Support #1578: Vector input to CbmMcbm2018EventBuilder - CbmRoot - CBM

Actions

Copy link

Support #1578

closed

Vector input to CbmMcbm2018EventBuilder

Added by Vikas Singhal over 2 years ago. Updated over 1 year ago.

Status:

Closed

Priority:

Normal

Assignee:

Vikas Singhal

Target version:

APR21

Start date:

02/26/2020

Due date:

% Done:

100%

Estimated time:

Description

Dear Florian,

We try to use the CbmMcbm2018EventBuilder for simulated timeslice, but as it expects Branches of different subsystem digis in TClonesArray and STS, MUCH, TRD providing digi branch in vector after digitizer therefore it does event building except mentioned 3 subsystem for simulated timeslices.
Can you modify CbmMcbm2018EventBuilder accordingly.

Regards
Vikas

Actions

Copy link

Updated by Florian Uhlig over 2 years ago

Status changed from New to In Progress
Assignee changed from Florian Uhlig to Vikas Singhal
% Done changed from 0 to 40

Volker modified meanwhile all digi classes and also adapted the task CbmMcbm2018EventBuilder. Could you please check if it is working now.

Actions

Copy link

Updated by Pierre-Alain Loizeau over 2 years ago

My commit r15878 probably breaks it again due to the branch names. Will be fixed SoonTM (this evening or tomorrow) when implementing #1585 in all mCBM classes.

Actions

Copy link

Updated by Vikas Singhal over 2 years ago

Dear Pierre,

Today I tried with cbmroot r15896, and tried to do event building of simulated data. So just for update getting 2 issue:-

1. By default, event builder is looking for CbmMuchBeamTimeDigi but for simulated timeslice CbmMuchDigi is the branch name. Therefore I disable line number 67 in CbmMcbm2018EventBuilder.

fDigiMan->UseMuchBeamTimeDigi();

2. After this got another error which you mentioned yesterday during Software meeting (getting boost any_cast error.) ===================
I run_reco_tb_event: Starting run
[INFO] FairRunAna::Run() After checking, the run will run from event 0 to 2.
[INFO] Begin of TS 0
[FATAL] Failed boost any_cast in Digimanager::Get for a digi of type CbmMuchBeamTimeDigi ==================

Update accordingly.
Vikas

Actions

Copy link

Updated by Volker Friese over 2 years ago

Dear Vikas,

what is the status here?

Actions

Copy link

Updated by Pierre-Alain Loizeau over 2 years ago

Assignee changed from Vikas Singhal to Pierre-Alain Loizeau

It should be checked if the issue is solved in the development branch. If it is the case we should backport the fix to trunk and release.

=> After rereading this issue and the #1585, I think I understand again where the problem lies: the event builder is using CbmMuchBeamTimeDigi, which is only avavilable for real data.
=> We would need a switch/flag to select between CbmMuchDigi and CbmMuchBeamTimeDigi, I will try to implement this this evening
=> This hack will anyway pop up again when we cleanup the situation with the 2 Digi classes, so no need for a separate issue to remember that this should be cleaned up.

=> r.16468 should provide the needed functionality. Please let me know if this is enough or if we need further tuning.

Actions

Copy link

Updated by Pierre-Alain Loizeau over 2 years ago

Status changed from In Progress to Feedback
Assignee changed from Pierre-Alain Loizeau to Vikas Singhal
% Done changed from 40 to 90

Actions

Copy link

Updated by Volker Friese over 2 years ago

Target version changed from APR20 to APR21

Since this affects the development branch mCBM2020, which will be merged to the main directory in git, this issue need not block APR20. I thus change the target version.

Actions

Copy link