Logo ROOT  
masterv6.32v6.30v6.28v6.26v6.24v6.22v6.20v6.18v6.16v6.14v6.12v6.10v6.08v6.06

Reference Guide
 
•All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Properties Friends Macros Modules Pages
Loading...
Searching...
No Matches
TMVAMinimalClassification.C File Reference
Tutorials » TMVA tutorials

Detailed Description

View in nbviewer Open in SWAN
Minimal self-contained example for setting up TMVA with binary classification.

This is intended as a simple foundation to build on. It assumes you are familiar with TMVA already. As such concepts like the Factory, the DataLoader and others are not explained. For descriptions and tutorials use the TMVA online manual https://root.cern/manual/tmva/ or the more detailed examples provided with TMVA e.g. TMVAClassification.C. or the TMVA Users Guide https://github.com/root-project/root/blob/master/documentation/tmva/UsersGuide/TMVAUsersGuide.pdf

Sets up a minimal binary classification example with two slightly overlapping 2-D gaussian distributions and trains a BDT classifier to discriminate the data.

<HEADER> DataSetInfo : [dataset] : Added class "Signal"
: Add Tree of type Signal with 1000 events
<HEADER> DataSetInfo : [dataset] : Added class "Background"
: Add Tree of type Background with 1000 events
<HEADER> Factory : Booking method: BDT
:
: Rebuilding Dataset dataset
: Building event vectors for type 2 Signal
: Dataset[dataset] : create input formulas for tree
: Building event vectors for type 2 Background
: Dataset[dataset] : create input formulas for tree
<HEADER> DataSetFactory : [dataset] : Number of events in input trees
:
:
: Dataset[dataset] : Weight renormalisation mode: "EqualNumEvents": renormalises all event classes ...
: Dataset[dataset] : such that the effective (weighted) number of events in each class is the same
: Dataset[dataset] : (and equals the number of events (entries) given for class=0 )
: Dataset[dataset] : ... i.e. such that Sum[i=1..N_j]{w_i} = N_classA, j=classA, classB, ...
: Dataset[dataset] : ... (note that N_j is the sum of TRAINING events
: Dataset[dataset] : ..... Testing events are not renormalised nor included in the renormalisation factor!)
: Number of training and testing events
: ---------------------------------------------------------------------------
: Signal -- training events : 500
: Signal -- testing events : 500
: Signal -- training and testing events: 1000
: Background -- training events : 500
: Background -- testing events : 500
: Background -- training and testing events: 1000
:
<HEADER> DataSetInfo : Correlation matrix (Signal):
: ------------------------
: x y
: x: +1.000 -0.034
: y: -0.034 +1.000
: ------------------------
<HEADER> DataSetInfo : Correlation matrix (Background):
: ------------------------
: x y
: x: +1.000 -0.057
: y: -0.057 +1.000
: ------------------------
<HEADER> DataSetFactory : [dataset] :
:
<HEADER> Factory : Train all methods
<HEADER> Factory : [dataset] : Create Transformation "I" with events from all classes.
:
<HEADER> : Transformation, Variable selection :
: Input : variable 'x' <---> Output : variable 'x'
: Input : variable 'y' <---> Output : variable 'y'
<HEADER> TFHandler_Factory : Variable Mean RMS [ Min Max ]
: -----------------------------------------------------------
: x: 1.0188 0.56914 [ 0.00044777 1.9995 ]
: y: 1.5175 0.74452 [ 0.0054384 2.9981 ]
: -----------------------------------------------------------
: Ranking input variables (method unspecific)...
<HEADER> IdTransformation : Ranking result (top variable is best ranked)
: --------------------------
: Rank : Variable : Separation
: --------------------------
: 1 : y : 5.193e-01
: 2 : x : 5.434e-02
: --------------------------
<HEADER> Factory : Train method: BDT for Classification
:
<HEADER> BDT : #events: (reweighted) sig: 500 bkg: 500
: #events: (unweighted) sig: 500 bkg: 500
: Training 800 Decision Trees ... patience please
: Elapsed time for training with 1000 events: 0.403 sec
<HEADER> BDT : [dataset] : Evaluation of BDT on training sample (1000 events)
: Elapsed time for evaluation of 1000 events: 0.0693 sec
: Creating xml weight file: dataset/weights/_BDT.weights.xml
: Creating standalone class: dataset/weights/_BDT.class.C
: out.root:/dataset/Method_BDT/BDT
<HEADER> Factory : Training finished
:
: Ranking input variables (method specific)...
<HEADER> BDT : Ranking result (top variable is best ranked)
: -----------------------------------
: Rank : Variable : Variable Importance
: -----------------------------------
: 1 : x : 5.205e-01
: 2 : y : 4.795e-01
: -----------------------------------
<HEADER> Factory : === Destroy and recreate all methods via weight files for testing ===
:
: Reading weight file: dataset/weights/_BDT.weights.xml
<HEADER> Factory : Test all methods
<HEADER> Factory : Test method: BDT for Classification performance
:
<HEADER> BDT : [dataset] : Evaluation of BDT on testing sample (1000 events)
: Elapsed time for evaluation of 1000 events: 0.0558 sec
<HEADER> Factory : Evaluate all methods
<HEADER> Factory : Evaluate classifier: BDT
:
<HEADER> BDT : [dataset] : Loop over test events and fill histograms with classifier response...
:
<HEADER> TFHandler_BDT : Variable Mean RMS [ Min Max ]
: -----------------------------------------------------------
: x: 1.0177 0.58666 [ 0.0011208 1.9999 ]
: y: 1.4705 0.77233 [ 0.024000 2.9933 ]
: -----------------------------------------------------------
:
: Evaluation results ranked by best signal efficiency and purity (area)
: -------------------------------------------------------------------------------------------------------------------
: DataSet MVA
: Name: Method: ROC-integ
: dataset BDT : 0.875
: -------------------------------------------------------------------------------------------------------------------
:
: Testing efficiency compared to training efficiency (overtraining check)
: -------------------------------------------------------------------------------------------------------------------
: DataSet MVA Signal efficiency: from test sample (from training sample)
: Name: Method: @B=0.01 @B=0.10 @B=0.30
: -------------------------------------------------------------------------------------------------------------------
: dataset BDT : 0.485 (0.705) 0.609 (0.797) 0.794 (0.924)
: -------------------------------------------------------------------------------------------------------------------
:
<HEADER> Dataset:dataset : Created tree 'TestTree' with 1000 events
:
<HEADER> Dataset:dataset : Created tree 'TrainTree' with 1000 events
:
<HEADER> Factory : Thank you for using TMVA!
: For citation information, please visit: http://tmva.sf.net/citeTMVA.html
#include "TMVA/DataLoader.h"
#include "TMVA/Factory.h"
#include "TFile.h"
#include "TString.h"
#include "TTree.h"
//
// Helper function to generate 2-D gaussian data points and fill to a ROOT
// TTree.
//
// Arguments:
// nPoints Number of points to generate.
// offset Mean of the generated numbers
// scale Standard deviation of the generated numbers.
// seed Seed for random number generator. Use `seed=0` for random
// seed.
// Returns a TTree ready to be used as input to TMVA.
//
TTree *genTree(Int_t nPoints, Double_t offset, Double_t scale, UInt_t seed = 100)
{
TRandom rng(seed);
Double_t x = 0;
Double_t y = 0;
TTree *data = new TTree();
data->Branch("x", &x, "x/D");
data->Branch("y", &y, "y/D");
for (Int_t n = 0; n < nPoints; ++n) {
x = rng.Rndm() * scale;
y = offset + rng.Rndm() * scale;
data->Fill();
}
// Important: Disconnects the tree from the memory locations of x and y.
data->ResetBranchAddresses();
return data;
}
//
// Minimal setup for performing binary classification in TMVA.
//
// Modify the setup to your liking and run with
// `root -l -b -q TMVAMinimalClassification.C`.
// This will generate an output file "out.root" that can be viewed with
// `root -l -e 'TMVA::TMVAGui("out.root")'`.
//
void TMVAMinimalClassification()
{
TString outputFilename = "out.root";
TFile *outFile = new TFile(outputFilename, "RECREATE");
// Data generation
TTree *signalTree = genTree(1000, 0.0, 2.0, 100);
TTree *backgroundTree = genTree(1000, 1.0, 2.0, 101);
TString factoryOptions = "AnalysisType=Classification";
TMVA::Factory factory{"", outFile, factoryOptions};
TMVA::DataLoader dataloader{"dataset"};
// Data specification
dataloader.AddVariable("x", 'D');
dataloader.AddVariable("y", 'D');
dataloader.AddSignalTree(signalTree, 1.0);
dataloader.AddBackgroundTree(backgroundTree, 1.0);
TCut signalCut = "";
TCut backgroundCut = "";
TString datasetOptions = "SplitMode=Random";
dataloader.PrepareTrainingAndTestTree(signalCut, backgroundCut, datasetOptions);
// Method specification
TString methodOptions = "";
factory.BookMethod(&dataloader, TMVA::Types::kBDT, "BDT", methodOptions);
// Training and Evaluation
factory.TrainAllMethods();
factory.TestAllMethods();
factory.EvaluateAllMethods();
// Clean up
outFile->Close();
delete outFile;
delete signalTree;
delete backgroundTree;
}
Int_t
int Int_t
Definition RtypesCore.h:45
UInt_t
unsigned int UInt_t
Definition RtypesCore.h:46
Double_t
double Double_t
Definition RtypesCore.h:59
TRangeDynCast
ROOT::Detail::TRangeCast< T, true > TRangeDynCast
TRangeDynCast is an adapter class that allows the typed iteration through a TCollection.
Definition TCollection.h:358
data
Option_t Option_t TPoint TPoint const char GetTextMagnitude GetFillStyle GetLineColor GetLineWidth GetMarkerStyle GetTextAlign GetTextColor GetTextSize void data
Definition TGWin32VirtualXProxy.cxx:104
offset
Option_t Option_t TPoint TPoint const char GetTextMagnitude GetFillStyle GetLineColor GetLineWidth GetMarkerStyle GetTextAlign GetTextColor GetTextSize void char Point_t Rectangle_t WindowAttributes_t Float_t Float_t Float_t Int_t Int_t UInt_t UInt_t Rectangle_t Int_t Int_t Window_t TString Int_t GCValues_t GetPrimarySelectionOwner GetDisplay GetScreen GetColormap GetNativeEvent const char const char dpyName wid window const char font_name cursor keysym reg const char only_if_exist regb h Point_t winding char text const char depth char const char Int_t count const char ColorStruct_t color const char Pixmap_t Pixmap_t PictureAttributes_t attr const char char ret_data h unsigned char height h offset
Definition TGWin32VirtualXProxy.cxx:245
TCut
A specialized string object used for TTree selections.
Definition TCut.h:25
TFile
A ROOT file is an on-disk file, usually with extension .root, that stores objects in a file-system-li...
Definition TFile.h:53
TMVA::Factory
This is the main MVA steering class.
Definition Factory.h:80
TRandom
This is the base class for the ROOT Random number generators.
Definition TRandom.h:27
TString
Basic string class.
Definition TString.h:139
TTree
A TTree represents a columnar dataset.
Definition TTree.h:79
y
Double_t y[n]
Definition legend1.C:17
x
Double_t x[n]
Definition legend1.C:17
n
const Int_t n
Definition legend1.C:16
Author
Kim Albertsson

Definition in file TMVAMinimalClassification.C.