Another difference between 6.18 and 6.22

Roofit and RooStats
1

Dear all,
I have recently reported a difference between 6.18 and 6.22 when it comes to RooFit/RooStats. Last time it was solved by adding this command:

RooAbsReal::defaultIntegratorConfig()->method1D().setLabel("RooAdaptiveGaussKronrodIntegrator1D");

But in the case I am reporting here it won’t help.
I attach the macro and the root file needed by macro.
Please run the macro like this:

root -l exercise_2.C

exercise_2.C (5.0 KB) Workspace_mumufit.root (14.9 KB)

The numerical results are identical between 6.18 and 6.22 but the plot for the posterior integration looks wrong. I join both of them, filename should be self-explaining.
exercise_2_618.pdf (17.0 KB) exercise_2_622.pdf (17.1 KB)

Any help will be greatly appreciated.

Best regards,
Marco Bomben

2

@moneta can perhaps explain? Thanks!

3

Hi,
Looking at th plots, it looks the difference is only in the gray area. Not in the curve itself.
If this is the case is then probably a graphics problem. Let me double check this

Lorenzo

4

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