On Wed, Dec 21, 2016 at 5:37 PM, Rafayel Paremuzyan <[log in to unmask]> wrote:

Hi All,

I was just doing simple tests on new tweak pass,
and noticed something strange in DSTs.

Attached figure shows momentum distribution of electrons, from different data sources.
Black is e- momentum from pass6 DST, from the collection fsparticles.
Blue is e- momentum from new(devel) DST, from the collection fsparticles.
Green is e- momentum from pass6 lcio file, from GBLTracks
Red is e- momentum from new (devel) lcio file, from GBLTracks

Note Black and Blue represent same file number (137) from run 5772, and hence have same number of events.
The normalization between lcio and DST doesn't necessarily should be the same, since thery are from
different collections and different files. I am just guessing fsparticles might have not only GBL tracks, but
seed tracks as well?

I should have answered this before. Yes, fs particles contains both seed and GBL tracks. You need to require the type to be > 32.

But the most imprtant is momentum distribution of new DSTs look very different from others, and goes down
even below 50 Mev, while from old data, momentum starts from 100-150 MeV.

I'm downloading one of the processed files now to check. However, I took a pass 6 v0 skim file

( hps_005772.0_v0_R3.8.slcio) , processed it using the TweakDriver and then ran the DST maker on the tweaked slcio and the momentum scale looks fine (see attached). I'm not sure what the problem is, but I'll start by checking that the slcio files generated at JLab are being processed correctly.

I have looked into positron momentum distributions as well, that one also is very different from others.
different files from same 5772 also showed this strange shape.

I haven't actually worked wth DSTs enough, could DST experts comment on this,
does it really indicates that something is wrong with DSTs? should it be fixed before we start the pass?

Rafo

########################################################################
Use REPLY-ALL to reply to list

To unsubscribe from the HPS-SOFTWARE list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1