Stefan Fröse
Stefan Fröse
@adonath I'm still a bit confused. Isn't this line here exactly allowing interpolation between energies, which should not be allowed by default? https://github.com/gammapy/gammapy/blob/90459a5bb99d7b22f0230627c05268d94212ffbc/gammapy/irf/core.py#L327 I also did a little bit of...
@adonath thanks a lot for the explanations! :)
@AtreyeeS I think best is to allow for nearest interpolation if wanted by the user :)
@AtreyeeS @cosimoNigro @LauraOlivera @QRemy @registerrier you all took part in the discussion :) react to this message if you want to be added as co-author :)
@registerrier @QRemy hi 😊 thanks for all review and helpful comments! I agree with you that your idea fits better into the gammapy framework without increasing the computational effort so...
@QRemy can you take a look? :)
But for ts=ts_asimov the significance should be zero always
ts is the ts value calculated on the measured data for a H1 against H0 (bkg only) while ts_asimov is the same but calculated on the Asimov dataset
ts = cash statistics with signal=0 and bkg fit - cash statistic with signal and bkg fit ts_asimov = cash statistics with signal=0 and bkg=bkg fit (fitted on asimov dataset)...
How is it meant to be used? I mean for which case do you want to convert ts to sigma? You have a dataset with a model and want to...