Hi Pawel, hope you are doing fine. From our last two meetings on Dec 9 & 15, we agreed on finding out what is there in the efficiency files you have made for auau pp, and maybe other species. Would you please let us know 1) what the afiles lready produced have and what they lack 2) your input on making efficiency files to be placed in a common area, especially what options should neccessarily be used so that the files contain a complete information. 3) we need volunteers to divide the task among us. Would you be able to organize it? I am working on making efficiency files for run05. 4) The options below suggest we identify runs where detectors may have died option -g (--t5effic) - default value is "T3-T5" but for runs when T3 was dead one should use "T2-T5". option -T (--tpc) - default value is "FullBfs" but for runs when T3 was dead one should use "FullBfsNoT3" ======================================================= On my part, I did try the efficiency scripts and got some plots. However, I needed to make some changes to the macro /brahms_app/ps_app/fs/receffnew/scripts/MapEfficiency.C 1) //root>MapEfficiency("run04",63,8,861,1,"CXP","T1",kTRUE,kTRUE,4,1000000) --->does not work root>MapEfficiency.C("run04_<species>",200,4,861,"CXP",1,"T1",kFALSE,kFALSE,4,1000000) ----> works where <species> = auau,cucu etc 2) May be what you have in your directory works fine, but I needed to make some changes to the code I checked out from CVS . The macro MapEfficiency breaks if there are no files to process. So, I needed to add two variables Int_t NFilesA = 0,NFilesB = 0; which are passed to GetEfficiency(...,Int_t& NFilesA). These variables are incremented in GetEfficiency() depending on how many files are there to process. If there are no files,their values would be 0 and the code would not break when you try to plot histograms or write them to a root file. if(anaA)GetEfficiency(Period.Data(),det.Data(),VarRequested.Data(),distCut, EffX_A,DistX_A,EffY_A,DistY_A,EffZ_A,DistZ_A,EffCXAX_A,DistCXAX_A, NameA,Eff_A,Err_A,1,MaxEvents,UpperX,LowerX,iroot, runList,NFilesA); if(anaB)GetEfficiency(Period.Data(),det.Data(),VarRequested.Data(),distCut, EffX_B,DistX_B,EffY_B,DistY_B,EffZ_B,DistZ_B,EffCXAX_B,DistCXAX_B, NameB,Eff_B,Err_B,1,MaxEvents,UpperX,LowerX,iroot, runList,NFilesB); Then , wherever they appear,I changed the conditionals if(anaA) , if(anaB) to if(anaA && NFilesA), if(anaB && NFilesB) 3) For the macro /brahms_app/ps_app/fs/receffnew/detmatch/InspectOffsets.C //root>Inspect("T4T5",11301) - will print offsets and sigmas for T4T5 combination. ---> Does not work root>Inspect("T4T5",11301,"run04","_<species>_NEW") - will print offsets and sigmas for T4T5 combination. ############################################# cheers, Selemon, -----Original Message----- From: Pawel Staszel [mailto:ufstasze@if.uj.edu.pl] Sent: Mon 12/5/2005 9:20 AM To: Bekele, Selemon Subject: Re: [Brahms-dev-l] Making ltr files Dear Selemon, Sorry for the late answer (see my notes below). Bekele, Selemon wrote: >Hi Pawel, > > thanks for the reply. Things are working now. >I managed to make one efficiency file using one run. >I would like to use more runs for a given field setting. >I checked one of the files you made and saw that you >can put in as many runs as you want. My question is at >which point do you specify many runs, during the generation >of the offset files or during the production run? > >This is what I do for pre-production >---------------------- > bratmain fsRecEff.C -r 10578 -n 15 --prod=3 --ltrver=0 --gtrprod=3 --gtrver=0 --offsetmode=1 --iroot=10 -e 20000 --tree=true >---------------------- > >Now to use many files, is the following what I should do before the production run? > > root>.L UpdateEfficParamFile.C > root>AddRun(10578,0,<period>_<species>) > root>AddRun(10579,0,<period>_<species>) > root>AddRun(10580,0,<period>_<species>) > root>AddRun(10581,0,<period>_<species>) > for set of runs you can do e.q. : AddRun(10578,10581,<period>_<species>), then it will add all runs between 10578 and 10581. > >and for production >---------------------- > bratmain fsRecEff.C -r 10578 10579 10580 10581 -n 15 --prod=3 --ltrver=0 --gtrprod=3 --gtrver=0 --offsetmode=0 --iroot=10 -e 20000 --tree=true >---------------------- > >do I have to use -r before each run number? > > No, you should use condor scripts, condorOffset.sh - to run offset runs and condorProd_EncT5.sh to run "production" runs. The command line looks like e.g. >./condorProd_EncT5.sh 10578 10579 10580 10581 ...... (the is no limit for number of runs you put). However, there are some parameter in the condor scripts that you may need to modify. Is is gtr production number, ltr production number and iroot. iroot is just used to disentangle between different output files. Usually I use iroot=10 for offset jobs and iroot=1 for production jobs. By this is absolutely arbitrary. The runs are merged when you run script MapEfficeincy.C. First you should edit the script and add the list of runs in the GetRunArray method in the was as it is done for other settings. MapEfficieincy loops over all runs and produces averaged efficiency maps. These maps (histograms) can be then used by the EfficiencyCalculator. See also README. Cheers, Pawel. >I want to do this correctly so that I can generate efficiency files >for all settings which are going to be kept in a common area. That way > everybody can access them and there would be no need for any one to > make them again. I appreciate your help in this matter. It would be great >if you could send me an example just as you would type them on the command line >when you made the efficiency files for AuAu@200 GeV. > >Thanks in advance, > > >-----Original Message----- >From: Pawel Staszel [mailto:ufstasze@if.uj.edu.pl] >Sent: Wed 11/30/2005 7:41 AM >To: Bekele, Selemon >Subject: Re: [Brahms-dev-l] Making ltr files > >Hi Selemon, >Looks like a standard problem with db access. >Make sure that you have these lines: > >brahms-db0.rcf.bnl.gov BrahmsMain query query-only >brahms-db0.rcf.bnl.gov BrahmsCalib query query-only >brahms-db0.rcf.bnl.gov RUNDB query query-only >brahms-db0.rcf.bnl.gov BrahmsGeom query query-only >brahms-db0.rcf.bnl.gov FILECATALOG query query-only > >in your ~/.bratdbrc, otherwise you need to specify user and password "by >hand". >Hope it will help, >Pawel. > >Bekele, Selemon wrote: > > > >>Hi Pawel, >> >> hope you are doing fine. I was trying to start making efficincy files. I get the error found below when I do >> >>--------------- >>fsRecEff.C -r 10578 -n 15 --prod=3 --ltrver=0 --gtrprod=3 --gtrver=0 --offsetmode=0 --iroot=10 -e 20000 --tree=true >>--------------- >> >>********************************************* >> >> B R A T M A I N >> >> BRAT Version: 2.30/04 >> BRAT Date: Tue Jan 4 18:09:34 2005 >> ROOT version: 4.04/02 >> Host: rcas0029.rcf.bnl.gov >> Started: Tue Nov 29 18:13:48 2005 >> Script: fsRecEff.C >> >>********************************************* >>Configuration script fsRecEff.C >>trying to connect to file catalog >>Error in <TMySQLServer::TMySQLServer>: connection to database FILECATALOG on brahms-db0.rcf.bnl.gov failed (error: Access denied for user 'query'@'rcas0029.rcf.bnl.gov' (using password: NO)) >>Error in <BrRdbmDb::Connect>: User query couldn't connect to database FILECATALOG @ brahms-db0.rcf.bnl.gov >>Error in <BrRdbmMainDb::ConnectToDb>: couldn't connect to database FILECATALOG >>Error: illegal pointer to class object fileCatalogsDb 0x0 668 FILE:fsRecEff.C LINE:118 >>Fatal in <BrMainModule::Instance()>: BrMainModule not instantized yet >>aborting >>Generating stack trace... >>/usr/bin/addr2line: bratmain: No such file or directory >>/usr/bin/addr2line: bratmain: No such file or directory >>0x080494be in __gxx_personality_v0 + 0x492 from bratmain >>0x02c4578a in __libc_start_main + 0xda from /lib/tls/libc.so.6 >>0x080490fd in std::ios_base::Init::~Init() + 0x31 from bratmain >>./runRecoProd: line 1: 7119 Aborted (core dumped) bratmain fsRecEff.C -r 10578 -n 15 --prod=3 --ltrver=0 --gtrprod=3 --gtrver=0 --offsetmode=0 --iroot=10 -e 20000 --tree=true >> >> >>---------------------------------------- >> >>Selemon, >> >> >> >> >> >> > > > -- ---------------------------------------------------------------------- | Pawel Staszel | | Jagiellonian University | | Institute of Physics email: ufstasze@if.uj.edu.pl | | Reymonta 4 phone: (+48) 12 663 5712 | | Krakow 30059 FAX: (+48) 12 633 7086 | | Poland | ---------------------------------------------------------------------- _______________________________________________ Brahms-dev-l mailing list Brahms-dev-l@lists.bnl.gov http://lists.bnl.gov/mailman/listinfo/brahms-dev-lReceived on Fri Dec 16 18:14:09 2005
This archive was generated by hypermail 2.1.8 : Fri Dec 16 2005 - 18:14:23 EST