Minuit fit problem with large statistics

Roofit and RooStats
21

Hi @moneta,

I see your code performs several fits and many of them fail because of precision issues. It is difficult to me to find the problem mentioned before with Minos.

yes, right. The code I provided includes also the likelihood profiling, which performs several fits;

Can you please simplify the code, so I can exactly reproduce the Minos problem ?

yes, I am attaching a new version in which only the fit and Minos analysis are performed;

Thank you,
Antonio

test_code_v2.tar.gz (6.8 KB)

22

Hello Antonio,

please note the @moneta will be on vacation the next two weeks. If you check in once a week by writing a short message, the topic won’t get closed.

If you have a CERN account, you can also create a JIRA ticket on https://root.cern (button with the bug at the very bottom), and add a link to this thread.

23

Hi @StephanH,

thanks for the info.

I opened a ticket on JIRA, as you have suggested:

https://sft.its.cern.ch/jira/browse/ROOT-10954

Cheers,
Antonio