STAR QA Board

["Hu, Yu" <yuhu@bnl.gov>]

Dear Ashik and Li-Ke,

Some short updates: 

1. I'm not sure if you guys already solved the issue of saving the repeated bad runs from different qualities.  I prepared a new version which saved a new file named <badruns_allcache.txt>, I also put some bash commands to help you guys find those runs in <checkrepeated.sh> as a reference.

I pushed a newer version on Github including a newer README.   https://github.com/EllipticFlow/run-by-run-qa-algorithm.git

I also uploaded a version to the my rcf by hand: /star/u/huyu/Runbyrun_qa/version-1.1

2. I have a quick look at ashik_hist_111.root, may I know why the  "runidvssdcaxyrms" has 1279 entries and all the others have 887 runs? Maybe I'm looking at the wrong root file.

Thank you.

Best,

Yu

---

From: ASHIK IKBAL <ashikhep@gmail.com>
Sent: Monday, November 29, 2021 14:43
To: Hu, Yu <yuhu@bnl.gov>
Cc: star-qaboard-l@lists.bnl.gov <star-qaboard-l@lists.bnl.gov>; STAR centrality focus group <star-centrality-l@lists.bnl.gov>; cebra <cebra@physics.ucdavis.edu>; Grigory Nigmatkulov <nigmatkulov@gmail.com>; Prithwish Tribedy <ptribedy@icloud.com>; Ma, Rongrong <marr@rcf.rhic.bnl.gov>
Subject: Re: Run-by-run QA for BES-II 19.6 GeV datasets

 

Dear Yu,

Thanks for the explanation on the algorithm. Your explanation on getting a few different badruns from the algorithm makes sense to me.  Of course, the root files are not exactly same (in terms of no. of events, though there are same runs):

Ashik: /star/data01/pwg/ashik/19gev_qa/qa_analysis_new/version-1/ashik_hist_111.root

Li-Ke: /star/data01/pwg/ashik/19gev_qa/qa_analysis_new/version-1/like_hist_111.root

We can discuss this before the next QA-board meeting.

Thanks,

Ashik

On Mon, Nov 29, 2021 at 2:25 PM Hu, Yu <yuhu@bnl.gov> wrote:

Dear Ashik and Li-Ke,

Thanks a lot for the hard work. I definitely like to join you in the discussion and maybe we can do something more before the meeting.

 Would you mind providing the root files to us? Maybe I can try to mark out the unique runs from a plot.  And we should try to confirm them after going back and checking the shift log.

Thank you.

Best,

Yu

ps:

It is understandable if you have small fractions of different bad runs. Let me explain a little bit here if you are interested:

The bad runs are found from 2 steps, 

1) global steps, where based on the whole mean and RMS it cuts out some runs as bad, same with all the previous analysis;

2) local step, where the code will try to find out some stable run regions(separated by jumps) and then have cuts on each stable run region

So if you are not using identical root files, in step 1, the mean and RMS will also be not identical. 

Then there might be some different runs. But I will assume they are just on the n-sigma cut edge, which should be very little, no more than 1 or 2 runs.

But in the next step, the stable run region is based on the error(statistics) on the runs you are interested in. The different statistics will lead to a different grouping, and then each small run-regions will have different "local-mean", "local-weighted-error", and "local-RMS". So some runs, especially at the beginning of the run region and the end of the run regions, will be different just because of the groups. There will also be a small number of runs.  

So overall some different bad runs an acceptable by this algorithm, and it's the best we can do in the blind analysis.

But now we have all the information available, we should check those runs directly with run log and see if they are really bad.

---

From: ASHIK IKBAL <ashikhep@gmail.com>
Sent: Monday, November 29, 2021 12:57
To: star-qaboard-l@lists.bnl.gov <star-qaboard-l@lists.bnl.gov>; STAR centrality focus group <star-centrality-l@lists.bnl.gov>
Cc: cebra <cebra@physics.ucdavis.edu>; Grigory Nigmatkulov <nigmatkulov@gmail.com>; Prithwish Tribedy <ptribedy@icloud.com>; Ma, Rongrong <marr@rcf.rhic.bnl.gov>; Hu, Yu <yuhu@bnl.gov>
Subject: Run-by-run QA for BES-II 19.6 GeV datasets

 

Hi Dan, Grigory, Rongrong, Prithwish, Yu, and all,

I and Li-Ke have been running the QA codes over the full statistics for BES-II 19.6 GeV datasets. We have the output root files having the required QA variables and we have analyzed both the root files (from Li-Ke and me) separately. We found 86 consistent bad runs and also found a few extra badruns that are not common between us. We can discuss this in the next QA-board meeting --- I would like to request Prithwish and Yu to be present in the discussion, if they are free. 

Please find the attached slides for the details.

Thanks,

Ashik

--

Ashik Ikbal

Research Associate

Kent State University

Department Of Physics

Kent, OH 44242, USA

Contact no.s: +1 631-504-7647 (USA)

                      +91 8250216234 (India)

Alternate e-mails: asheikh2@kent.edu

                             ashik@rcf.rhic.bnl.gov

                             ashik.ikbal.sheikh@cern.ch

--

Ashik Ikbal

Research Associate

Kent State University

Department Of Physics

Kent, OH 44242, USA

Contact no.s: +1 631-504-7647 (USA)

                      +91 8250216234 (India)

Alternate e-mails: asheikh2@kent.edu

                             ashik@rcf.rhic.bnl.gov

                             ashik.ikbal.sheikh@cern.ch