2020 Workshop:Model systems engineering

From CedarWiki
Revision as of 20:43, 25 June 2020 by Hirsch.m (Talk | contribs)

Jump to: navigation, search

Contents

Time, Connection

Thursday, June 25, 2020

3-5pm Mountain (Denver) Time <=> 5-7pm Eastern (New York) Time

Zoom, YouTube

  1. Zoom was used for the session
  1. YouTube Live: https://youtu.be/k_c_jiI8ndY

We will post the slides and will request that the presenters allow archiving the video presentation.

Conveners

Michael Hirsch
Matthew Zettergren

Agenda

All times Mountain (Denver) Time (UTC - 6 hours)

  • 3:00 - 4:05: Applications
    • 3:00 - 3:05: brief introduction
    • 3:05 - 3:20: Resen: REproducible Software ENvironment (Pablo Reyes)
    • 3:20 - 3:35: Melding ISR data into GITM (Doga Ozturk)
    • 3:35 - 3:50: 3D Matlab visualization techniques for modeled atmospheric dynamics (Jaime Aguilar)
    • 3:50 - 4:05: Modeling Voltage Level Data for Geospace Sensors (John Swoboda)
  • 4:05 - 4:45: Tutorials (Michael Hirsch, Matthew Zettergren)
    • 4:05 - 4:15: Using Git and CI to connect, collaborate, integrate and test disparate geospace models
    • 4:15 - 4:25: Using CMake to connect multi-language and multi-repo models across computing platforms
    • 4:25 - 4:35: PyGemini examples and status update
    • 4:35 - 4:45: Managing Development of a Large Physics-Based Geospace Model
  • 4:45 - 5:00: Community discussion and next steps

Audience

  • Users, developers and managers of scientific software, for modeling and analysis of the geospace environment
  • Share the “how” of geospace model software development
  • conserve human effort – easier onboarding of new code users
  • minimize scientific software bit rot
  • preserve models and data analyses for future discoveries

Description (long form as proposed)

This session seeks to connect the power of models that may have limited public access to the broader community. This increases the rate of science discovery and value to the funding agencies. Sessions will demonstrate:

  • connecting models to modern scripting languages such as Python
  • improving build systems so it takes minutes instead of a day to do first model setup
  • calls for community involvement to help make a model more ready for public consumption

Justification ST #5: Fuse the Knowledge Base across Disciplines

  • Encourage modelers who straddle the gap between software engineering and physics to make their models more accessible by joining forces with Python veterans
  • Demonstrate ways to make models that are used in non-public applications modular enough to be publicly shared, while able to plug in non-public modules
  • Progress is measured over the next year or so by increased number of modelers sharing core functionality of their model, and enhancing model connections to the Python data scientists

ST #6: Manage, Mine, Manipulate Geoscience Data and Models

  • For many models, not having to rely on the PI or their delegate accelerates science discovery
  • Junior scientists can help develop example codes and videos for the models, greatly raising awareness, science output and value to funding agencies
  • Progress might be measured by mining papers for citations / keywords used such as links to software repos used, which can themselves be mined for use of continuous integration tools, build system type and specific software libraries

Workshop Summary

This is where the final summary workshop report will be.

Presentation Resources

Upload presentation and link to it here. Links to other resources.

Upload Files Here

  • Add links to your presentations here, including agendas, that are uploaded above. Please add bullets to separate talks.