Child pages
  • 10th December 2014
Skip to end of metadata
Go to start of metadata

Meeting notes for Sims telecon, December 3rd 2014:

1-866-330-1200
5182287#

https://www4.gotomeeting.com/join/821177359

Attending: ajc, scott, michael, george, darko, veljko, brittany, cathy, lynne, kem, simon, peter, steve R

Note: sims people will try to meet the friday after the AAS meeting ( scott, lynne, kem, cathy, peter, simon)

Action Everyone will look at their calendars to define the the next in person meeting (March 2015)

Action For next quarter of work we will identify likely dependencies on other simulations efforts

 

Current status

All:

  • Write up the simulation framework codes into papers: papers will be based off the SPIE versions (ie expanded). This would provide a set of refereed papers for opsim, catsim, maf that people can cite.
    • We need to schedule time for this in the next quarter

Peter:

  • Sky Model 
    • Progress made with formatting and processing data
      • data in the right format, and discussed the 
      • Want to define brightness as fn of sun angle, moon elevation etc
      • will initially ignore unresolved stars
      • will continue in the next quarter
    • Evaluate ESO model
      • DONE - no source code is available  so we can't deploy it locally
      • will write up the results of the comparison
  • Calibration sims (as needed)
    • none requested but code refactored
      • ajc to ask Z and RHL about needs
  • Community support of MAF
    • Lynne supporting this
    •  in the process of adding focal plane maps to MAF
    • work with the community to help them write metrics and to curate contributed metrics … we’ve done this, set up a github repo, and been working with the members of the community (pretty much Phil Marshall, Mike Lund and Ashish Mahabel with Simon) to resolve problems they may have had with MAF and to help write new metrics. With the limited set available, we could easily run these metrics on new opsim output together with our cadence metrics.
  • Dust maps
    • decided not to upgrade beyond implementing the map available from sims in heal pix
    • DONE

 

Lynne:

  • Solar system
    • Implement solar system metric
    • will complete in next quarter
  • Visualization
    • Freeze the current web interface and then give to Cathy to write more opsim development metrics and plots
      • Conversion transition to MAF
      • We cleaned up a lot, made a new visualization page in MAF outputs, identified and killed some bugs, made a bunch of enhancements to the viz layer, added six-panel plots in PDF and a way to easily copy MAF results for distribution, and now have one or two small issues (relating to the configuration summary and slew statistics) and then I think/hope we’ve converged. Peter has summarized the progress in a confluence page, https://confluence.lsstcorp.org/display/SIM/Replicating+SSTAR+with+MAF%2C+Results
      • Goal is close out this week
      • Kem, remaining config summary stuff not essential for transition of MAF to Cathy. We still have to deal with slew stats but these should be accessible from summary_report.py (Peter - it is not clear what these numbers mean in the summary_report)
    • Mozilla work on iterative viz
      • not gone very far should look at this again in the next quarter
    • Nan-Chen work on collaborative viz
      • design set up and ready for review next quarter


Scott:

  • Complete catsim clean up
    • mostly complete - want to refactor photutils 
  • Implement variability model
    • SNe model in progress
    • for next quarter should look at adding  Steve R's model for variability
  • galsim
    • close to completion
    • ChrisW - tmv could be packaged as part of the install (and should fix the dependencies with python and boost) so we can install galsim using eups (but with mininimal dependence on the LSST stack) 
  • Implement footprint model in catsim then MAF
    • DONE - slow (need to speed up) - 45 mins to 45 hrs!
    • simon - no effort on speed for DM so can address this in the next quarter


Bryce

  • Finish bulge star density evaluation and write up (internal report then paper)
  • Finish PCA on stars/galaxies and implement for cats and phosim (write up as a paper)

 

Cathy

  • Finalize what functionality needs to be in MAF to replace SSTAR
    • DONE
  • Learn to write metrics and implement any required metrics for opsim in MAF
  • Implement new plots in MAF required for evaluation of Tier 1 runs
    • Next quarter implement Zeljko's summary plots (extended SRD requirements, including efficiency, open shutter time etc) 


Francisco

      Just back Thursday from wedding.

  • Generate a breakdown of tasks for the refactoring of opsim to produce telemetry data and to modularize the scheduler component
    • DONE
  • Produce a design for the OCS/Opsim framework and submit for design review (including libraries, data structures, api and methodology for the communication)
    • next quarter
  • Start the refactoring

 

Michael (the person formally known as New Hire)

  • Learn the opsim code base and LSST tools
    • Implement eups packaging 
      • done and in testing
    • configuration refactoring
      • next quarter  
    • wants discussion on naming conventions for proposal types on opsim 
  • Work with Francisco and support the OCS/Scheduler development plan

 

Kem

  • LSE-190, Scheduler Requirements.
    • need to talk to francisco about some of the details
  • Evaluate Tier 1 and rerun if necessary
    • started rerun with tier 1 with v3.2 on ops
  • Scheduler experts workshop
    • ajc needs to do this proposal to George and will finish by next phone con

Veljko:

  • Event Broker
    • Generate small number of diasources based on variability (nominal catalog vs time specific catalog) - base off variabilityReferenceCatalog.py
      • DONE
    • Create initial voevent package and broadcast locally for the brightest sources (small numbers)
      • DONE
    • Need for design review and output
      • next quarter
    • UCDs for the parameters in the schema
    • Visualize output (e.g. distribution on sky) using existing voevent tools
    • Modify the output parameters for the catalogs to reflect the object schema (or diasource schema) that DM expects

 

 

  • No labels