c-dilks / trgmon15pAu

run 15 pAu trigger monitor

This is the code that builds www.star.bnl.gov/protected/spin/dilks/trgmon15pAu/200/trigger.html

Procedure

1. Execute getTrgEv to download all needed data from the RunLog database
   • trigger.dat is the main file that will be used
   • there are if statements in getTrgEv which select what goes into trigger.dat, such as requirents for nonzero no. of BBC events and JP1 events
   • an index is assigned to each run number, starting from the first one that is appended to trigger.dat; this index is an integer assigned to each run number so plots of things vs. run number are plotted vs. the index instead so they are easier to read
2. Execute mkHTML to build the HTML pages
   • this runs trgMon.C, which is important and described below
   • each page is written using other csh scripts
   • data files and HTML files will be copied to ./200 or ./500 depending on sqrt(s) being used
   • if RunLog SQL database is inaccessible, all the dat files that are used to build the HTML page are always maintained in ./200 or ./500; check the SQL server port map to see if the database has been moved and update getTrgEv accordingly
3. Execute copyToProtected, which will copy all the files to the proper directory for HTML access

trgMon.C

This script attempts to determine which runs are good to keep, by comparing the trigger rate of a run to the average trigger rate within an epoch. An epoch is a duration of time where the trigger rate is approximately constant; the epoch separators are inserted manually in trgMon.C for each trigger.

Trigger rate is computed as: (#events * prescale)/(#min. bias * min. bias prescale)

• The trigger rates are all drawn in TCanvas *ev_canv

Then the trigger rate is normalised by the epoch mean in each epoch;

• These normalised plots are drawn in TCanvas *nv_canv

Then cutoffs, defined under the epoch cuts, are used to decide whether a run should be marked as good or bad; the lower cutoff is called qa_cut_low and is less than unity and the higher cutoff is called qa_cut_high and is greater than unity

• The cutoff lines define a region around unity in the epoch mean normalised trigger rate plots (cno plots) for which runs are marked as good

A run is only marked as good if the duration is greater than 3 minutes (time_cut)

Sometimes, a run will cause anomolously high or low trigger rates and radically throw off the epoch mean; for these runs, we can manually mark them as bad and remove them from the computation of the epoch mean; this is done via the exclusion_list

Any other bad runs to be manually excluded are added to the variable exclusion_list

Trigger ratios are plotted to TCanvas *rat_canv

Luminosity plots are drawn automatically, to TCanvas *lum_canv

Prescale vs. run index are drawn to TCanvas *prescale_canv

Everything is written to runlog.root