Pion LT Software Meetings » History » Revision 28
Revision 27 (Nathan Heinrich, 04/16/2021 05:07 PM) → Revision 28/36 (Stephen Kay, 04/19/2021 04:18 PM)
{{>TOC}}
h1. Pion LT Software Meetings
Pion LT software meetings. Please add your talk to the meeting ahead of time. Add the newest meeting to the top of the list!
----
h2. 2021.04.19 PionLT Software Meeting
* Nathan Presented his Selections for Heep data
** [[https://redmine.jlab.org/attachments/download/1226/PionLT_Elasitics_Selection_Nathan.ods]]
** [[https://redmine.jlab.org/attachments/download/1227/Nathan_PionLT_Heep_selection_2021.pdf]]
** Indicated chosen settings include timings for minimum stats goals
** HeeP singles, two spect at some angle/P on one or two settings would be desirable
** HeeP delta scan can be done two ways, should ideally do both
*** Vary angle
*** Vary field
** Question from Casey - Current variation for a Heep setting? Would be nice if we can fit it in
* Stephen showed and discussed new RFtime hcana class and new MM calculations added to THcSecondaryKine
** Both in and working, MM calculation still be tested a little
** Pushed to JLab repo
* Discussion of remote counting house, still going ahead?
** Info given from Brad, but not progressed much further yet (due to financial constraints fitting it into the previous FY)
** Garth going to inquire further
----
h2. 2021.04.12 PionLT Software Meeting
* Discussed HeeP studies further, including event requirements
** Clarified that 10k events is the lower limit for singles
** 50k events for coin runs
* Need to tweak all DEF-cuts files, want *All Events*, can cut out EDTM manually after as needed
** Carlos wants to use EDTM events so can't remove them
** Ali will need to tweak some cuts accordingly to include a removal of the EDTM events
* Stephen highlighted that HeeP physics analysis scripts are in major need of a refresh
** Still using old TProof based method, should switch to Python
** Need more robust file checking/finding in script too
* Garth highlighted that Peter Bosted had made it very clear that the Physics settings where the current was varied were *incredibly* useful
** Want to try working some runs like this into the upcoming run plan
** Allow us to check FADC efficiencies as fn of rate/current
----
h2. 2021.04.05 PionLT Software Meeting
* I (Stephen) wasn't there so some minutes would be nice!
----
h2. 2021.03.29 PionLT Software Meeting
* Jacob updated on HeeP studies
** Fixed rate issues seen previously
** Still seeing odd results, can't reconstruct earlier rates
*** Factor ~10 out, probably some odd units?
* Carlos clarified issues Stephen was having with new Helicity stuff in the replay
* Vijay suggested switching to Etottracknorm in some Physics analysis (and including it in the DEF-files)
** Should definitely be kept, doesn't make a huge difference for low level analysis though
*** Working this into scripts will take a little bit of work, can update it as we go along
** Also including Aero X/Y in all DEF-files now too
----
h2. 2021.03.22 PionLT Software Meeting
* Casey reported that he had managed to get hcana to compile with root 6.18.00
** Stephen mentioned he'd had some conflicts with 6.18.04 - will test 6.18.00 too
* Jacob seeing some replayed files with streamer info missing errors
** Stephen mentions this is likely just an incomplete or failed replay, if it keeps happening though more info will be needed
* Casey reported on some more updates to his timing window scripts
** Now work with HMS windows too
** Mentioned that Bill Henry is looking at some potentially useful pulse amp plots, useful for other detectors
** Pulse int vs timing windows for different detectors and other plots
* Stephen gave an overview of upcoming software tasks
** Big item on the near horizon is the upcoming second replay of the kaon data, not independent of the pion software work!
*** Need updated report files, new efficiencies in the file, new information in the file
*** DEF-Files still need updating and cleaning, what can we trim out?
*** Which trees do we actually need in each type of replay
*** Which types of events should we actually specify -> ALL_Events vs ALL_Events_No_EDTM issue Richard saw recently
*** Helicity scalers included now too
*** Want to cross check electronic live time multiple ways
** A point was raised that we should aim to verify selection cutting in the DEF files yields the same result as post processing the replays with the "same" cuts
** Carlos pointed out that we need to be sure to compare the *exact* same value in our report files and in our data
** Casey requested calibration replay and shell scripts
*** Stephen forwarded these on
** SHMS Calibration code was not in latest hallc_replay repository
*** Casey mentioned Bill Henry hasn't merged it yet, some work is apparently still on going
*** Casey is going to follow up on this with Bill as he had some other items to discuss with him too
** SIMC comparisons were raised
*** Not quite there yet in our analysis, will be coming very soon (within next month or two) though
*** Garth mentioned he has a slightly different SIMC branch with different generators available
** Need to make sure updates to kaon analysis feed into the pion analysis and vice versa
* Garth showed and discussed the latest beam schedule
** Long beamtime for PionLT this year
** Final part of experiment actually expected to run next year (2022)
** Going to be a busy year or two!
----
h2. 2021.03.15 PionLT Software Meeting
* Replay and Python script running discussion
** Git changes, new branch for hallc_replay_lt/UTIL_PION/UTIL_BATCH - _offline_pionlt_
*** Need to modify repo setup script, some issues if branch doesn't exist
*** Will add instructions on how to get and update new branch
*** Will also modify so it automatically checks if the branch exists and gets it if not
** Pathing changes for c-pionlt and directory structure now all done, working well
** Some issues with replay scripts, both related to hcana updates
*** How the pcal_geom file is read in has been changed, variable needs to be VAL, VAL and not VAL VAL (comma instead of space) now
*** Helicity stuff no longer working, needed to be commented out for scripts to run
*** Carlos mentioned there's a new map for this, will need to update from hallc_replay
** Garth raised that Brad was potentially working on an improved GUI for checking online EDTM rates
*** Double check/follow up on this?
** Discussion of analysis scripts in more detail
*** How the MM is calculated, discussed adjusted values
*** How CT is calculated for each particle species
*** How RF time is utilised in PID
*** Vijay showed RF vs delta and RF vs MM plots and how these can be utilised in PID - "RFvDelta":https://redmine.jlab.org/attachments/download/1200/Kaon_RFtime_delta.pdf, "RFvMM":https://redmine.jlab.org/attachments/download/1201/Kaon_mm_FRtime.pdf, both with Kaon PID cuts applied
*** Stephen will send on some documentation on these topics to Jacob/Julie
** To do list for software
*** Update hallc_replat_lt with hallc_reply maps
*** Install new hcana in c_kaonlt area
*** Push all hallc_replay_lt fixes
*** Update git script, add instructions
----
h2. 2021.03.08 PionLT Software Meeting
* HeeP studies discussion
** Discussed minimal criteria for singles runs
*** No setting shorter than 12 minutes
*** No setting longer than one hour
*** No setting with less than 10k events
*** Want elastics rates of under 100 kHz
** Criteria/Discussion for coin runs
*** Want roughly 50k elastics
*** Multiple would be good
*** ~3 hours for one previous setting (at 50% data taking eff), this is OK
** Discussed solid angles for each spectrometer
*** HMS - 8 mSr, 8.1?
*** SHMS - 3.5 mSr, Carlos states > 4 mSr in his thesis
** Both Nathan and Jacob had some spreadsheets with settings analysed, a few numbers did not match previous settings, both looking into this (probably a solid angle off or similar)
** Garth sent on Fortran code for cross sections
* Analysis Code Updates
** Stephen has finalised the pathing/directory structure updates for
*** hallc_replay_lt
*** UTIL_PION
*** UTIL_BATCH
** All have a new branch, *offline_pionlt*
** Further testing and updates needed, Stephen will work through updating the scripts for the calibrations now that the new structure is set up
*** Some are outdated now, DC, Hodoscopes e.t.c (grab wrong DB files to begin with)
* Numerous people reported that the farm was very slow over the weekend, if this continues, should raise with Brad/SciComp
----
h2. 2021.03.01 PionLT Software Meeting
* HeeP studies
** Students still working on this, Jacob looking at rate calculations
* Stephen - updates on new git and directory setup scripts
** Scripts in place but still some work on hallc_replay_lt and UTIL_BATCH needed
*** Issues with pathing that need to be corrected for new area
*** Should have these finalised by mid/late this week
** Jacob has tested and seems to be working OK
** Will work on updating/testing DC/Calorimeter calibration scripts next
* Jacob conducting sanity checks on new matrix elements, noticed discrepancy between kaonlt/pionlt hcana and previous versions, likely due to changes in hcana from April last year
** Mark gave some talk on this last year - "Talk1":https://hallcweb.jlab.org/DocDB/0010/001060/002/update_april2020.pdf , "Talk2":https://hallcweb.jlab.org/DocDB/0010/001055/001/HCANA_Hodo_changes_april2020.pdf, "Talk3":https://hallcweb.jlab.org/DocDB/0010/001056/001/April%202020%20HCANA%20changes%20for%20other%20detectors.pdf
* Meeting time discussed, remaining at 13:00 ET (which means it shifts to 11:00 Regina time)
----
h2. 2021.02.22 PionLT Software Meeting
* Updates from the group
** Garth
*** HeeP data discussion, one per beam energy
**** KaonLT and PionLT Phase I had nearly every beam energy setting
*** Updates about the new target config, phase II and III of PionLT likely going to have new target ladder
*** Thicker aluminium cans, CAD design done but not completely finalised, will know soon
*** KaonLT beamline will be used in 2021 run, new beamline for 2022 run
*** Start of run pushed back slightly, 2022 run will be May to December, PionLT might be slotted into parts of this
** Stephen
*** Demo2 script fixed, bug was slightly stupid, run not in range but the script did not report this
**** Will look at tweaks to kaonlt.py to fix this
*** New hcana installed in pionlt area, used root 6.14.04, scons issue with newer root (6.18.04)
*** Will create setup scripts for use in migrating over to this area for working
*** Minor issue with /cache/c-pionlt directory, permissions related
* Discussed plans and potential problems for upcoming run regarding COVID procedures, how many people can be on shift, how many people will travel and so on
* Will need to shift meeting time in a few weeks due to daylight savings time change for some people
----
h2. 2021.02.15 PionLT Software Meeting
* Updates from the group
** Carlos
*** Previously looked at optimisation of kinematics for HeeP data
*** Document on "DocDB":https://hallcweb.jlab.org/DocDB/0010/001036/002/HC_SoftwareMeeting_Oct03_2019_pdf.pdf
*** Will revisit this and try to make some notes on code
** Stephen
*** No major updates, resolved some issues with scripts Jacob/Junaid were having
*** /group/c-pionlt now exists, will try to install hcana here with a more up to date version of root
*** Garth mentioned some changes to farm software, cmake, singularity, Jupyter - shouldn't affect us much
** Nathan
*** Combing hodo script DBase files, will try and update, including batch scripts
** Jacob
*** Sanity checks of running old scripts and comparing to new
*** Should be in place to combine in new matrix elements soon
** Junaid
*** demo1 being slow, file probably too large
*** Farm also being slow could be an issue
** Ali
*** Sym link hardcoding, switch location of outputting
* Idea of a directory setup script, something people can run and it will set up a tree of directories to save files to
** Can try making a version of this later this week, will need to modify output of lots of scripts!
----
h2. 2021.02.08 PionLT Software Meeting
* Updates from the group
** Stephen - Tried to replicate errors Jacob was having with Demo2, could not replicate issues and didn't see any script differences
*** May be a versioning issue, Stephen and Richard will investigate further
*** JLab environment/root version issue likely cause
** Jacob - Questions about the pionlt directories, all but the /group/c-pionlt one seem to be set up
*** Stephen following up with Brad on this
** Junaid - Some environment issues
*** Stephen will try to resolve later today
** Casey - Gave an update on his timing window scripts
*** https://github.com/mrcmor100/hallc_replay_XEM/tree/timing
*** Scripts look very nice, compare timing windows to existing ones and upload plots to network accessible location
*** Stephen will take a look at how they can be adapted for our coincidence running
* HeeP running discussion
** Minimum accessible spectrometer angles - SHMS 5.5, HMS 10.5, opening angle ??
** Want higher for HeeP running for practicality (no hall access e.t.c.)
----
h2. 2021.02.01 PionLT Software Meeting
* Updates from the group
** Stephen - Cointime peak script updates and fixes
** Jacob - Issues with Demo2 cuts
*** Stephen thinks this may be due to lack of updating of the script after some changes Richard made, will investigate
** Nathan - Hodoscope script issues, tracked down to .database file. Investigating fixes
** Junaid - Software setup on Lark but python/root issues here, working on JLab account. Will investigate root compilation on Lark
*** Stephen raised that issue in getting root to compile isn't actually related to the python version used, other package/component causing issues
** Ali - Working on improving report files and def files
* "Casey Presentation":https://redmine.jlab.org/attachments/download/1175/pionLTgroup_presentation.pptx
** Discussion of process and software used by XEM experiment
** Nice reference time script in place, will exchange with Stephen/Richard and see what can be merged with each
** Structure and plan for analysis look similar to our own, will try to collaborate as much as possible on having a nice working set of software for online usage
** Raised point of online BCM/BPM monitoring
*** Dave mentioned it would be nice to track the stability of the BCM/BPMs over the run period, at least once per shift produce a plot/document checking this
*** Stephen added that this could also track the prescale factors (reported vs calculated)
*** Investigate idea of a once per shift script that checks last X runs and produces plots above
----
h2. 2021.01.25 PionLT Preparation Meeting
* Garth Discussed run plan kinematics, purpose of Heep runs, and Luminosity scans
** Homework: students should look at distributed Heep kinematics and try to find some sets of singles (e') and coin (e'+p) runs which would overlap either in energy or angle with the physics kinematics, for the purpose of establishing angle and momentum offsets.
** Jacob also looked through Garth's physics kinematics spreadsheet, and did not find any additional errors, beyond the one GH had already identified. That error turned out to be an error in the kinematics table in the Jeopardy (2019) proposal, and GH subsequently distributed a corrected spreadsheet.
[[https://redmine.jlab.org/attachments/download/1176/pion_runplan_21jan26.ods]]
[[https://redmine.jlab.org/attachments/download/1178/elas_9p900gev.out]]
[[https://redmine.jlab.org/attachments/download/1179/elas_9p200gev.out]]
[[https://redmine.jlab.org/attachments/download/1180/elas_8p000gev.out]]
[[https://redmine.jlab.org/attachments/download/1182/elas_6p000gev.out]]
----
h2. 2021.01.12 PionLT Software Meeting
* "Stephen Presentation":https://redmine.jlab.org/documents/122
** Discussion of analysis tasks and status as of 12/01/21. The non-annotated version of the notes is also available "here":https://redmine.jlab.org/documents/123
** Homework for next week is the following -
*** Setup and try the batch calibration scripts and the python analysis script demo
*** For Stephen - Check with Brad regarding new area for files