[Clas12_rgb] plan for completion of checks of new CVT tracking, for pass2

silvia at jlab.org silvia at jlab.org
Mon Nov 7 03:34:29 EST 2022


Dear all,
the report of the RGB review has not officially come out yet, but I can
anticipate that will contain an explicit request to complete our studies
to understand which version of CVT reconstruction we will use for pass2.
As many of you heard, both in the review itself and during the CLAS
meeting, the new CVT reconstruction relies on some cuts, in particular on
the pt and on the z vertex, which were hardcoded in the version we used
for our test cooking, but will be included as parameters of the yaml file
in an upcoming version.
We should make sure to pick values of the cuts which are OK for all the
analyses that will use pass2 RGB data. For the 28 October review, some
analyzers carried out comparative tests using the pass1, pass2_old_cvt,
pass2_new_cvt reconstructions. It is apparent that for the DVCS analyses
the new tracking is very good, already in the 8c.3.2 version.
We need that all analyzers who are relying on the CD do a similar test,
and quickly.
I propose the following plan:
1) Like Adam did, you should run your analysis on the selected 15 runs
using the pass1, pass2_old_cvt, pass2_new_cvt reconstructions and compare
yields and, possibly, kinematic distributions and resolutions (on missing
masses, for instance, if it applies to your analysis).
2) If there is an improvement with the new CVT tracking, this is good. If
there is a worsening, we should understand to which cut of the CVT
tracking (pt or z vertex) this is due to, and the optimal values of the
cuts on these variables should be established, either looking at the data
or at the MC simulations.

The data locations for the 15 runs are:
Old CVT tracking:
/volatile/clas12/rg-b/production/recon/pass0/v28.22_8.3.2/dst

New CVT tracking:
/volatile/clas12/rg-b/production/recon/pass0/v28.22_8c.3.2/dst

Please get in touch with me if you need more information or to discuss in
detail.
Thanks a lot and best regards,
Silvia




More information about the Clas12_rgb mailing list