search

alcap-org / AlcapDAQ

Alcap DAQ
alcap-org.github.io
8 stars 0 forks source link

running the calibration #243

Closed benkrikler closed 10 years ago

benkrikler commented 10 years ago

Hi guys,

I'm sorry to repeat a question that's more or less been asked before, but I'm struggling to find the right set of commands to do this and it seemed daft to waste more time on it. I want to run the calibration over the active target datasets, probably for all channels, though the muon counters and SiR* channels seem the most important.

How can I use the run scripts to just set things running through all runs in that dataset? Just the basic commands to set up the run db and launch the jobs would be a help.

For what it's worth, could we just calibrate all channels for all runs now, and then just have one standard, canonical calibration DB?

jrquirk commented 10 years ago

Summarized from here, I think this is it

$ ./rootana -i tree.root -o out.root -m configurations/pedestalsandnoise_calib.cfg -c
$ scripts/merge_calibration_database.py calibration.db calib.run.PedestalAndNoise.csv
$ ./rootana -i tree.root -o out.root -m configurations/timeoffset_calib.cfg
$ scripts/merge_calibration_database.py calibration.db calib.run.CoarseTimeOffset.csv

The last one takes a while. You can disable channels by modifying timeoffset_calib.cfg.

For the most part I think we've all been using the same calib DBs, but I'm not against having one place. If you run it all I'll use it. If it's going t0o slow for you I can help run. We just need to make sure to save all of the CSV files until we're certain everything is okay.

benkrikler commented 10 years ago

Thanks John. I'm more interested in the run scripts though. I read through the wiki on the production scripts and our previous exchanges about setting up the DBs and running, but I'm still not very sure what commands I need to issue to get running over a particular dataset with the scripts. I could write this all in a bash script easily, but if we've got your production scripts then it would be nice to use them since they already take care of everything and in a very nice way. A simple bullet-point walk-through would be really useful.

jrquirk commented 10 years ago
  1. Put calibration.db in home directory (grid jobs for these scripts are based there)
  2. Run pedestal/noise calibration
    • AlcapDAQ/analyzer/batch/jobscripts/run_production.py --production=rootana --new=pedcalib --version=3 --modules=~/AlcapDAQ/analyzer/rootana/configurations/pedestalsandnoise_calib.cfg --database=~/production.db --dataset=Al100 --calib
    • AlcapDAQ/analyzer/batch/jobscripts/run_production.py --production=rootana --modules=~/AlcapDAQ/analyzer/rootana/configurations/pedestalsandnoise_calib.cfg --database=~/production.db --dataset=Al50b --calib
    • AlcapDAQ/analyzer/batch/jobscripts/run_production.py --production=rootana --modules=~/AlcapDAQ/analyzer/rootana/configurations/pedestalsandnoise_calib.cfg --database=~/production.db --dataset=Al50awithNDet2 --calib
    • ...
  3. Merge into calibration database
    • AlcapDAQ/analyzer/rootana/scripts/merge_calibration_database.py ~/calibration.db ~/calib.run0????.PedestalAndNoise.csv
  4. Run time offset calibration
    • AlcapDAQ/analyzer/batch/jobscripts/run_production.py --production=rootana --new=timecalib --version=3 --modules=~/AlcapDAQ/analyzer/rootana/configurations/timeoffset_calib.cfg --database=~/production.db --dataset=Al100 --calib
    • AlcapDAQ/analyzer/batch/jobscripts/run_production.py --production=rootana --modules=~/AlcapDAQ/analyzer/rootana/configurations/timeoffset_calib.cfg --database=~/production.db --dataset=Al50b --calib
    • AlcapDAQ/analyzer/batch/jobscripts/run_production.py --production=rootana --modules=~/AlcapDAQ/analyzer/rootana/configurations/timeoffset_calib.cfg --database=~/production.db --dataset=Al50awithNDet2 --calib
    • ...
  5. Merge this calibration into your database
    • AlcapDAQ/analyzer/rootana/scripts/merge_calibration_database.py ~/calibration.db ~/calib.run0????.CoarseTimeOffset.csv
  6. Add the Energy table to calibration.db
    • sqlite3 ~/calibration.db "CREATE TABLE Energy(run INT, channel TEXT, gain REAL, offset REAL)"
  7. Run production just like above. You only need to calibrate datasets you want to run that we haven't yet. Also, omit the --calib flag.
benkrikler commented 10 years ago

Thanks John that's really helpful. I hit problems with tilde expansion but replacing this with $HOME seemed to do it.