// input: - Input file (result from TMVA),
// - use of colors or grey scale
// - use of TMVA plotting TStyle
void correlations( TString fin = "TMVA.root",TString outputdir="plots", Bool_t isRegression = kFALSE,
Bool_t greyScale = kFALSE, Bool_t useTMVAStyle = kTRUE )
{
// set style and remove existing canvas'
TMVAGlob::Initialize( useTMVAStyle );
// checks if file with name "fin" is already open, and if not opens one
TFile* file = TMVAGlob::OpenFile( fin );
// signal and background or regression problem
Int_t ncls = (isRegression ? 1 : 2 );
TString hName[2] = { "CorrelationMatrixS", "CorrelationMatrixB" };
if (isRegression) hName[0]= "CorrelationMatrix";
const Int_t width = 600;
for (Int_t ic=0; ic<ncls; ic++) {
TH2* h2 = file->Get( hName[ic] );
if(!h2) {
cout << "Did not find histogram " << hName[ic] << " in " << fin << endl;
continue;
}
TCanvas* c = new TCanvas( hName[ic],
Form("Correlations between MVA input variables (%s)",
(isRegression ? "" : (ic==0 ? "signal" : "background"))),
ic*(width+5)+200, 0, width, width );
Float_t newMargin1 = 0.13;
Float_t newMargin2 = 0.15;
if (TMVAGlob::UsePaperStyle) newMargin2 = 0.13;
c->SetGrid();
c->SetTicks();
c->SetLeftMargin ( newMargin2 );
c->SetBottomMargin( newMargin2 );
c->SetRightMargin ( newMargin1 );
c->SetTopMargin ( newMargin1 );
gStyle->SetPalette( 1, 0 );
gStyle->SetPaintTextFormat( "3g" );
h2->SetMarkerSize( 1.5 );
h2->SetMarkerColor( 0 );
Float_t labelSize = 0.040;
h2->GetXaxis()->SetLabelSize( labelSize );
h2->GetYaxis()->SetLabelSize( labelSize );
h2->LabelsOption( "d" );
h2->SetLabelOffset( 0.011 );// label offset on x axis
h2->Draw("colz"); // color pads
c->Update();
// modify properties of paletteAxis
TPaletteAxis* paletteAxis = (TPaletteAxis*)h2->GetListOfFunctions()->FindObject( "palette" );
paletteAxis->SetLabelSize( 0.03 );
paletteAxis->SetX1NDC( paletteAxis->GetX1NDC() + 0.02 );
h2->Draw("textsame"); // add text
// add comment
TText* t = new TText( 0.53, 0.88, "Linear correlation coefficients in %" );
t->SetNDC();
t->SetTextSize( 0.026 );
t->AppendPad();
// TMVAGlob::plot_logo( );
c->Update();
TString fname = outputdir + TString("/");
fname += hName[ic];
TMVAGlob::imgconv( c, fname );
}
}
// input: - Input file (result from TMVA)
// - use of TMVA plotting TStyle
void mvas( TString fin = "TMVA.root", HistType htype = MVAType, Bool_t useTMVAStyle = kTRUE )
{
// set style and remove existing canvas'
TMVAGlob::Initialize( useTMVAStyle );
// switches
const Bool_t Save_Images = kTRUE;
// checks if file with name "fin" is already open, and if not opens one
TFile* file = TMVAGlob::OpenFile( fin );
// define Canvas layout here!
Int_t xPad = 1; // no of plots in x
Int_t yPad = 1; // no of plots in y
Int_t noPad = xPad * yPad ;
const Int_t width = 600; // size of canvas
// this defines how many canvases we need
TCanvas *c = 0;
// counter variables
Int_t countCanvas = 0;
// search for the right histograms in full list of keys
TIter next(file->GetListOfKeys());
TKey *key(0);
while ((key = (TKey*)next())) {
if (!TString(key->GetName()).BeginsWith("Method_")) continue;
if( ! gROOT->GetClass(key->GetClassName())->InheritsFrom("TDirectory") ) continue;
TString methodName;
TMVAGlob::GetMethodName(methodName,key);
TDirectory* mDir = (TDirectory*)key->ReadObj();
TIter keyIt(mDir->GetListOfKeys());
TKey *titkey;
while ((titkey = (TKey*)keyIt())) {
if (!gROOT->GetClass(titkey->GetClassName())->InheritsFrom("TDirectory")) continue;
TDirectory *titDir = (TDirectory *)titkey->ReadObj();
TString methodTitle;
TMVAGlob::GetMethodTitle(methodTitle,titDir);
cout << "--- Found directory for method: " << methodName << "::" << methodTitle << flush;
TString hname = "MVA_" + methodTitle;
if (htype == ProbaType ) hname += "_Proba";
else if (htype == RarityType ) hname += "_Rarity";
TH1* sig = dynamic_cast<TH1*>(titDir->Get( hname + "_S" ));
TH1* bgd = dynamic_cast<TH1*>(titDir->Get( hname + "_B" ));
if (sig==0 || bgd==0) {
if (htype == MVAType)
cout << "mva distribution not available (this is normal for Cut classifier)" << endl;
else if(htype == ProbaType)
cout << "probability distribution not available (this is normal for Cut classifier)" << endl;
else if(htype == RarityType)
cout << "rarity distribution not available (this is normal for Cut classifier)" << endl;
else if(htype == CompareType)
cout << "overtraining check not available (this is normal for Cut classifier)" << endl;
else cout << endl;
}
else {
cout << endl;
// chop off useless stuff
sig->SetTitle( Form("TMVA response for classifier: %s", methodTitle.Data()) );
if (htype == ProbaType)
sig->SetTitle( Form("TMVA probability for classifier: %s", methodTitle.Data()) );
else if (htype == RarityType)
sig->SetTitle( Form("TMVA Rarity for classifier: %s", methodTitle.Data()) );
else if (htype == CompareType)
sig->SetTitle( Form("TMVA overtraining check for classifier: %s", methodTitle.Data()) );
// create new canvas
TString ctitle = ((htype == MVAType) ?
Form("TMVA response %s",methodTitle.Data()) :
(htype == ProbaType) ?
Form("TMVA probability %s",methodTitle.Data()) :
(htype == CompareType) ?
Form("TMVA comparison %s",methodTitle.Data()) :
Form("TMVA Rarity %s",methodTitle.Data()));
TString cname = ((htype == MVAType) ?
Form("output_%s",methodTitle.Data()) :
(htype == ProbaType) ?
Form("probability_%s",methodTitle.Data()) :
(htype == CompareType) ?
Form("comparison_%s",methodTitle.Data()) :
Form("rarity_%s",methodTitle.Data()));
c = new TCanvas( Form("canvas%d", countCanvas+1), ctitle,
countCanvas*50+200, countCanvas*20, width, (Int_t)width*0.78 );
// set the histogram style
TMVAGlob::SetSignalAndBackgroundStyle( sig, bgd );
// normalise both signal and background
TMVAGlob::NormalizeHists( sig, bgd );
// frame limits (choose judicuous x range)
Float_t nrms = 4;
cout << "--- Mean and RMS (S): " << sig->GetMean() << ", " << sig->GetRMS() << endl;
cout << "--- Mean and RMS (B): " << bgd->GetMean() << ", " << bgd->GetRMS() << endl;
Float_t xmin = TMath::Max( TMath::Min(sig->GetMean() - nrms*sig->GetRMS(),
bgd->GetMean() - nrms*bgd->GetRMS() ),
sig->GetXaxis()->GetXmin() );
Float_t xmax = TMath::Min( TMath::Max(sig->GetMean() + nrms*sig->GetRMS(),
bgd->GetMean() + nrms*bgd->GetRMS() ),
sig->GetXaxis()->GetXmax() );
Float_t ymin = 0;
Float_t maxMult = (htype == CompareType) ? 1.3 : 1.2;
Float_t ymax = TMath::Max( sig->GetMaximum(), bgd->GetMaximum() )*maxMult;
// build a frame
Int_t nb = 500;
TString hFrameName(TString("frame") + methodTitle);
TObject *o = gROOT->FindObject(hFrameName);
if(o) delete o;
TH2F* frame = new TH2F( hFrameName, sig->GetTitle(),
nb, xmin, xmax, nb, ymin, ymax );
frame->GetXaxis()->SetTitle( methodTitle + ((htype == MVAType || htype == CompareType) ? " response" : "") );
if (htype == ProbaType ) frame->GetXaxis()->SetTitle( "Signal probability" );
else if (htype == RarityType ) frame->GetXaxis()->SetTitle( "Signal rarity" );
frame->GetYaxis()->SetTitle("Normalized");
TMVAGlob::SetFrameStyle( frame );
// eventually: draw the frame
frame->Draw();
c->GetPad(0)->SetLeftMargin( 0.105 );
frame->GetYaxis()->SetTitleOffset( 1.2 );
// Draw legend
TLegend *legend= new TLegend( c->GetLeftMargin(), 1 - c->GetTopMargin() - 0.12,
c->GetLeftMargin() + (htype == CompareType ? 0.40 : 0.3), 1 - c->GetTopMargin() );
legend->SetFillStyle( 1 );
legend->AddEntry(sig,TString("Signal") + ((htype == CompareType) ? " (test sample)" : ""), "F");
legend->AddEntry(bgd,TString("Background") + ((htype == CompareType) ? " (test sample)" : ""), "F");
legend->SetBorderSize(1);
legend->SetMargin( (htype == CompareType ? 0.2 : 0.3) );
legend->Draw("same");
// overlay signal and background histograms
sig->Draw("samehist");
bgd->Draw("samehist");
if (htype == CompareType) {
// if overtraining check, load additional histograms
TH1* sigOv = 0;
TH1* bgdOv = 0;
TString ovname = hname += "_Train";
sigOv = dynamic_cast<TH1*>(titDir->Get( ovname + "_S" ));
bgdOv = dynamic_cast<TH1*>(titDir->Get( ovname + "_B" ));
if (sigOv == 0 || bgdOv == 0) {
cout << "+++ Problem in \"mvas.C\": overtraining check histograms do not exist" << endl;
}
else {
cout << "--- Found comparison histograms for overtraining check" << endl;
TLegend *legend2= new TLegend( 1 - c->GetRightMargin() - 0.42, 1 - c->GetTopMargin() - 0.12,
1 - c->GetRightMargin(), 1 - c->GetTopMargin() );
legend2->SetFillStyle( 1 );
legend2->SetBorderSize(1);
legend2->AddEntry(sigOv,"Signal (training sample)","P");
legend2->AddEntry(bgdOv,"Background (training sample)","P");
legend2->SetMargin( 0.1 );
legend2->Draw("same");
}
Int_t col = sig->GetLineColor();
sigOv->SetMarkerColor( col );
sigOv->SetMarkerSize( 0.7 );
sigOv->SetMarkerStyle( 20 );
sigOv->SetLineWidth( 1 );
sigOv->SetLineColor( col );
sigOv->Draw("e1same");
col = bgd->GetLineColor();
bgdOv->SetMarkerColor( col );
bgdOv->SetMarkerSize( 0.7 );
bgdOv->SetMarkerStyle( 20 );
bgdOv->SetLineWidth( 1 );
bgdOv->SetLineColor( col );
bgdOv->Draw("e1same");
ymax = TMath::Max( ymax, TMath::Max( sigOv->GetMaximum(), bgdOv->GetMaximum() )*maxMult );
frame->GetYaxis()->SetLimits( 0, ymax );
// for better visibility, plot thinner lines
sig->SetLineWidth( 1 );
bgd->SetLineWidth( 1 );
// perform K-S test
cout << "--- Perform Kolmogorov-Smirnov tests" << endl;
Double_t kolS = sig->KolmogorovTest( sigOv );
Double_t kolB = bgd->KolmogorovTest( bgdOv );
cout << "--- Goodness of signal (background) consistency: " << kolS << " (" << kolB << ")" << endl;
TString probatext = Form( "Kolmogorov-Smirnov test: signal (background) probability = %5.3g (%5.3g)", kolS, kolB );
TText* tt = new TText( 0.12, 0.74, probatext );
tt->SetNDC(); tt->SetTextSize( 0.032 ); tt->AppendPad();
}
// redraw axes
frame->Draw("sameaxis");
// text for overflows
Int_t nbin = sig->GetNbinsX();
Double_t dxu = sig->GetBinWidth(0);
Double_t dxo = sig->GetBinWidth(nbin+1);
TString uoflow = Form( "U/O-flow (S,B): (%.1f, %.1f)%% / (%.1f, %.1f)%%",
sig->GetBinContent(0)*dxu*100, bgd->GetBinContent(0)*dxu*100,
sig->GetBinContent(nbin+1)*dxo*100, bgd->GetBinContent(nbin+1)*dxo*100 );
TText* t = new TText( 0.975, 0.115, uoflow );
t->SetNDC();
t->SetTextSize( 0.030 );
t->SetTextAngle( 90 );
t->AppendPad();
// update canvas
c->Update();
// save canvas to file
TMVAGlob::plot_logo(1.058);
if (Save_Images) {
if (htype == MVAType) TMVAGlob::imgconv( c, Form("plots/mva_%s", methodTitle.Data()) );
else if (htype == ProbaType) TMVAGlob::imgconv( c, Form("plots/proba_%s", methodTitle.Data()) );
else if (htype == CompareType) TMVAGlob::imgconv( c, Form("plots/overtrain_%s", methodTitle.Data()) );
else TMVAGlob::imgconv( c, Form("plots/rarity_%s", methodTitle.Data()) );
}
countCanvas++;
}
}
}
}
void variables( TString fin = "TMVA.root", TString dirName = "InputVariables_Id", TString title = "TMVA Input Variables",
Bool_t isRegression = kFALSE, Bool_t useTMVAStyle = kTRUE )
{
TString outfname = dirName;
outfname.ToLower(); outfname.ReplaceAll( "input", "" );
// set style and remove existing canvas'
TMVAGlob::Initialize( useTMVAStyle );
// obtain shorter histogram title
TString htitle = title;
htitle.ReplaceAll("variables ","variable");
htitle.ReplaceAll("and target(s)","");
htitle.ReplaceAll("(training sample)","");
// checks if file with name "fin" is already open, and if not opens one
TFile* file = TMVAGlob::OpenFile( fin );
TDirectory* dir = (TDirectory*)file->Get( dirName );
if (dir==0) {
cout << "No information about " << title << " available in directory " << dirName << " of file " << fin << endl;
return;
}
dir->cd();
// how many plots are in the directory?
Int_t noPlots = TMVAGlob::GetNumberOfInputVariables( dir ) +
TMVAGlob::GetNumberOfTargets( dir );
// define Canvas layout here!
// default setting
Int_t xPad; // no of plots in x
Int_t yPad; // no of plots in y
Int_t width; // size of canvas
Int_t height;
switch (noPlots) {
case 1:
xPad = 1; yPad = 1; width = 550; height = 0.90*width; break;
case 2:
xPad = 2; yPad = 1; width = 600; height = 0.50*width; break;
case 3:
xPad = 3; yPad = 1; width = 900; height = 0.4*width; break;
case 4:
xPad = 2; yPad = 2; width = 600; height = width; break;
default:
xPad = 3; yPad = 2; width = 800; height = 0.55*width; break;
}
Int_t noPadPerCanv = xPad * yPad ;
// counter variables
Int_t countCanvas = 0;
Int_t countPad = 0;
// loop over all objects in directory
TCanvas* canv = 0;
TKey* key = 0;
Bool_t createNewFig = kFALSE;
TIter next(dir->GetListOfKeys());
while ((key = (TKey*)next())) {
if (key->GetCycle() != 1) continue;
if (!TString(key->GetName()).Contains("__Signal") &&
!(isRegression && TString(key->GetName()).Contains("__Regression"))) continue;
// make sure, that we only look at histograms
TClass *cl = gROOT->GetClass(key->GetClassName());
if (!cl->InheritsFrom("TH1")) continue;
TH1 *sig = (TH1*)key->ReadObj();
TString hname(sig->GetName());
// create new canvas
if (countPad%noPadPerCanv==0) {
++countCanvas;
canv = new TCanvas( Form("canvas%d", countCanvas), title,
countCanvas*50+50, countCanvas*20, width, height );
canv->Divide(xPad,yPad);
canv->Draw();
}
TPad* cPad = (TPad*)canv->cd(countPad++%noPadPerCanv+1);
// find the corredponding backgrouns histo
TString bgname = hname;
bgname.ReplaceAll("__Signal","__Background");
TH1 *bgd = (TH1*)dir->Get(bgname);
if (bgd == NULL) {
cout << "ERROR!!! couldn't find background histo for" << hname << endl;
exit;
}
// this is set but not stored during plot creation in MVA_Factory
TMVAGlob::SetSignalAndBackgroundStyle( sig, (isRegression ? 0 : bgd) );
sig->SetTitle( TString( htitle ) + ": " + sig->GetTitle() );
TMVAGlob::SetFrameStyle( sig, 1.2 );
// normalise both signal and background
if (!isRegression) TMVAGlob::NormalizeHists( sig, bgd );
else {
// change histogram title for target
TString nme = sig->GetName();
if (nme.Contains( "_target" )) {
TString tit = sig->GetTitle();
sig->SetTitle( tit.ReplaceAll("Input variable", "Regression target" ) );
}
}
// finally plot and overlay
Float_t sc = 1.1;
if (countPad == 1) sc = 1.3;
sig->SetMaximum( TMath::Max( sig->GetMaximum(), bgd->GetMaximum() )*sc );
sig->Draw( "hist" );
cPad->SetLeftMargin( 0.17 );
sig->GetYaxis()->SetTitleOffset( 1.70 );
if (!isRegression) {
bgd->Draw("histsame");
TString ytit = TString("(1/N) ") + sig->GetYaxis()->GetTitle();
sig->GetYaxis()->SetTitle( ytit ); // histograms are normalised
}
TLatex *text = new TLatex();
text->SetNDC();
text->SetTextSize(0.06);
float overlap = computeOverlap(sig,bgd);
text->DrawLatex(0.8,0.8,Form("%.2f",overlap));
// Draw legend
if (countPad == 1 && !isRegression) {
TLegend *legend= new TLegend( cPad->GetLeftMargin(),
1-cPad->GetTopMargin()-.15,
cPad->GetLeftMargin()+.4,
1-cPad->GetTopMargin() );
legend->SetFillStyle(1);
legend->AddEntry(sig,"Signal","F");
legend->AddEntry(bgd,"Background","F");
legend->SetBorderSize(1);
legend->SetMargin( 0.3 );
legend->Draw("same");
}
// redraw axes
sig->Draw("sameaxis");
// text for overflows
Int_t nbin = sig->GetNbinsX();
Double_t dxu = sig->GetBinWidth(0);
Double_t dxo = sig->GetBinWidth(nbin+1);
TString uoflow = "";
if (isRegression) {
uoflow = Form( "U/O-flow: %.1f%% / %.1f%%",
sig->GetBinContent(0)*dxu*100, sig->GetBinContent(nbin+1)*dxo*100 );
}
else {
uoflow = Form( "U/O-flow (S,B): (%.1f, %.1f)%% / (%.1f, %.1f)%%",
sig->GetBinContent(0)*dxu*100, bgd->GetBinContent(0)*dxu*100,
sig->GetBinContent(nbin+1)*dxo*100, bgd->GetBinContent(nbin+1)*dxo*100 );
}
TText* t = new TText( 0.98, 0.14, uoflow );
t->SetNDC();
t->SetTextSize( 0.040 );
t->SetTextAngle( 90 );
t->AppendPad();
// save canvas to file
if (countPad%noPadPerCanv==0) {
TString fname = Form( "plots/%s_c%i", outfname.Data(), countCanvas );
TMVAGlob::plot_logo();
TMVAGlob::imgconv( canv, fname );
createNewFig = kFALSE;
}
else {
createNewFig = kTRUE;
}
}
if (createNewFig) {
TString fname = Form( "plots/%s_c%i", outfname.Data(), countCanvas );
TMVAGlob::plot_logo();
TMVAGlob::imgconv( canv, fname );
createNewFig = kFALSE;
}
return;
}
