[Tpe] Fwd: Re: gpart vs npart

[Larry Weinstein]

Dear Folks,

Here is the explanation of gpart vs npart from Stepan.  I suggest we 
follow his recommendation and use gpart >= 2 with at least two tracks 
with dc_stat > 0.

- Larry

-------- Original Message --------
Subject: 	Re: gpart vs npart
Date: 	Tue, 24 May 2011 11:58:00 -0400
From: 	Stepan Stepanyan <stepanya at jlab.org>
To: 	Larry Weinstein <weinstei at jlab.org>



Dear Larry,

Since you are looking for two charged particles, it always
should be npart>=2. In fact for elastic npart=2. If find
more events with elastic kinematics, that means one of
tracks did not pass timing information. It can be either
time-based tracking (less likely) or TOF information was
way OFF - PID (if SC counters had too high occupancy this
is more likely). Since PID is not an issue for elastic events,
you may want to use gpart>=2 events, but make sure
you select those that have 2 time based tracks, dc_stat>0
for both tracks.

Regards, Stepan

On 5/24/11 11:51 AM, Larry Weinstein wrote:
>  Dear Stepan,
>
>  Thanks!  Is there any reason to analyze gpart tracks or should we
>  limit all our analysis to npart tracks?  I ask because Puneet found
>  about 10% more events passing our elastic cuts when he required gpart
>  >= 2 rather than npart>= 2.
>
>  - Larry
>
>  PS: I am also appalled at the dearth of documentation on our offline
>  software.
>
>  Stepan Stepanyan wrote:
>>  Dear Larry,
>>
>>  There is no documentation on that. I sue to have a short writeup
>>  link from old offline page but it disappeared for some reason.
>>
>>  As for gpart/npart - gpart is number of "geometrically matched
>>  particles", no timing information. Npart is number of particles
>>  that were found after timing analysis (time based tracks and in
>>  time neutral hits). EVNT bank has "gpart" entries, first "npart" of
>>  which are particles that passed timing analysis.
>>
>>  Regards, Stepan
>>
>>  On 5/24/11 10:49 AM, Larry Weinstein wrote:
>>>  Dear Stepan,
>>>
>>>  Is there any documentation on the difference between gpart and
>>>  npart?  Are there any cuts on the data other than the number of HBT
>>>  and TBT tracks per event?
>>>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailman.jlab.org/pipermail/tpe/attachments/20110524/da2613df/attachment.html