HOW TO RUN CalibGenCAL Mark Strickman's document tells some of this: http://heseweb.nrl.navy.mil/glast/CAL_ATDP/FM107/SummaryData/ LAT-TD-05595-01%20calibGenCAL_v3_description%20init%20release.pdf I also found it searching “Lat docs” : https://oraweb.slac.stanford.edu:8080/pls/slacquery/bbrdownload/ LAT-TD-05595-01-calibGenCAL_v3_description_init_release.pdf?P_FRAME=GLAST&P_DOC_ID=16716 David's stuff is in /nfs/slac/g/svac/dsmith/ChuckP/calibGenCAL/v3r5p2 Using /nfs/slac/g/svac/dsmith/ChuckP/CMTPATH I build against nearly the same version of EngineeringModel as the pipeline. I run three executables: 1) runCIFit.exe which reads in ciFit_option.xml Personally, I make files like ciFit_optionFM104.xml, and then cp ciFit_optionFM104.xml ciFit_option.xml before execution. You have to specify two input files, which you find as follows: The online script "calibGEN" runs calu_collect_ci_singlex16.digi six times in a row. You need the third (for low energy diode) and the fourth (for high energy diode) element for the LE1 and HE1 entries, respectively. The output of this is ../output/*ci.intnonlin*.xml 2) Once you have ci.intnonlin.xml, you can run runMuonCalib.exe which reads in muonCalib_option.xml It generated MevPerDac and adc2nrg file besides pedestal and light asymmetry. You have to specify as input the intnonlin file, and a muon file. The muon file serves two purposes: a) pedestal determination using 10,000 events. So you need a NON ZERO SUPPRESSED muon file. b) Determination of the asymmetry. They ask for 1,000,000 but half of that will do. So you need a REALLY BIG FILE. The files I've used to get asym were taken at NRL ("Pshp") using an external muon trigger (scintillator paddles) and were non-zero suppressed. If you have a file like that, you can runMuonCalib.exe only once, and it does first pedestals, then asym. Sasha discourages us from updating the asym files too often. So I've used the mu_optical files from the CPT's to get pedestals, and then ignore the asym output. If I want to update the asym output for a new version of calibGenCAL, I re-use the Pshp files. When we get the calibGen calibDAC muTrg data next week, Sasha suggested chaining all the six (?) muTrg files in the muonCalib_option.xml list for asym. Individual files have masked crystals, but using all of them gives all crystals. Pedestals would come from... CPT? I'm not sure if muTrg gives us non-zero suppressed muon files. muonCalib_option.xml has flags that allow you to read pedestals from the file if you run the job twice (first for peds, second for asym, with different muon input files). The output is then ../output/*peds* and ../output/*asym*. It also produces mevperDac and adc2nrg files 3) runMuTrigEff.exe which reads in MuTrigEff_option.xml To generate FLE bias table it's enough to run runMuTrigEff ONCE giving as an input the first singlex16 file and the first and the second muon files, produced by muTrig suit The output of is files NOT used by SAS, which are the FLE and LAC settings files. The generation of FleFheBias file is included now into runMuTrigEff - you have to edit the following line muTrigEff_options.xml file: (boolean) generate xml output At present, the ciThold file is "fake" and was made by hand by Zach, with 3400 for every channel. How to check that it's okay =========================== I have a couple of root *.C macros that I look at sometimes, and sometimes I copy the .txt files to windows and use excel to plot the values vs channel index. I confess that checking the outputs is a weak link in my chain.