Problem with TMVA

Hello, I have an issue with TMVA. While I am trying to do the training and the testing (BDT) I get the following result:
— Tools : error in matrix diagonalization; printed S and B

and (for one of many variables):
– TFHandler_Factory : Variable Mean RMS [ Min Max ]
— TFHandler_Factory : -----------------------------------------------------------------------------------------
— TFHandler_Factory : ratio: -nan -nan [ 1.7977e+308 -1.7977e+308 ]

and the training stops with the following message:

— Tools : signal and background histograms have different or invalid dimensions:

I looked online and they are proposing that in:

TMVA::Factory *factory = new TMVA::Factory( “TMVAClassification”, outputFile,
"!V:!Silent:Color:DrawProgressBar:Transformations=I;D;P;G,D:AnalysisType=Classification" );

I leave only Transformations=I;

But won’t this affect my analysis?

Since I m a rookie, does anyone know what to do?

Cheers
George

Hi,

Can you post please your macro and your data so I can look at your problem

Cheers

Lorenzo

TMVAClassification.C (29.1 KB)

This is the macro although I think that the problem is not there. The issue is the calculation of the decorrelation matrix and I found online the proposal to erase its calculation when I call the TMVA factory but I am worried that this will affect the procedure…?

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.