Child pages
  • 13th December Simulations Meeting
Skip to end of metadata
Go to start of metadata

13th December Simulations Meeting

Attending: ajc, john, dave monet, lynne, peter, kem, zeljko, kirk, jim, simon, abi, debbie, cathy, michael wv

 

Actions:

  1. ajc: to put image renaming code into simulation stack
  2. ajc: to invite Alex Abate to present at a sims phonecon on her work
  3. everyone: read Lynne's design review pages
  4. ajc: will send requests for people to read opsim review documents (feedback due 1st week of January)

After DESC, FDR and all of the associated preparations, meeting focuses on updates and work plans for the next few weeks of work 

Change in simulations development: simulations will be treated as a single team to remove some of the silo effect (with resources moved around as necessary). Everyone should expect to be reviewing code and designs from other members of the group.

  1. Debrief from the FDR (ajc)
    1. We prepared a lot of talks but only got to present the overview and calsim (a few questions). Tools and simulations were noted at report out as a highlight of systems engineering. Much work ahead but we should take a moment to acknowledge all of the work that has gone in to get us to FDR and how sims were integral to the success of this project (congratulations are due to all of the team). John - simulations were described as dessert by the committee. Kem - as they weren't directly related to systems engineering in the FDR.
       
  2. Debrief from DESC meeting (Debbie)
    1. Very productive. Power user session for phosim was very well attended and people working actively with phosim (including sensors, atmosphere, science). Having these events more often would be useful. Starting recipes at github (example scripts in python) for using phosim (see comments). Plan is for recipes for phosim, catsim, opsim and dm stack. Developing in 6 month timescale. Welcome contributions (for example the configuration parameters - documenting them in detail). Spoke with Chris Walter about DM and his recipes.

      Simulation mock catalogs got a whole day. Cosmo sim people want to use catsim as a framework for their validation studies. People keen on integrating into the validation framework. Simon - Risa said we should share validation data. Some of it is not public so it is a little tricky but should be addressed. Hands on walk through was useful. No actions but need to be responsive to science groups

      Kem - Steve Ridgway said no a enormous interest in opsim but supernova working group interested in rolling cadence. Lensing group (time delay) wanted better sampling on long time scales. Simon - Eric Gawiser gave dithering talk and is keen to search the parameter space on how dithering works. Expressed concerns about time it takes for opsim to run as wanted to put in any arbitrary dithering schemes (not clear if he really needs to run opsim). MWV - DB people at Pitt interested in helping with opsim in context of DB access and optimizations.

      Simon - made the worlds most boring movie (see comments). Has the first end-to-end run from catalogs to DB to images to DM stack. Should make the process into that into a set of scripts. Renaming code should be put into version code. MWR - files should be self documenting.

      Kirk - presentation on direct simulation of galaxy properties (for photoz). Alex Abate gave a talk on these simulations. Action for ajc to have Alex give a presentation

  3. Project requirements for simulations looking out over the next 3 months
    1. SPIE papers
      1 general paper and 3 on opsim and OCS (see comments for details of abstracts)
    2. photoz requirements for out of band leaks for the filters
         - kirk's slides on impact of leaks
         - Kirk - design studies dont factor in the tolerance in non-uniformity of the surface.
    3. M1/M3 acceptance
      • Need to plan out the testing plan
    4. new project software and plans for rolling that out (e.g. Stash, upgrade of confluence, Jira Agile)
    5. Any issues with current tools
      • John - need to be able to make the pages public
      • Kem - what about greenhopper (for project management) Simon - greenhopper = JIRA agile.


    Postponed for next week:

  4. Updates from opSim of review planning and June workshop
  5. Updates from catalogs 
  6. Updates from phosim 

  • No labels

4 Comments

  1. Unknown User (djbard)

    From the DESC meeting: we (some of us at SLAC) have started working on a 'recipe book' for phosim, to create user-generated worked examples. It should be a good complement to the documentation provided by phosim. Over the next months, this will be extended to catsim recipes, and in the longer term (~6 months?) DM and opsim recipes. The recipe should be in the form of well-documented python scripts (we're using sphinx to generate web-based documentation). Eventually, we want to evolve a library of functions that users can piece together to make a custom phosim run.

    https://github.com/DarkEnergyScienceCollaboration/ImageSimulationRecipes

  2. Please give comments on the MAF design review documentation. Is this understandable? Is it sufficient for the design review on the building blocks? Is there anything else we need to cover? 

    Metric Analysis Framework, Design Review