reweighting samples with madspin about genproductions HOT 9 CLOSED

The issue is that madspin calculates probabilities based on the input ME. As this is changing through EFT reweighting it doesn't work properly. So I agree on the veto as long as we have no central fix by Oliver.,

from genproductions.

cms-sw/cmssw#43784 (comment)

Not too much related to the bug reported in the link above but this reminded me of the issue

from genproductions.

@efeyazgan for the gen scripts

@bbilin @menglu21 @agrohsje @Cvico @DickyChant

from genproductions.

Yes, I think that makes to veto such submissions. I will implement this sometime today. I'll also take a look at the other issue you pointed to.

from genproductions.

@efeyazgan hello efe, thanks for quick response! i haven't checked the issue myself though. i think someone from eft analysis should give some insights on this buggy behavior if it is really true

from genproductions.

From CMStalk, the conclusion seems to be what you suggested, but OK, I will wait for feedback from others before implementing this.

from genproductions.

To implement this, to understand if reweighting is on, would it be sufficient to check the existence of the reweight_card or do you know of a better way to do it?

from genproductions.

technically that would be enough. existence of both madspin_card.dat and reweight_card.dat.

if @agrohsje can confirm the bug/issue we can implement this i think

from genproductions.

#3658 addressed, closing it here

from genproductions.