[Gate-users] problem with gjs, gjm, and root files

[Hermann Fuchs]

Dear Parisa,

I assume you are using the newest Gate release (8.0).
It seems gjs is not working smoothly any more. Previously, the jobs were 
splitted by time.
E.g. A (default) simulation time of 1 second was assumed. A split in 10 
jobs would result in job1 simulation from 0 to 0.1 seconds, job2 from 
0.1 to 0.2 seconds etc.
In the newest release this does not work any more. Instead every split 
simulates all particles.
So the number of particles per split is always the same as in the 
initial simulation (10 splits, eg.g 10x the particles of the initial 
simulation etc.).

Anyway, the root merger hadd will correctly merge root files. I am not 
sure about how it handles the eventID, though.

I hope I could help a bit,
Hermann


Am 2017-08-29 11:11, schrieb Parisa Khateri:
> Dear Gate developers/users,
> 
> Recently, I have tried gjs and gjm to split a simulation to smaller
> parts to be able to run them in parallel on a cluster.
> I used the following command to split the mac file:
> gjs -numberofsplits 20 -clusterplatform condor -condorscript
> path_to_condor_script/condor.script path_to_mac_file/my_macfile.mac
> 
> Then I ran them in parallel and got the output root files in which the
> event_IDs start from zero! So when I merge them the merged root file
> is a mess in the branch of event_IDs. My csorter works based on
> event_IDs. I think you can imagine where it ends up with :-(
> I appreciate if anybody can give me a hand to solve this issue.
> 
> Thanks in advance,
> Cheers,
> Parisa
> _______________________________________________
> Gate-users mailing list
> Gate-users at lists.opengatecollaboration.org
> http://lists.opengatecollaboration.org/mailman/listinfo/gate-users