cms-analysis / HiggsAnalysis-CombinedLimit

CMS Higgs Combination toolkit.
https://cms-analysis.github.io/HiggsAnalysis-CombinedLimit/latest
Apache License 2.0
75 stars 385 forks source link

-S option not accepted anymore #853

Closed campaneros closed 5 months ago

campaneros commented 1 year ago

When I tried running a datacard with the option -S 0 since I do not have the systematic in the datacard, it gives me the error -S invalid option. If iI try to run without the option -S it gives me the error asking me to put -S because I do not have the systematic

nucleosynthesis commented 1 year ago

Can you tell us which version of combine you are using (tag or branch)?

The option -S 0 indeed has been deprecated since some time, however

I run on the main branch with the datacard below and use the command,

combined datacard.txt

I get the expected output,

 <<< Combine >>>
>>> random number generator seed is 123456
>>> method used is AsymptoticLimits

 -- AsymptoticLimits ( CLs ) --
Observed Limit: r < 0.5068
Expected  2.5%: r < 0.3549
Expected 16.0%: r < 0.5137
Expected 50.0%: r < 0.7969
Expected 84.0%: r < 1.2860
Expected 97.5%: r < 1.9879

Done in 0.00 min (cpu), 0.00 min (real)

datacard.txt

imax 1  number of channels
jmax 1  number of backgrounds
kmax *  number of nuisance parameters (sources of systematical uncertainties)
------------
bin         bin1
observation 0
------------
bin           bin1   bin1
process       ggh4G  Bckg
process         0      1
rate           4.76  1.47
------------
nucleosynthesis commented 5 months ago

It seems there was indeed an erroneous suggestion to re-run with -S 0. Instead one should use the --toysNoSys option. This suggestion has now been removed from the output.