Root v4 vs v5

Roofit and RooStats
1

Hi,

Calculating chi^2 probabilities in Root v4.04/02b vs Root v5.22/00a yields
different results when comparing identical histograms.

Any obvious reason why?

Thanks,
Marc

2

Hi,

yes, the implementation of the chi2 test was improved a lot, see e.g. root.cern.ch/root/html/TH1#TH1:Chi2Test which also gives external references.

Cheers, Axel.