sPHENIX discussion of software

["Frawley, Anthony" <afrawley AT fsu.edu>]

Hi Mike,

The value has gone down since the early running with the TPC in (I had a macro that previously required all layers that got a lot of tracks, now it would get almost none). But that was pre-ALD-charge.

Krista and I were just looking at the results of running 100 pion events with what is now the default, and the momentum resolution looks quite good with no cuts at all on the number of hits per track. And the DCA resolution looks good. So we do not see any obvious performance issues associated with this. Perhaps that means we just don't need all 60 hits.

Tony

---

From: michael.p.mccumber AT gmail.com <michael.p.mccumber AT gmail.com> on behalf of Michael P. McCumber <mccumber AT bnl.gov>
Sent: Thursday, June 23, 2016 1:56 PM
To: Frawley, Anthony
Cc: Sourav Tarafdar; sphenix-software-l AT lists.bnl.gov
Subject: Re: [Sphenix-software-l] Issue with new.1 and new.2 libraries?

 

Hi Tony,

If I remember correctly, this has been with us since the "good efficiency" tune. The "good momentum resolution" tune had more hits. But I don't have this quantity under eval test, so I'm not 100% sure. Do you have evidence that the value has changed since the TPC was put in master? 

Mike

--

Michael P. McCumber, PhD
Los Alamos National Laboratory
505-709-8161

On Wed, Jun 22, 2016 at 10:57 PM, Frawley, Anthony <afrawley AT fsu.edu> wrote:

And I should also have mentioned that the purity is quite good - purity = nhits in most cases.

Tony

---

From: sphenix-software-l-bounces AT lists.bnl.gov <sphenix-software-l-bounces AT lists.bnl.gov> on behalf of Frawley, Anthony <afrawley AT fsu.edu>
Sent: Wednesday, June 22, 2016 10:55 PM
To: Michael P. McCumber; Sourav Tarafdar
Cc: sphenix-software-l AT lists.bnl.gov

Subject: Re: [Sphenix-software-l] Issue with new.1 and new.2 libraries?

 

Hi Mike,

After I updated the maps+tpc macro I see tracks with today's build (new.4). However I see an odd thing that I think is new - the number of hit layers for 3 layers of maps + 60 TPC layers is on average about 54. I attach a plot of the histogram of "nhits" from the ntp_track ntuple, written by the tracking evaluator. The tracks are from 10,000 80 pion events (pT = 0.5-40 GeV/c).

Previously, most tracks had a hit in every layer. I note that in the example you showed in your email, nhits is 51. The tracks seem to be of good quality and have good momentum resolution. 

I am using the maps+tpc macro straight out of the box.

Tony

---

From: michael.p.mccumber AT gmail.com <michael.p.mccumber AT gmail.com> on behalf of Michael P. McCumber <mccumber AT bnl.gov>
Sent: Wednesday, June 22, 2016 10:02 PM
To: Sourav Tarafdar
Cc: Frawley, Anthony; sphenix-software-l AT lists.bnl.gov
Subject: Re: [Sphenix-software-l] Issue with new.1 and new.2 libraries?

 

Hi guys,

There have been a number of changes in the last 4 days that require synchronous updates to both the coresoftware and the macros repo. However, I just took the very latest macros, backed out a change in the MAP+TPC macro to cluster the old way with just the TPC code (which would have otherwise caused a crash stop), and I see reconstructed tracks from new.1 (Jun 20th build).

root [1] ntp_track->Show(1)

======> EVENT:1

 event           = 0

 trackID         = 1

 px              = 8.65076

 py              = -2.27025

 pz              = 0.879521

 charge          = -1

 quality         = 0.443519

 chisq           = 43.0213

 ndf             = 97

 nhits           = 51

 layers          = 1.87904e+09

 dca2d           = -0.000504471

 dca2dsigma      = 0.00121597

 pcax            = -3.88547e-05

 pcay            = -0.000303789

 pcaz            = 2.13742

My macro checkout, clustering backport delta, and eval output are in /phenix/u/mccumber/macros. I'll leave that up for comparison in case you want to track down the issue while the new.1 is still extant.

Mike

--

Michael P. McCumber, PhD
Los Alamos National Laboratory
505-709-8161

On Wed, Jun 22, 2016 at 6:34 PM, Sourav Tarafdar <sourav.pheonix AT gmail.com> wrote:

Hi Tony,

Actually even with updates in macros I also noticed the issue with new.1 and new.2 libraries. I think this has been since last couple of days. So currently I am using either new.3 and new.4 though not sure about the reason for not getting reconstructed tracks with new.1 and new.2 libraries.

Best regards

-Sourav

On Wed, Jun 22, 2016 at 6:28 PM, Frawley, Anthony <afrawley AT fsu.edu> wrote:

Sorry, disregard this. I was coming back to this after a few weeks, and I missed an update from June 9 to the macros.

Tony

---

From: Frawley, Anthony
Sent: Tuesday, June 21, 2016 11:00 PM
To: sphenix-software-l AT lists.bnl.gov
Cc: Frawley, Anthony
Subject: Issue with new.1 and new.2 libraries?

 

Hi All,

When I run the sPHENIX tracking simulation (using the standard maps+TPC macro) I get no reconstructed tracks. This occurs with the new.2 and new.1 libraries, but when I switch to new.4 I get reconstructed tracks again - all running the same macros.

Does anyone have any ideas?

Thanks

Tony

_______________________________________________
Sphenix-software-l mailing list
Sphenix-software-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-software-l

_______________________________________________
Sphenix-software-l mailing list
Sphenix-software-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-software-l