Dan Guest (34ae248a) at 28 Mar 18:06
Updated documentation to use Derivation_tf rather than Reco_tf.
Hi @dguest, updated documentation for making new test files.
Hi @dguest, updated documentation for making new test files.
Ok true.. But then I need 30 files regardless and I have already copied all of that by hand and wouldn't want to go back and delete bunch of lines per file for 30 files again... I agree it's neater though.
Review checklist:
This is a good compromise for now. I wouldn't copy the whole configuration for each variation though: you can make a base file, where "track_systematics": []
and similar for jet systematics, and then make each variation more like
{
"file": "retag_minimal_base.json",
"track_systematics": ["something"]
}
this will only be around 4 lines per file rather than 60
Just an intermediate solution until in-parallel submission is done so that people don't have to change the config file 30x and then push changes 30x to git.
I just made a folder with config files containing track variations that can be specified as input config files in a bash for loop which calls grid-submit -c EMPFlowRetag_minimal_configs/EMPFlowRetag_minimal_${track_systematic}
etc.
Also includes changes from my other MR, so first merge the other one i guess.
Review checklist:
I would do this in the Athena
place below, and make it inherit from .run_athena
, that way you can also run the test with the rich
augmenter.
cheers
Okay good to know, thanks for checking, I'll adjust the documentation to include the AMI tag too
You need to add it to the .gitlab-ci.yaml
file, that's the one that actually steers the pipelines, the thing you edited is just the script it calls.
good question, I just looked, see here
https://atlas-ami.cern.ch/?subapp=tagsShow&userdata=p5981
I don't think there's any big difference from the default you'd get from that release, but there are a few extra settings. I probably wouldn't remake the test file for this but we should have the correct instructions.
I do asetup Athena,24.0.18 too - (probably naive question) is there anything else that the p-tag is specifying other than athena release?
I would be very surprised if it were clever enough but anything is possible.
There were new changes to the main upstream branch after I rebased...
Ok, i rebased. Why is the retag_minimal
test not in the CI tests as part of the pipeline? I added it to test-dumper.
I didn't implement it correctly? Can't see where I messed up.
Well the script I ran to actually create it was: Derivation_tf.py --CA True --inputAODFile ../data/mc20_13TeV.800030.Py8EG_A14NNPDF23LO_flatpT_Zprime_Extended.recon.AOD.e7954_s3681_r13144/AOD.27306425._000155.pool.root.1 --outputDAODFile DAOD_FTAG1.e7954_s3681_r13144_p5981.small.pool.root --formats FTAG1 --maxEvents 10
I guess it's not as clever as to get p5981 from the output DAOD file right?
That might explain why the metadata doesn't look the same between the file you created and the one where you truncated. But we also know something else isn't the same...
ok! So the derivation that you are creating here isn't the same as the one that has the p-tag you specified. You still need the --AMIConfig
(or similar) argument to make sure any specific flags are picked up. Sorry, I should have explained this before.
Hi @dguest, updated documentation for making new test files.