MainPage:Nuclear:DVCS-3:Software
| ⇐ Back to the Main_Page |
| ⇐ Back to DVCS-3 |
DVCS-3 Software and Analysis How-To
by Salina Ali
Ideas and Progress
- Update 9/1/2018 - ana_mod1run_ttree.C is working fine (no bugs). Need to run for accidentals and attempt to subtract them and check missing mass squared
- Update 7/23/2018 - Submitted runs I have already for wf analysis, need to do ntup analysis, e.g. ana_mod1run.C script. Changed some things in ntup analysis. Andres pointed out some inconsistencies.
- Idea: Perhaps upgrade ana script without using the TNtuple Class, and instead use TTree? Worth the time? (Update 9/1/2018: Yes as TNtuple has a limit on the amount of branches it can hold..see ana_mod1run_ttree.C for the updated script.)
- Update 7/20/2018 - Success with waveform analysis .xml and .pl files (Andres helped with some errors). Shared with Mongi. Need to do for 100+ runs. Having problems with ana_mod1run.C script.
- Update 7/14/2018 - Currently working on testing a script that reads run list and loops around to produce the wf2p split files by itself. This will be a .xml script. Instructions are here: xml auger command files
- Idea 7/13/2018 - Mongi and I think perhaps good idea to use most recent version of analyzer (1.6.3) and root 6.12.06 and upgrade scripts.Currently using analyzer 1.5.36 and root 5.34.36. However, not sure if we should proceed as things may break along the way.
- Update 7/13/2018 - Installed soft-2 in /work/halla/dvcs/disk1/salina/
- Update 7/12/2018 - ana_mod1runkin48.C is the most recent working "ana" script outputting the rootfiles. I used this to test my wf2p_kin48_4** files as I need to do the waveform analysis for about 160 runs!
- Update 7/11/2018 - dowf_sfa.C is the most recent working waveform analysis script. Does not include any compiling capabilities (e.g. headers like #include blah blah..) wf2p_kin48_4_**.root are output split by split.
GEant4 Simulation for pi0
Bill's simulation is suitable only for photons - Mongi has transformed it to pi0's. Currently trying to get this to work with visuals.
Jobs via Batch/Auger
All information in this section is referencing this page, and Bill's guide for DVCS GEant4 simulations via Bill's elog entry #516
Submitting GEant4 Simulation Jobs
Go to /volatile/halla/dvcs/salina/hallasim/HallASim
Submitting jobs to Auger or Batch
- Need minimum 2 scripts for job submission..but can do multiple things via .xml
DVCS Replay on JLab's farm
Step 1- Get libraries for replays
- The DVCS database, or "DB"
- The main DB should include Frederic's corrections back in June 2018: see the eLOG entry https://hallaweb.jlab.org/dvcslog/12+GeV/517 here.]
- onldvcs
Compiling libraries
Need to compile libraries before doing replay. Tried to follow Bishnu/Alexa 's instructions but didn't work on ifarm 1401 and 1402. To compile, do
analyzergSystem->Load("libDVCS.so")gSystem->Load("libTDVCSCalo.so")
- Make sure libTDVCSCalo.so is loaded to the onldvcs library.
DVCS software install guide
DVCS-specific libraries
Carlos's past instructions are also very helpful and still useful. I use "make" instead of the "cvs".
soft-2
- Copy soft-2 library to your own work directory. Should only need to if doing work specific to soft-2. Or else can source /work/halla/dvcs/disk1/DVCS3/soft-2
cp -r /work/halla/dvcs/disk1/DVCS3/soft-2 .
- Also copy UsingNamespaceStd.h and rules.make
- Then do
cd soft-2make cleanmake biglib
- Should install.. can test by making a shell script and sourcing.. need to make sure root is the same version of analyzer as when you compiled the soft-2 library. For example, if you used analyzer 1.5.3 with root 5.34.34 to compile and build soft-2, that should be used whenever you load "libDVCS.so" or anything from the soft-2 library
Set path like:
DVCS gen
- Hasn't been changed in years, can just copy into working directory
- Can use mine or Mongi's in /work/halla/dvcs/disk1/mongi/dvcs_gen
DB
- Updates more frequently, current working one is /work/halla/dvcs/disk1/salina/DB as of 12/12/2018