Star-fst-l mailing list

[Daniel Brandenburg <dbrandenburg.ufl AT gmail.com>]

Hi Shenghui,

I found that your code does not compile. Specifically the following issues:

1) The StFstXXX containers are not defined in StContainers.h (this causes several further issues)

2) In a few places "Ist" is still used in function names etc.

3) There is no test, do you have a test to run the code and verify that it works?

I already indicated to S&C that this code was ready, so they are waiting. But these issues need to be resolved before we can move forward.

Best,

Daniel

On Mon, Oct 25, 2021 at 5:58 PM Zhang, Shenghui <shenghui AT uic.edu> wrote:

Hi Flemming,

Sorry for the late response and please find my replies in line.

Best,

Shenghui

Hi Shenghui,

Thanks for posting the slides and all the work that went into this.
Some of the code snippets are hard to read from slides though.

I have a couple of comments/questions:

i) Open questions:
--- The Topo map is used for mid-rapidity tracking and is not relevant 
for FST,sTGC and FCS.
     There maybe a need for something similar for the forward tracking. 
Daniel?
     ome of the other Ist related stuff may be for mi-d-rapidity 
tracking.

A: For the open questions, I only want to make a remider about the FST parts should be included in these codes. I think someone else should take care of these.

ii) slide 7
     How are this calculations and tables related to what is implemented 
for the DAQ code? The DAQ code needs this info right away
and this cannot be redone later (pedestal, CMN etc) except for a subset 
of events. This was done for IST in the late stages, but is planed to be
done online for FST. I think this needs some discussion.

A: These calculations and tables are only for the non-zero data and pedestal run and are used to write calibration DBs.

iii) It would be good to have a writeup on definitions of the coordinate 
transformations that are needed. I know its being implemented in
the geometry descriptions, but it has to be used in the HitMakers and 
should be well understood. 

A: Right. I think Gavin have the detailed definitions of the coordinate transformations. When the alignment has beed finished, I will check with Gavin to make sure the transformation I used in StFstHitMaker/StFstHit is correct.

You say the calibrationMakecreated the CMN 
data, but thats too late. as it's applied in the online DAQ. How is this 
coordinated?

A: The calibrationMaker created the CMN data are used to only tell us the size of CMN. The CMN subtraction is performed in StFstRawHitMaker. The detailed iformation can be found in slide 4 of 

https://drupal.star.bnl.gov/STAR/system/files/FST_Software_Status_09282021.pdf

iv) I think som more thought should go into marking bad channels , apv 
chips as this is already need for online. We shoud not have two diffrent 
sources for this.

A: In the CalibrationMaker, we only marker as bad channels and apv chips when there are no any adc information for the channels and apv chips. The rejected bad channels and chips in the StFstRawHitMaker are based on the selections shown below. 

482 |   |   |   //skip dead chips and bad mis-configured chips

483 |   |   |   if (mConfigVec[apvId - 1] < 1 || mConfigVec[apvId - 1] > 9) { //1-9 good status code

484 |   |   |   |   LOG_DEBUG << "Skip: Channel belongs to dead/bad/mis-configured APV chip geometry index: " << apvId << " on wedge " <    < wedge << endm;

485 |   |   |   |   continue;

486 |   |   |   }

487 

488 |   |   |   //skip current channel marked as suspicious status

489 |   |   |   if (mRmsVec[elecId] < mChanMinRmsNoiseLevel ||

490 |   |   |   |   |   mRmsVec[elecId] > mChanMaxRmsNoiseLevel ||

491 |   |   |   |   |   mRmsVec[elecId] > 99.0)

492 |   |   |   {

493 |   |   |   |   LOG_DEBUG << "Skip: Noisy/hot/dead channel electronics index: " << elecId << endm;

494 |   |   |   |   continue;

495 |   |   |   }

) StFstHit
-- looks ok if one chekcs what's already in StHit; It would be usefull 
in the comments to define units, and precise definitions.E.g. the is
    a size in Z directions (units ..) . I guess this comes from 
segmentation in 4 radial segments of the sensors not Z which is nominal 
along the beam.

A: The detailed description have been included in all the codes.

 

How is charge uncertainty evaluated. I guess one can 
read the code, but general brief descriptions will be good.

A: The charge uncertainty is that of the cluster, whose calculation can be found in slide 7 (sigma_cluster) of https://drupal.star.bnl.gov/STAR/system/files/FST_Software_Status_09282021.pdf

slide 11)
   - why is the collection split over sensors. I would guess you need to 
access it on a per wedge(or mybe even ring)  when matching hits, but 
this
     of course depends how genfit wants the hits presented.The data 
structure layout should math what is needed in tracking, and it may be 
different
   for the forward tracking than for mid-rapidity with HFT.

A: As we can see from the right-bottom figure of slide 4, the transformation for local position to global one is performed sensor by sensor. So we need the sensor collection.

best Flemming

---

From: videbaek <videbaek AT bnl.gov>
Sent: Thursday, October 21, 2021 5:58 PM
To: Zhang, Shenghui <shenghui AT uic.edu>
Cc: Star-fst L <star-fst-l AT lists.bnl.gov>
Subject: Re: [Star-fst-l] Offline software

 

Hi Shenghui,

Thanks for posting the slides and all the work that went into this.
Some of the code snippets are hard to read from slides though.

I have a couple of comments/questions:

i) Open questions:
--- The Topo map is used for mid-rapidity tracking and is not relevant
for FST,sTGC and FCS.
     There maybe a need for something similar for the forward tracking.
Daniel?
     ome of the other Ist related stuff may be for mi-d-rapidity
tracking.
ii) slide 7
     How are this calculations and tables related to what is implemented
for the DAQ code? The DAQ code needs this info right away
and this cannot be redone later (pedestal, CMN etc) except for a subset
of events. This was done for IST in the late stages, but is planed to be
done online for FST. I think this needs some discussion.
iii) It would be good to have a writeup on definitions of the coordinate
transformations that are needed. I know its being implemented in
the geometry descriptions, but it has to be used in the HitMakers and
should be well understood. You say the calibrationMakecreated the CMN
data, but thats too late. as it's applied in the online DAQ. How is this
coordinated?
iv) I think som more thought should go into marking bad channels , apv
chips as this is already need for online. We shoud not have two diffrent
sources for this.
) StFstHit
-- looks ok if one chekcs what's already in StHit; It would be usefull
in the comments to define units, and precise definitions.E.g. the is
    a size in Z directions (units ..) . I guess this comes from
segmentation in 4 radial segments of the sensors not Z which is nominal
along the beam. How is charge uncertainty evaluated. I guess one can
read the code, but general brief descriptions will be good.

slide 11)
   - why is the collection split over sensors. I would guess you need to
access it on a per wedge(or mybe even ring)  when matching hits, but
this
     of course depends how genfit wants the hits presented.The data
structure layout should math what is needed in tracking, and it may be
different
   for the forward tracking than for mid-rapidity with HFT.


best Flemming



On 2021-10-20 05:01, Zhang, Shenghui wrote:
> Hi all,
>
>  Please find the detailed information of FST offline software below:
> 
> https://nam04.safelinks.protection.outlook.com/?url="https%3A%2F%2Fdrupal.star.bnl.gov%2FSTAR%2Fsystem%2Ffiles%2FFST_Software_Status_10202021.pdf&amp;data=04%7C01%7Cshenghui%40uic.edu%7Cdcc0c248b6ad470d058008d994e64633%7Ce202cd477a564baa99e3e3b71a7c77dd%7C0%7C0%7C637704539186276885%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=Xvgx3%2BvJY6UJTpugr8DN4y%2FFDiLR84L42UltbdvV7ck%3D&amp;reserved=0
>
>  Any comments, suggetions or questions?
>
>  The codes have been uploded to RCF:
>  /star/u/shenghui/shenghui/FST/offline/StRoot/
>
>  Best,
>  Shenghui
> _______________________________________________
> Star-fst-l mailing list
> Star-fst-l AT lists.bnl.gov
> https://nam04.safelinks.protection.outlook.com/?url="https%3A%2F%2Flists.bnl.gov%2Fmailman%2Flistinfo%2Fstar-fst-l&amp;data=04%7C01%7Cshenghui%40uic.edu%7Cdcc0c248b6ad470d058008d994e64633%7Ce202cd477a564baa99e3e3b71a7c77dd%7C0%7C0%7C637704539186276885%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=W%2FVBzCr55k78tHG3CLAvtp5jpkm6OJlNog7nzP8cgHA%3D&amp;reserved=0

--
Flemming Videbaek
senior scientist
videbaek @ bnl.gov
Brookhaven National Lab
Physics Department
Bldg 510D
Upton, NY 11973

phone: 631-344-4106
cell     :  631-681-1596

_______________________________________________
Star-fst-l mailing list
Star-fst-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/star-fst-l