search

ytatus94 / Higgsino

The NUHM2 Higgsino study
0 stars 0 forks source link

Judita's mail 20170713 #31

Open ytatus94 opened 7 years ago

ytatus94 commented 7 years ago

Hi Chris,

I remembered something. We will never get 100k per production approved in the request. So no need to test that. At most it will be 20k per production, so I think we can safely revert to multiplier of 10 when MET filter is used.

@ Yu-Ting, we were discussing wether to increase the multiplier with the MET filter, like you suggested to 50, and Chris will make a few tests.

Cheers, Judita

ytatus94 commented 7 years ago

P.S. I have changed the evt_multiplier to 10 in the JO in the CHRIS folder, just so you know.

ytatus94 commented 7 years ago

Yes, the original setup was 10 for the 2L filter, and 10 for the MET filter, but with this the jobs were failing. They were also failing for tested extra x3 and x5 in the 2L filter, so then they stopped failing with x10 in the 2L filter. In the higgsino JO they use 10X3 for 2L and 10 for MET filter, for this mass splitting. It looks that we need a higher multiplier.

ytatus94 commented 7 years ago

I have run the 4 productions with METx10 and 10k events each, but the check with METx10 and 20k would be useful. If that is not successful, then perhaps we would also test METx50 with 20k events.