[Halld-physics] Check your MC samples!

[Alexander Austregesilo]

Dear Colleagues,

Last summer, we discovered a bug in the mcsmear hit merging routine 
which eliminated any hits in the TOF for a seemingly random fraction of 
jobs. The bug was fixed, but it was recently discovered that the impact 
of this effect is considerably larger than expected (see 
https://github.com/JeffersonLab/halld_recon/issues/780). Some of the 
affected MC samples wrongly show up to a 10% higher efficiency.

If you are currently using MC samples generated before Fall 2023, please 
check the used version set. The first version set which contain the 
fixed mcsmear code for each run period is listed below:

recon-2017_01-ver03_36.xml
recon-2017_01-ver04_3.xml
recon-2018_01-ver02_28.xml
recon-2018_08-ver02_27.xml
recon-2019_11-ver01_9.xml

All version sets which have a smaller last number than these may be 
affected and we strongly advise you to regenerate them to avoid any 
systematic errors. As always, we recommend using the latest version as 
more improvements were implemented in the mean time.

If you have questions or are unsure about the choice, please confirm 
with Justin, Peter, Sean or me.

Best regards,

Alex

-- 
Alexander Austregesilo

Staff Scientist - Experimental Nuclear Physics
Thomas Jefferson National Accelerator Facility
Newport News, VA
aaustreg at jlab.org
(757) 269-6982