Project

General

Profile

Actions

Bug #536

closed

Feature #608: MUCH development June 16

Problem when using CbmMatchRecoToMC with much

Added by Florian Uhlig over 6 years ago. Updated over 3 years ago.

Status:
Closed
Priority:
High
Assignee:
Target version:
Start date:
12/18/2015
Due date:
05/13/2016
% Done:

100%

Estimated time:
50.00 h
Spent time:

Description

When using the task CbmMatchRecoToMC in macro/run/run_reco.C when the much detector is in the setup one gets many times the following error message

Error in <TClonesArray::At>: index -1 out of bounds

For me it looks like one of the data members of the much data classes is not filled correctly.


Files

links.png (36.2 KB) links.png All links point to the index -1 Volker Friese, 01/12/2016 05:11 PM
much_timebin0.patch (619 Bytes) much_timebin0.patch Julian Book, 01/26/2016 01:59 PM
CbmMuchDigiMatch.cxx (8.68 KB) CbmMuchDigiMatch.cxx Vikas Singhal, 01/27/2016 07:57 AM

Related issues

Related to CbmRoot - Bug #361: Bug in MuCh GEM digi matchingClosedVikas Singhal09/08/201510/01/2015

Actions
Has duplicate CbmRoot - Bug #537: Error in execution of "much_reco.C" and "much_analysis.C" in NOV15ClosedFlorian Uhlig12/23/2015

Actions
Has duplicate CbmRoot - Bug #538: FACING ERROR IN MATCHING OF STS TRACK AND MUCH TRACKClosedVolker Friese01/06/2016

Actions
Has duplicate CbmRoot - Bug #532: Error in Littrack while running much_reco.CClosed12/16/2015

Actions
Actions #1

Updated by Volker Friese over 6 years ago

  • Status changed from New to Assigned
  • Assignee set to Andrey Lebedev

Andrey, could you please have a look?

Actions #2

Updated by Anna Senger over 6 years ago

  • Assignee changed from Andrey Lebedev to Partha Pratim Bhaduri

In MuCh there is no matching between MC and digis (clusters, hits). The MuCh group has to update MuCh software.

Actions #3

Updated by Volker Friese over 6 years ago

  • File links.png links.png added
  • Assignee changed from Partha Pratim Bhaduri to Vikas Singhal

Anna is right (see figure attached). The match objects are empty. This is connected to #361 as temporary solution to decrease the file size and surely has to be fixed. For the time being, one could think of removing MUCH matching from CbmMatchRecoToMC, in order to avoid the annoying output.

I wonder whether the matching of MUCH data to MC is used by anybody, since it is not operational for quite some months.

I should add that not only the links of GEM digis but also those of straw digis are empty.

Actions #4

Updated by Julian Book over 6 years ago

this issue has been fixed by Vikas in revision r9171
CbmMatchRecoToMC works as it should however since the matches were always reseted another bug was hidden. Now you get negative timebins in
CbmMuchDigiMatch::AddCharge: Warning t=16.648596 < fT0=1460.421878
which causes tons of errors alla
Error in <TArrayD::operator[]>: index -1073 out of bounds (size: 456, this: 0x1a358cf8)
this can be avoided with the patch attached but the much group should find out why the fT0 get that large for some events.

the fix throughs an now an error (not a warning):
Error in <CbmMuchDigiMatch::AddCharge>: Warning t=30.335243 < fT0=1460.421878, negative timebin! use bin=0 instead

Actions #5

Updated by Partha Pratim Bhaduri over 6 years ago

Hello,
Interestingly such warning does not appear, when I ran the reconstruction macro with only Pluto. However for UrQMD tracks I also see this warning message.

Partha

Actions #6

Updated by Vikas Singhal over 6 years ago

  • Status changed from Assigned to In Progress
  • % Done changed from 0 to 70

Yes Partha,

That warning is in if statement, so in all instances code does not go in that block.
I will update you more what that code is doing.
But are you getting any warning message in the modified codes which I have sent you?

Kindly Update.
Vikas

Actions #7

Updated by Vikas Singhal over 6 years ago

Dear Julian,
I had modified the CbmMuchDigiMatch.cxx file for the Warning on negative timebins.

I could not commit into the repository as I do not have the permission to write under data/much.

attached the modified CbmMuchDigiMatch.cxx file.

Regards
Vikas

Julian Book wrote:

this issue has been fixed by Vikas in revision r9171
CbmMatchRecoToMC works as it should however since the matches were always reseted another bug was hidden. Now you get negative timebins in
CbmMuchDigiMatch::AddCharge: Warning t=16.648596 < fT0=1460.421878
which causes tons of errors alla
Error in <TArrayD::operator[]>: index -1073 out of bounds (size: 456, this: 0x1a358cf8)
this can be avoided with the patch attached but the much group should find out why the fT0 get that large for some events.

the fix throughs an now an error (not a warning):
Error in <CbmMuchDigiMatch::AddCharge>: Warning t=30.335243 < fT0=1460.421878, negative timebin! use bin=0 instead

Actions #9

Updated by Vikas Singhal over 6 years ago

  • Status changed from In Progress to Resolved
  • Estimated time set to 50.00 h
Actions #10

Updated by Julian Book over 6 years ago

before you close the ticket could you please explain:
why does the t<fT0 only pops up for the urqmd part and not for pluto particles?
is your current fix,not taking the charge of these point into account but still adding them as cbmlink to the list, valid?

Actions #11

Updated by Florian Uhlig over 6 years ago

  • Due date set to 05/13/2016
  • Parent task set to #608
Actions #12

Updated by Florian Uhlig over 6 years ago

  • Target version changed from FEB16 to June16

Florian Uhlig wrote:

When using the task CbmMatchRecoToMC in macro/run/run_reco.C when the much detector is in the setup one gets many times the following error message

Error in <TClonesArray::At>: index -1 out of bounds

For me it looks like one of the data members of the much data classes is not filled correctly.

Actions #13

Updated by Volker Friese over 3 years ago

  • Status changed from Resolved to Closed
  • % Done changed from 70 to 100

It seems this issue can be closed since there is not input since three years.

Actions

Also available in: Atom PDF