[Hps-analysis] [EXTERNAL] Re: Track Chisq Cuts

Tue Jun 7 20:45:25 EDT 2022

I see, this is nice. Looks like we do currently use mostly defaults.
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JeffersonLab_hps-2Djava_blob_master_steering-2Dfiles_src_main_resources_org_hps_steering_recon_PhysicsRun2021-5Fpass0-5Frecon-5Fevio.lcsim-23L269&d=DwIF-g&c=CJqEzB1piLOyyvZjb8YUQw&r=JaSEOiNc_6InrJmbYDvKU2tZqhhONpIkbyl_AUnkDSY&m=arZMnawUI9fy16o-1QnkHcWWuh8PV_rZ6eHl2eeWdA4Of9N6db9hsLNs712tb3FW&s=HzyAyfJRg1U2pxaBkWPYhsBvE2RX7XoaWaic-lNdx-w&e= 

________________________________
From: Robert Johnson <rjohnson at ucsc.edu>
Sent: Tuesday, June 7, 2022 5:42 PM
To: Bravo, Cameron B. <bravo at slac.stanford.edu>
Cc: Graf, Norman A. <ngraf at slac.stanford.edu>; Graham, Mathew Thomas <mgraham at slac.stanford.edu>; hps-analysis at jlab.org <hps-analysis at jlab.org>
Subject: Re: [Hps-analysis] [EXTERNAL] Re: Track Chisq Cuts

Only the default values are hardcoded. They all can be, and may be, modified by the steering file.

Robert P. Johnson
Professor
Physics Department
Santa Cruz Institute for Particle Physics
University of California at Santa Cruz
Santa Cruz, CA 95064
831-459-2125
rjohnson at ucsc.edu<mailto:rjohnson at ucsc.edu>

On Tue, Jun 7, 2022 at 5:16 PM Bravo, Cameron B. <bravo at slac.stanford.edu<mailto:bravo at slac.stanford.edu>> wrote:
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JeffersonLab_hps-2Djava_blob_master_tracking_src_main_java_org_hps_recon_tracking_kalman_KalmanParams.java&d=DwIF-g&c=CJqEzB1piLOyyvZjb8YUQw&r=JaSEOiNc_6InrJmbYDvKU2tZqhhONpIkbyl_AUnkDSY&m=arZMnawUI9fy16o-1QnkHcWWuh8PV_rZ6eHl2eeWdA4Of9N6db9hsLNs712tb3FW&s=oAcKArbLuplqA0VBSPd1f0kbtO0Hfu4bho-BozgCLZ4&e= <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_JeffersonLab_hps-2Djava_blob_master_tracking_src_main_java_org_hps_recon_tracking_kalman_KalmanParams.java&d=DwMF-g&c=CJqEzB1piLOyyvZjb8YUQw&r=JaSEOiNc_6InrJmbYDvKU2tZqhhONpIkbyl_AUnkDSY&m=_dIpefaElthIovRGpd51k2QSLzXqneNsG1eYNUkorr_FLz5DlskHP2uNzXUvQw3q&s=lW22qIYz7qremO_4SZDK0LwuvDyYZ55EzFOpyQI2pz0&e=>

Probably comes from one of these hardcoded numbers.
________________________________
From: Hps-analysis <hps-analysis-bounces at jlab.org<mailto:hps-analysis-bounces at jlab.org>> on behalf of Bravo, Cameron B. <bravo at slac.stanford.edu<mailto:bravo at slac.stanford.edu>>
Sent: Tuesday, June 7, 2022 5:10 PM
To: Graf, Norman A. <ngraf at slac.stanford.edu<mailto:ngraf at slac.stanford.edu>>; Graham, Mathew Thomas <mgraham at slac.stanford.edu<mailto:mgraham at slac.stanford.edu>>; hps-analysis at jlab.org<mailto:hps-analysis at jlab.org> <hps-analysis at jlab.org<mailto:hps-analysis at jlab.org>>
Subject: [Hps-analysis] [EXTERNAL] Re: Track Chisq Cuts

It is what I said. If I require exactly 14 hits I do see a ~cut at 25, but also do see a few events sprinkled above 25 like you do.
________________________________
From: Graf, Norman A. <ngraf at slac.stanford.edu<mailto:ngraf at slac.stanford.edu>>
Sent: Tuesday, June 7, 2022 5:03 PM
To: Bravo, Cameron B. <bravo at slac.stanford.edu<mailto:bravo at slac.stanford.edu>>; Graham, Mathew Thomas <mgraham at slac.stanford.edu<mailto:mgraham at slac.stanford.edu>>; hps-analysis at jlab.org<mailto:hps-analysis at jlab.org> <hps-analysis at jlab.org<mailto:hps-analysis at jlab.org>>
Subject: RE: Track Chisq Cuts

Hi Cameron,

Thanks for that feedback. Are these top tracks with 14 hits?

Norman

From: Bravo, Cameron B. <bravo at slac.stanford.edu<mailto:bravo at slac.stanford.edu>>
Sent: Tuesday, June 7, 2022 4:39 PM
To: Graf, Norman A. <ngraf at slac.stanford.edu<mailto:ngraf at slac.stanford.edu>>; Graham, Mathew Thomas <mgraham at slac.stanford.edu<mailto:mgraham at slac.stanford.edu>>; hps-analysis at jlab.org<mailto:hps-analysis at jlab.org>
Subject: Re: Track Chisq Cuts

Hello Norman,

I see plenty of KF tracks with chi2/ndf over 25 in run 14166. This is all KF tracks in the top in partition 37 of that run using a detector I made last week.

I did require chi2/ndf be less than 1000 for this plot.

[cid:18140c2d5314cff311]

Cheers,

Cameron

________________________________

From: Hps-analysis <hps-analysis-bounces at jlab.org<mailto:hps-analysis-bounces at jlab.org>> on behalf of Graf, Norman A. <ngraf at slac.stanford.edu<mailto:ngraf at slac.stanford.edu>>
Sent: Tuesday, June 7, 2022 3:59 PM
To: Graham, Mathew Thomas <mgraham at slac.stanford.edu<mailto:mgraham at slac.stanford.edu>>; hps-analysis at jlab.org<mailto:hps-analysis at jlab.org> <hps-analysis at jlab.org<mailto:hps-analysis at jlab.org>>
Subject: [Hps-analysis] [EXTERNAL] Track Chisq Cuts

Hello All,

Here is a followup to my presentation at today’s meeting with somewhat higher statistics. Recall that track1 is associated with a calorimeter cluster, whereas track2 has no such requirement. To get around the bimodal momentum distribution in top tracks I require 14 hits on tracks on the top and 13 hits on tracks on the bottom (due to the dead sensor). I plot here the chisq/ndf for Møller candidates with such tight restrictions on the number of hits. Not only is the track fit much worse for top tracks, there is a hard cutoff at 25, which is clearly cutting into the number of tracks on the top. I have not explicitly applied this cut in my analysis.

Looking at a more general sample of tracks, I notice a similar cutoff at higher values of chisq, although it varies by number of hits on the track. Before I dive into the code does this look familiar to anyone? Is there an explicit maximum chisq cut being applied in the current track reconstruction?

Thanks,

Norman

_______________________________________________
Hps-analysis mailing list
Hps-analysis at jlab.org<mailto:Hps-analysis at jlab.org>
https://mailman.jlab.org/mailman/listinfo/hps-analysis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/hps-analysis/attachments/20220608/8783bb3d/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.png
Type: image/png
Size: 60688 bytes
Desc: image001.png
URL: <https://mailman.jlab.org/pipermail/hps-analysis/attachments/20220608/8783bb3d/attachment-0001.png>