataSet Name: \u001b[1mmymc-dataset2\u001b[0m\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Splitmode is: \"RANDOM\" the mixmode is: \"SAMEASSPLITMODE\"\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Create training and testing trees -- looping over class \"Signal\" ...\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Weight expression for class 'Signal': \"\"\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Create training and testing trees -- looping over class \"Background\" ...\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Weight expression for class 'Background': \"\"\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Number of events in input trees (after possible flattening of arrays):\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :     Signal          -- number of events       : 5449   / sum of weights: 5449 \n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :     Background      -- number of events       : 5449   / sum of weights: 5449 \n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :     Signal     tree -- total number of entries: 5449 \n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :     Background tree -- total number of entries: 5449 \n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Preselection: (will NOT affect number of requested training and testing events)\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :     No preselection cuts applied on event classes\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Weight renormalisation mode: \"NumEvents\": renormalises all event classes \n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :  such that the effective (weighted) number of events in each class equals the respective \n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :  number of events (entries) that you demanded in PrepareTrainingAndTestTree(\"\",\"nTrain_Signal=.. )\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :  ... i.e. such that Sum[i=1..N_j]{w_i} = N_j, j=0,1,2...\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :  ... (note that N_j is the sum of TRAINING events (nTrain_j...with j=Signal,Background..\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :  ..... Testing events are not renormalised nor included in the renormalisation factor! )\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : --> Rescale Signal     event weights by factor: 1\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : --> Rescale Background event weights by factor: 1\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Number of training and testing events after rescaling:\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : ---------------------------------------------------------------------------\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Signal     -- training events            : 3000 (sum of weights: 3000) - requested were 3000 events\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Signal     -- testing events             : 1449 (sum of weights: 1449) - requested were 1449 events\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Signal     -- training and testing events: 4449 (sum of weights: 4449)\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Background -- training events            : 3000 (sum of weights: 3000) - requested were 3000 events\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Background -- testing events             : 1449 (sum of weights: 1449) - requested were 1449 events\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Background -- training and testing events: 4449 (sum of weights: 4449)\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Create internal training tree\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] : Create internal testing tree\n",
      "--- DataSetInfo              : Dataset[mymc-dataset2] : Correlation matrix (Signal):\n",
      "--- DataSetInfo              : ----------------------------------------------------------------\n",
      "--- DataSetInfo              :               var0    var1    var2 var0-var1 var0*var2 var1+var2\n",
      "--- DataSetInfo              :      var0:  +1.000  -0.008  +0.012    +0.857    +0.922    +0.001\n",
      "--- DataSetInfo              :      var1:  -0.008  +1.000  +0.028    -0.521    +0.003    +0.810\n",
      "--- DataSetInfo              :      var2:  +0.012  +0.028  +1.000    -0.004    +0.331    +0.609\n",
      "--- DataSetInfo              : var0-var1:  +0.857  -0.521  -0.004    +1.000    +0.785    -0.416\n",
      "--- DataSetInfo              : var0*var2:  +0.922  +0.003  +0.331    +0.785    +1.000    +0.197\n",
      "--- DataSetInfo              : var1+var2:  +0.001  +0.810  +0.609    -0.416    +0.197    +1.000\n",
      "--- DataSetInfo              : ----------------------------------------------------------------\n",
      "--- DataSetInfo              : Dataset[mymc-dataset2] : Correlation matrix (Background):\n",
      "--- DataSetInfo              : ----------------------------------------------------------------\n",
      "--- DataSetInfo              :               var0    var1    var2 var0-var1 var0*var2 var1+var2\n",
      "--- DataSetInfo              :      var0:  +1.000  -0.008  -0.004    +0.652    +0.664    -0.009\n",
      "--- DataSetInfo              :      var1:  -0.008  +1.000  +0.031    -0.764    +0.018    +0.746\n",
      "--- DataSetInfo              :      var2:  -0.004  +0.031  +1.000    -0.026    +0.692    +0.688\n",
      "--- DataSetInfo              : var0-var1:  +0.652  -0.764  -0.026    +1.000    +0.415    -0.572\n",
      "--- DataSetInfo              : var0*var2:  +0.664  +0.018  +0.692    +0.415    +1.000    +0.474\n",
      "--- DataSetInfo              : var1+var2:  -0.009  +0.746  +0.688    -0.572    +0.474    +1.000\n",
      "--- DataSetInfo              : ----------------------------------------------------------------\n",
      "--- DataSetFactory           : Dataset[mymc-dataset2] :  \n"
     ]
    }
   ],
   "source": [
    "// TMVA ANN: MLP Multilayer Perceptrons\n",
    "factory->BookMethod(loader1,TMVA::Types::kMLP, \"MLP\", \"!V:NeuronType=tanh:VarTransform=N:NCycles=150:HiddenLayers=N+5:TestRate=5:!UseRegulator\" );\n",
    "\n",
    "//Boosted Decision Trees\n",
    "factory->BookMethod(loader2,TMVA::Types::kBDT, \"BDT\",\"!V:NTrees=5:MinNodeSize=2.5%:MaxDepth=2:BoostType=AdaBoost:AdaBoostBeta=0.5:UseBaggedBoost:BaggedSampleFraction=0.5:SeparationType=GiniIndex:nCuts=20\" );"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training the Methods"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- Factory                  :  \n",
      "--- Factory                  : Train all methods for Classification ...\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Splitmode is: \"RANDOM\" the mixmode is: \"SAMEASSPLITMODE\"\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Create training and testing trees -- looping over class \"Signal\" ...\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Weight expression for class 'Signal': \"\"\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Create training and testing trees -- looping over class \"Background\" ...\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Weight expression for class 'Background': \"\"\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Number of events in input trees (after possible flattening of arrays):\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :     Signal          -- number of events       : 5449   / sum of weights: 5449 \n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :     Background      -- number of events       : 5449   / sum of weights: 5449 \n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :     Signal     tree -- total number of entries: 5449 \n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :     Background tree -- total number of entries: 5449 \n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Preselection: (will NOT affect number of requested training and testing events)\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :     No preselection cuts applied on event classes\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Weight renormalisation mode: \"NumEvents\": renormalises all event classes \n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :  such that the effective (weighted) number of events in each class equals the respective \n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :  number of events (entries) that you demanded in PrepareTrainingAndTestTree(\"\",\"nTrain_Signal=.. )\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :  ... i.e. such that Sum[i=1..N_j]{w_i} = N_j, j=0,1,2...\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :  ... (note that N_j is the sum of TRAINING events (nTrain_j...with j=Signal,Background..\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :  ..... Testing events are not renormalised nor included in the renormalisation factor! )\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : --> Rescale Signal     event weights by factor: 1\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : --> Rescale Background event weights by factor: 1\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Number of training and testing events after rescaling:\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : ---------------------------------------------------------------------------\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Signal     -- training events            : 3000 (sum of weights: 3000) - requested were 3000 events\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Signal     -- testing events             : 1449 (sum of weights: 1449) - requested were 1449 events\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Signal     -- training and testing events: 4449 (sum of weights: 4449)\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Background -- training events            : 3000 (sum of weights: 3000) - requested were 3000 events\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Background -- testing events             : 1449 (sum of weights: 1449) - requested were 1449 events\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Background -- training and testing events: 4449 (sum of weights: 4449)\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Create internal training tree\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] : Create internal testing tree\n",
      "--- DataSetInfo              : Dataset[mymc-dataset1] : Correlation matrix (Signal):\n",
      "--- DataSetInfo              : --------------------------------\n",
      "--- DataSetInfo              :             var0    var1    var2\n",
      "--- DataSetInfo              :    var0:  +1.000  -0.008  +0.012\n",
      "--- DataSetInfo              :    var1:  -0.008  +1.000  +0.028\n",
      "--- DataSetInfo              :    var2:  +0.012  +0.028  +1.000\n",
      "--- DataSetInfo              : --------------------------------\n",
      "--- DataSetInfo              : Dataset[mymc-dataset1] : Correlation matrix (Background):\n",
      "--- DataSetInfo              : --------------------------------\n",
      "--- DataSetInfo              :             var0    var1    var2\n",
      "--- DataSetInfo              :    var0:  +1.000  -0.008  -0.004\n",
      "--- DataSetInfo              :    var1:  -0.008  +1.000  +0.031\n",
      "--- DataSetInfo              :    var2:  -0.004  +0.031  +1.000\n",
      "--- DataSetInfo              : --------------------------------\n",
      "--- DataSetFactory           : Dataset[mymc-dataset1] :  \n",
      "--- Factory                  : \n",
      "--- Factory                  : current transformation string: 'I'\n",
      "--- Factory                  : Dataset[mymc-dataset1] : Create Transformation \"I\" with events from all classes.\n",
      "--- Id                       : Transformation, Variable selection : \n",
      "--- Id                       : Input : variable 'var0' (index=0).   <---> Output : variable 'var0' (index=0).\n",
      "--- Id                       : Input : variable 'var1' (index=1).   <---> Output : variable 'var1' (index=1).\n",
      "--- Id                       : Input : variable 'var2' (index=2).   <---> Output : variable 'var2' (index=2).\n",
      "--- Id                       : Preparing the Identity transformation...\n",
      "--- TFHandler_Factory        : -----------------------------------------------------------\n",
      "--- TFHandler_Factory        : Variable        Mean        RMS   [        Min        Max ]\n",
      "--- TFHandler_Factory        : -----------------------------------------------------------\n",
      "--- TFHandler_Factory        :     var0:     3.0763     1.7033   [   0.043380     9.9950 ]\n",
      "--- TFHandler_Factory        :     var1:     1.5553     1.4223   [ 0.00032097     4.9967 ]\n",
      "--- TFHandler_Factory        :     var2:     3.5090     1.1966   [  0.0038420     5.0000 ]\n",
      "--- TFHandler_Factory        : -----------------------------------------------------------\n",
      "--- TFHandler_Factory        : Plot event variables for Id\n",
      "--- TFHandler_Factory        : Create scatter and profile plots in target-file directory: \n",
      "--- TFHandler_Factory        : TMVAOutputDL.root:/mymc-dataset1/InputVariables_Id/CorrelationPlots\n",
      "--- TFHandler_Factory        :  \n",
      "--- TFHandler_Factory        : Ranking input variables (method unspecific)...\n",
      "--- IdTransformation         : Ranking result (top variable is best ranked)\n",
      "--- IdTransformation         : -----------------------------\n",
      "--- IdTransformation         : Rank : Variable  : Separation\n",
      "--- IdTransformation         : -----------------------------\n",
      "--- IdTransformation         :    1 : var1      : 1.432e-01\n",
      "--- IdTransformation         :    2 : var2      : 1.081e-01\n",
      "--- IdTransformation         :    3 : var0      : 1.014e-01\n",
      "--- IdTransformation         : -----------------------------\n",
      "--- Factory                  : Train method: MLP for Classification\n",
      "--- Norm                     : Preparing the transformation.\n",
      "--- TFHandler_MLP            : -----------------------------------------------------------\n",
      "--- TFHandler_MLP            : Variable        Mean        RMS   [        Min        Max ]\n",
      "--- TFHandler_MLP            : -----------------------------------------------------------\n",
      "--- TFHandler_MLP            :     var0:   -0.39047    0.34232   [    -1.0000     1.0000 ]\n",
      "--- TFHandler_MLP            :     var1:   -0.37756    0.56933   [    -1.0000     1.0000 ]\n",
      "--- TFHandler_MLP            :     var2:    0.40316    0.47901   [    -1.0000     1.0000 ]\n",
      "--- TFHandler_MLP            : -----------------------------------------------------------\n",
      "--- MLP                      : Dataset[mymc-dataset1] : Begin training\n",
      "--- MLP                      : Training Network\n",
      "--- MLP                      : Dataset[mymc-dataset1] : End of training                                              \n",
      "--- MLP                      : Dataset[mymc-dataset1] : Elapsed time for training with 6000 events: \u001b[1;31m4.84 sec\u001b[0m         \n",
      "--- MLP                      : Dataset[mymc-dataset1] : Create MVA output for Dataset[mymc-dataset1] : classification on training sample\n",
      "--- MLP                      : Dataset[mymc-dataset1] : Evaluation of MLP on training sample (6000 events)\n",
      "--- MLP                      : Dataset[mymc-dataset1] : Elapsed time for evaluation of 6000 events: \u001b[1;31m0.0101 sec\u001b[0m       \n",
      "--- MLP                      : Dataset[mymc-dataset1] : Creating weight file in xml format: \u001b[0;36mmymc-dataset1/weights/TMVAClassification_MLP.weights.xml\u001b[0m\n",
      "--- MLP                      : Dataset[mymc-dataset1] : Creating standalone response class: \u001b[0;36mmymc-dataset1/weights/TMVAClassification_MLP.class.C\u001b[0m\n",
      "--- MLP                      : Write special histos to file: TMVAOutputDL.root:/mymc-dataset1/MLP\n",
      "--- Factory                  : Training finished\n",
      "--- Factory                  : \n",
      "--- Factory                  : Ranking input variables (method specific)...\n",
      "--- MLP                      : Ranking result (top variable is best ranked)\n",
      "--- MLP                      : -----------------------------\n",
      "--- MLP                      : Rank : Variable  : Importance\n",
      "--- MLP                      : -----------------------------\n",
      "--- MLP                      :    1 : var0      : 2.780e+01\n",
      "--- MLP                      :    2 : var1      : 2.339e+01\n",
      "--- MLP                      :    3 : var2      : 6.350e-01\n",
      "--- MLP                      : -----------------------------\n",
      "--- Factory                  : \n",
      "--- Factory                  : === Destroy and recreate all methods via weight files for testing ===\n",
      "--- Factory                  : \n",
      "--- MethodBase               : Dataset[mymc-dataset1] : Reading weight file: \u001b[0;36mmymc-dataset1/weights/TMVAClassification_MLP.weights.xml\u001b[0m\n",
      "--- MLP                      : Dataset[mymc-dataset1] : Read method \"MLP\" of type \"MLP\"\n",
      "--- MLP                      : Dataset[mymc-dataset1] : MVA method was trained with TMVA Version: 4.2.1\n",
      "--- MLP                      : Dataset[mymc-dataset1] : MVA method was trained with ROOT Version: 6.05/03\n",
      "--- MLP                      : Building Network\n",
      "--- MLP                      : Initializing weights\n",
      "--- Factory                  : \n",
      "--- Factory                  : current transformation string: 'I'\n",
      "--- Factory                  : Dataset[mymc-dataset2] : Create Transformation \"I\" with events from all classes.\n",
      "--- Id                       : Transformation, Variable selection : \n",
      "--- Id                       : Input : variable 'var0' (index=0).   <---> Output : variable 'var0' (index=0).\n",
      "--- Id                       : Input : variable 'var1' (index=1).   <---> Output : variable 'var1' (index=1).\n",
      "--- Id                       : Input : variable 'var2' (index=2).   <---> Output : variable 'var2' (index=2).\n",
      "--- Id                       : Input : variable 'var3' (index=3).   <---> Output : variable 'var3' (index=3).\n",
      "--- Id                       : Input : variable 'var4' (index=4).   <---> Output : variable 'var4' (index=4).\n",
      "--- Id                       : Input : variable 'var5' (index=5).   <---> Output : variable 'var5' (index=5).\n",
      "--- Id                       : Preparing the Identity transformation...\n",
      "--- TFHandler_Factory        : -----------------------------------------------------------\n",
      "--- TFHandler_Factory        : Variable        Mean        RMS   [        Min        Max ]\n",
      "--- TFHandler_Factory        : -----------------------------------------------------------\n",
      "--- TFHandler_Factory        :     var0:     3.0763     1.7033   [   0.043380     9.9950 ]\n",
      "--- TFHandler_Factory        :     var1:     1.5553     1.4223   [ 0.00032097     4.9967 ]\n",
      "--- TFHandler_Factory        :     var2:     3.5090     1.1966   [  0.0038420     5.0000 ]\n",
      "--- TFHandler_Factory        :     var3:     1.5210     2.2372   [    -3.9769     9.9365 ]\n",
      "--- TFHandler_Factory        :     var4:     10.822     7.5844   [   0.014320     48.593 ]\n",
      "--- TFHandler_Factory        :     var5:     5.0643     1.7847   [    0.14162     9.9525 ]\n",
      "--- TFHandler_Factory        : -----------------------------------------------------------\n",
      "--- TFHandler_Factory        : Plot event variables for Id\n",
      "--- TFHandler_Factory        : Create scatter and profile plots in target-file directory: \n",
      "--- TFHandler_Factory        : TMVAOutputDL.root:/mymc-dataset2/InputVariables_Id/CorrelationPlots\n",
      "--- TFHandler_Factory        :  \n",
      "--- TFHandler_Factory        : Ranking input variables (method unspecific)...\n",
      "--- IdTransformation         : Ranking result (top variable is best ranked)\n",
      "--- IdTransformation         : -----------------------------\n",
      "--- IdTransformation         : Rank : Variable  : Separation\n",
      "--- IdTransformation         : -----------------------------\n",
      "--- IdTransformation         :    1 : var3      : 2.599e-01\n",
      "--- IdTransformation         :    2 : var1      : 1.432e-01\n",
      "--- IdTransformation         :    3 : var2      : 1.081e-01\n",
      "--- IdTransformation         :    4 : var0      : 1.014e-01\n",
      "--- IdTransformation         :    5 : var5      : 5.893e-02\n",
      "--- IdTransformation         :    6 : var4      : 5.046e-02\n",
      "--- IdTransformation         : -----------------------------\n",
      "--- Factory                  : Train method: BDT for Classification\n",
      "--- BDT                      : Dataset[mymc-dataset2] : Begin training\n",
      "--- BDT                      :  found and suggest the following possible pre-selection cuts \n",
      "--- BDT                      : as option DoPreselection was not used, these cuts however will not be performed, but the training will see the full sample\n",
      "--- BDT                      :  found cut: Sig if var 0 > 7.56526\n",
      "--- BDT                      :  found cut: Bkg if var 3 < -1.09256\n",
      "--- BDT                      :  found cut: Sig if var 3 > 7.48204\n",
      "--- BDT                      : <InitEventSample> For classification trees, \n",
      "--- BDT                      :  the effective number of backgrounds is scaled to match \n",
      "--- BDT                      :  the signal. Othersise the first boosting step would do 'just that'!\n",
      "--- BDT                      : re-normlise events such that Sig and Bkg have respective sum of weights = 1\n",
      "--- BDT                      :   sig->sig*1ev. bkg->bkg*1ev.\n",
      "--- BDT                      : #events: (reweighted) sig: 3000 bkg: 3000\n",
      "--- BDT                      : #events: (unweighted) sig: 3000 bkg: 3000\n",
      "--- BDT                      : Training 5 Decision Trees ... patience please\n",
      "--- BinaryTree               : The minimal node size MinNodeSize=2.5 fMinNodeSize=2.5% is translated to an actual number of events = 76.175 for the training sample size of 3047\n",
      "--- BinaryTree               : Note: This number will be taken as absolute minimum in the node, \n",
      "--- BinaryTree               :       in terms of 'weighted events' and unweighted ones !! \n",
      "--- BDT                      : <Train> elapsed time: \u001b[1;31m0.0241 sec\u001b[0m                              \n",
      "--- BDT                      : <Train> average number of nodes (w/o pruning) : 4\n",
      "--- BDT                      : Dataset[mymc-dataset2] : End of training                                              \n",
      "--- BDT                      : Dataset[mymc-dataset2] : Elapsed time for training with 6000 events: \u001b[1;31m0.0586 sec\u001b[0m         \n",
      "--- BDT                      : Dataset[mymc-dataset2] : Create MVA output for Dataset[mymc-dataset2] : classification on training sample\n",
      "--- BDT                      : Dataset[mymc-dataset2] : Evaluation of BDT on training sample (6000 events)\n",
      "--- BDT                      : Dataset[mymc-dataset2] : Elapsed time for evaluation of 6000 events: \u001b[1;31m0.00401 sec\u001b[0m       \n",
      "--- BDT                      : Dataset[mymc-dataset2] : Creating weight file in xml format: \u001b[0;36mmymc-dataset2/weights/TMVAClassification_BDT.weights.xml\u001b[0m\n",
      "--- BDT                      : Dataset[mymc-dataset2] : Creating standalone response class: \u001b[0;36mmymc-dataset2/weights/TMVAClassification_BDT.class.C\u001b[0m\n",
      "--- BDT                      : Write monitoring histograms to file: TMVAOutputDL.root:/mymc-dataset2/BDT\n",
      "--- Factory                  : Training finished\n",
      "--- Factory                  : \n",
      "--- Factory                  : Ranking input variables (method specific)...\n",
      "--- BDT                      : Ranking result (top variable is best ranked)\n",
      "--- BDT                      : --------------------------------------\n",
      "--- BDT                      : Rank : Variable  : Variable Importance\n",
      "--- BDT                      : --------------------------------------\n",
      "--- BDT                      :    1 : var3      : 4.718e-01\n",
      "--- BDT                      :    2 : var2      : 2.594e-01\n",
      "--- BDT                      :    3 : var0      : 1.709e-01\n",
      "--- BDT                      :    4 : var1      : 9.794e-02\n",
      "--- BDT                      :    5 : var4      : 0.000e+00\n",
      "--- BDT                      :    6 : var5      : 0.000e+00\n",
      "--- BDT                      : --------------------------------------\n",
      "--- Factory                  : \n",
      "--- Factory                  : === Destroy and recreate all methods via weight files for testing ===\n",
      "--- Factory                  : \n",
      "--- MethodBase               : Dataset[mymc-dataset2] : Reading weight file: \u001b[0;36mmymc-dataset2/weights/TMVAClassification_BDT.weights.xml\u001b[0m\n",
      "--- BDT                      : Dataset[mymc-dataset2] : Read method \"BDT\" of type \"BDT\"\n",
      "--- BDT                      : Dataset[mymc-dataset2] : MVA method was trained with TMVA Version: 4.2.1\n",
      "--- BDT                      : Dataset[mymc-dataset2] : MVA method was trained with ROOT Version: 6.05/03\n"
     ]
    }
   ],
   "source": [
    "factory->TrainAllMethods();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Testing and Evaluating the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- Factory                  : Test all methods...\n",
      "--- Factory                  : Test method: MLP for Classification performance\n",
      "--- MLP                      : Dataset[mymc-dataset1] : Evaluation of MLP on testing sample (2898 events)\n",
      "--- MLP                      : Dataset[mymc-dataset1] : Elapsed time for evaluation of 2898 events: \u001b[1;31m0.00825 sec\u001b[0m       \n",
      "--- Factory                  : Test method: BDT for Classification performance\n",
      "--- BDT                      : Dataset[mymc-dataset2] : Evaluation of BDT on testing sample (2898 events)\n",
      "--- BDT                      : Dataset[mymc-dataset2] : Elapsed time for evaluation of 2898 events: \u001b[1;31m0.00275 sec\u001b[0m       \n",
      "--- Factory                  : Evaluate all methods...\n",
      "--- Factory                  : Evaluate classifier: MLP\n",
      "--- TFHandler_MLP            : -----------------------------------------------------------\n",
      "--- TFHandler_MLP            : Variable        Mean        RMS   [        Min        Max ]\n",
      "--- TFHandler_MLP            : -----------------------------------------------------------\n",
      "--- TFHandler_MLP            :     var0:   -0.39553    0.33943   [   -0.99628     1.0000 ]\n",
      "--- TFHandler_MLP            :     var1:   -0.35571    0.57297   [    -1.0001    0.99955 ]\n",
      "--- TFHandler_MLP            :     var2:    0.40013    0.48671   [    -1.0012    0.99971 ]\n",
      "--- TFHandler_MLP            : -----------------------------------------------------------\n",
      "--- MLP                      : Dataset[mymc-dataset1] : Loop over test events and fill histograms with classifier response...\n",
      "--- Factory                  : Write evaluation histograms to file\n",
      "--- TFHandler_MLP            : Plot event variables for MLP\n",
      "--- TFHandler_MLP            : -----------------------------------------------------------\n",
      "--- TFHandler_MLP            : Variable        Mean        RMS   [        Min        Max ]\n",
      "--- TFHandler_MLP            : -----------------------------------------------------------\n",
      "--- TFHandler_MLP            :     var0:   -0.39553    0.33943   [   -0.99628     1.0000 ]\n",
      "--- TFHandler_MLP            :     var1:   -0.35571    0.57297   [    -1.0001    0.99955 ]\n",
      "--- TFHandler_MLP            :     var2:    0.40013    0.48671   [    -1.0012    0.99971 ]\n",
      "--- TFHandler_MLP            : -----------------------------------------------------------\n",
      "--- TFHandler_MLP            : Create scatter and profile plots in target-file directory: \n",
      "--- TFHandler_MLP            : TMVAOutputDL.root:/mymc-dataset1/Method_MLP/MLP/CorrelationPlots\n",
      "--- Factory                  : \n",
      "--- Factory                  : Evaluation results ranked by best signal efficiency and purity (area)\n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : DataSet              MVA              Signal efficiency at bkg eff.(error):                | Sepa-    Signifi- \n",
      "--- Factory                  : Name:                Method:          @B=0.01    @B=0.10    @B=0.30    ROC-integ    ROCCurve| ration:  cance:   \n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : mymc-dataset1        MLP            : 0.197(10)  0.631(12)  0.926(06)    0.896       0.894 | 0.498    1.371\n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : \n",
      "--- Factory                  : Testing efficiency compared to training efficiency (overtraining check)\n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : DataSet              MVA              Signal efficiency: from test sample (from training sample) \n",
      "--- Factory                  : Name:                Method:          @B=0.01             @B=0.10            @B=0.30   \n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : mymc-dataset1        MLP            : 0.197 (0.212)       0.631 (0.620)      0.926 (0.899)\n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : \n",
      "--- Dataset:mymc-dataset1    : Dataset[mymc-dataset1] : Created tree 'TestTree' with 2898 events\n",
      "--- Dataset:mymc-dataset1    : Dataset[mymc-dataset1] : Created tree 'TrainTree' with 6000 events\n",
      "--- Factory                  : Evaluate classifier: BDT\n",
      "--- BDT                      : Dataset[mymc-dataset2] : Loop over test events and fill histograms with classifier response...\n",
      "--- Factory                  : Write evaluation histograms to file\n",
      "--- TFHandler_BDT            : Plot event variables for BDT\n",
      "--- TFHandler_BDT            : -----------------------------------------------------------\n",
      "--- TFHandler_BDT            : Variable        Mean        RMS   [        Min        Max ]\n",
      "--- TFHandler_BDT            : -----------------------------------------------------------\n",
      "--- TFHandler_BDT            :     var0:     3.0511     1.6889   [   0.061879     9.9950 ]\n",
      "--- TFHandler_BDT            :     var1:     1.6099     1.4314   [ 9.9424e-05     4.9956 ]\n",
      "--- TFHandler_BDT            :     var2:     3.5015     1.2158   [ 0.00088645     4.9993 ]\n",
      "--- TFHandler_BDT            :     var3:     1.4412     2.2459   [    -3.6483     9.9615 ]\n",
      "--- TFHandler_BDT            :     var4:     10.770     7.6139   [ 0.00088366     48.019 ]\n",
      "--- TFHandler_BDT            :     var5:     5.1113     1.8033   [    0.17973     9.9030 ]\n",
      "--- TFHandler_BDT            : -----------------------------------------------------------\n",
      "--- TFHandler_BDT            : Create scatter and profile plots in target-file directory: \n",
      "--- TFHandler_BDT            : TMVAOutputDL.root:/mymc-dataset2/Method_BDT/BDT/CorrelationPlots\n",
      "--- Factory                  : \n",
      "--- Factory                  : Evaluation results ranked by best signal efficiency and purity (area)\n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : DataSet              MVA              Signal efficiency at bkg eff.(error):                | Sepa-    Signifi- \n",
      "--- Factory                  : Name:                Method:          @B=0.01    @B=0.10    @B=0.30    ROC-integ    ROCCurve| ration:  cance:   \n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : mymc-dataset2        BDT            : 0.000(00)  0.481(13)  0.742(11)    0.814       0.814 | 0.337    0.965\n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : \n",
      "--- Factory                  : Testing efficiency compared to training efficiency (overtraining check)\n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : DataSet              MVA              Signal efficiency: from test sample (from training sample) \n",
      "--- Factory                  : Name:                Method:          @B=0.01             @B=0.10            @B=0.30   \n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : mymc-dataset2        BDT            : 0.000 (0.000)       0.481 (0.443)      0.742 (0.720)\n",
      "--- Factory                  : -------------------------------------------------------------------------------------------------------------------\n",
      "--- Factory                  : \n",
      "--- Dataset:mymc-dataset2    : Dataset[mymc-dataset2] : Created tree 'TestTree' with 2898 events\n",
      "--- Dataset:mymc-dataset2    : Dataset[mymc-dataset2] : Created tree 'TrainTree' with 6000 events\n",
      "--- Factory                  :   \n",
      "--- Factory                  : \u001b[1mThank you for using TMVA!\u001b[0m\n",
      "--- Factory                  : \u001b[1mFor citation information, please visit: http://tmva.sf.net/citeTMVA.html\u001b[0m\n",
      "==> TMVAClassification is done!\n"
     ]
    }
   ],
   "source": [
    "factory->TestAllMethods();\n",
    "\n",
    "factory->EvaluateAllMethods();    \n",
    "\n",
    "// Save the output\n",
    "outputFile->Close();\n",
    "\n",
    "std::cout << \"==> TMVAClassification is done!\" << std::endl;\n",
    "\n",
    "delete factory;\n",
    "delete loader1;\n",
    "delete loader2;"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ploting ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- Found 1 classifier types\n",
      "--- Found 1 instance(s) of the method Method_MLP\n",
      "--- Found 1 classifier types\n",
      "--- Found 1 instance(s) of the method Method_MLP\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "TColor::TColor:0: RuntimeWarning: color 1367 already defined\n",
      "Info in <TCanvas::Print>: eps file plots/rejBvsS.eps has been created\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoYAAAHYCAYAAADQ5iljAAAABmJLR0QAAAAAAAD5Q7t/AAAgAElE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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "TFile *file=TFile::Open(\"TMVAOutputDL.root\");\n",
    "\n",
    "gDirectory->cd(\"mymc-dataset1\");\n",
    "TMVA::plot_efficiencies(file,  2,gDirectory);\n",
    "c->Draw();\n",
    "c->Draw();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--- Found 1 classifier types\n",
      "--- Found 1 instance(s) of the method Method_BDT\n",
      "--- Found 1 classifier types\n",
      "--- Found 1 instance(s) of the method Method_BDT\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "TCanvas::Constructor:0: RuntimeWarning: Deleting canvas with same name: c\n",
      "TColor::TColor:0: RuntimeWarning: color 1367 already defined\n",
      "Info in <TCanvas::Print>: eps file plots/rejBvsS.eps has been created\n"
     ]
    }
   ],
   "source": [
    "file->cd();\n",
    "gDirectory->cd(\"mymc-dataset2\");\n",
    "TMVA::plot_efficiencies(file,  2,gDirectory);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoYAAAHYCAYAAADQ5iljAAAABmJLR0QAAAAAAAD5Q7t/AAAgAElE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\nAIAjMIYAAAAAAICIYAwBAAAAAMCR/x9uPlc/eL91jgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "c->Draw();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "ROOT Prompt",
   "language": "c++",
   "name": "root"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}