GlueX Offline Meeting, November 15, 2017
GlueX Offline Software Meeting
Wednesday, November 15, 2017
11:00 am EST
JLab: CEBAF Center F326/327
BlueJeans: 968 592 007
- 1 Agenda
- 2 Communication Information
- 3 Minutes
- Review of minutes from the last meeting (all)
- Organizing AmpTools-Based Event Generators (Justin)
- Review of recent pull requests (all)
- Review of recent discussion on the GlueX Software Help List (all)
- Action Item Review (all)
- The BlueJeans meeting number is 968 592 007 .
- Join the Meeting via BlueJeans
Talks can be deposited in the directory
/group/halld/www/halldweb/html/talks/2017 on the JLab CUE. This directory is accessible from the web at https://halldweb.jlab.org/talks/2017/ .
- CMU: : Curtis Meyer
- FSU : Brad Cannon, Sean Dobbs
- JLab: : Alex Austregesilo, Alex Barnes, Thomas Britton, Mark Ito (chair), Elton Smith
- MIT : Cris Fanelli
- W&M : Justin Stevens
- Yerevan : Hrach Marukyan
There is a recording of this meeting on the BlueJeans site. Use your JLab credentials to access it.
New releases, AmpTools 0.9.3, sim-recon 2.19.0. These came out November 2. See the release notes for details.
Review of minutes from the last meeting
We went over the minutes from November 1.
- MCwrapper. Thomas reported on new features.
- He has been working on supporting output samples that are run-by-run proportional to the size of real runs within a user specified range.
- With Sebastian Cole and Sean he has been working on a feature where the BGRATE parameter to HDGeant(4) is generated automatically depending on the beam intensity of the corresponding real run. This will be implemented as a stand-alone C++ program using Richard's coherent bremsstrahlung library.
- HDvis. Some items from Thomas.
- Changes to recent changes to sim-recon broke the EVIO to JSON conversion. Thomas has fixed this.
- The GCC compatibility problem reported last time was solved by using the regex function from the boost libraries rather than getting them from std::.
- There is currently a problem where events pile up in the display rather than being rendered one by one. Cool visual effect, but not very useful.
- HDvis will not run on all nodes in the counting house. It depends on the particular video driver being used. Diagnosed by running glxgears.
- Work Disk Clean-Up. Mark reported that the initial rsync to the new work server is almost done after almost two weeks. We will make the decision on when to cut over depending on circumstances once the rsync is done.
Organizing AmpTools-Based Event Generators
Justin presented slides outlining a proposal for streamlining development and maintenance of our event generators, resident in the sim-recon tree, that depend on the AmpTools package. Please see his slides for details. Some summary points:
- New generators of this type are created by cutting and pasting code from existing generators. This results not only in duplication of code, but duplication of constants, such as particle masses and beam characteristics.
- There are sections of the code that could be isolated and coded in a more general way with behavior controlled by input parameters and constants obtained from known, trusted sources. This would make the code more concise and less prone to errors. Among these sections are:
- beam energy distribution
- beam polarization
- SDMEs used in amplitude generation
- decay chain specification
- properties of common resonances
- algorithms for generating according to phase space
- methods for creating the HDDM output
- event-by-event random number seed insertion (Alex A.)
- There was general agreement that we should go ahead with the plan.
- Sean suggested a document on the wiki putting words to the proposal outlined in the slides, i. e., a conceptual design. Justin agreed to write it.
- Alex A. volunteered to work on the decay chain specification, resonance properties, and phase space generation.
Review of recent pull requests
We went through the list. Some discussion was generated.
- Alex A's event-by-event random number specification in event generators has already been implemented (Pull Request #972).
- Alex A.'s code to compute moments from an AmpTools fit (Pull Request #958) requires the GNU Scientific Library. This is already available on all nodes at JLab, but it has to be added to the list of packages that must be installed since it is not always part of a minimal Linux distribution.
- Thomas will incorporate Sean's J/ψ generator into MCwrapper. The generator came into sim-recon with Pull Request #952
Review of recent discussion on the GlueX Software Help List
We looked over the list, without much comment.
gluex_install on Ubuntu
Mark mentioned that the complete build of the GlueX software stack on Ubuntu is broken at present. The Ubuntu-specific parts are behind the times; they did not get a refresh after HDGeant4 was introduced. In particular, some packages required by Geant4/HDGeant4 are not on the old list. Mark is working on it. There should be an announcement soon.