Skip to content

Project Meeting 2021.02.25

Ben Stabler edited this page Mar 2, 2021 · 17 revisions

Task 1 Project Management (milestone)

  • Should we establish AMPO resources for this project:
  • Contracting / bench update
    • Partners meet to discuss bench setup and future work orders
    • Plan to get Informal RFPs (IRFPs), which are like task orders on an on-call
    • Planning to get going in a few weeks
    • Partners follow-up with Joe if you missed the call
    • IRFPs will require a 3-5 page response
  • Agenda next week
    • Clint and Joel share recent work for ARC and SEMCOG

Task 2 Complete Estimation Mode (milestone)

  • Newman working through submodel examples and coordinated with Doyle on improvements as needed
  • We plan to circle back and discuss potential additional improvements once through all the submodels

Task 3 Telecommute Model in Cooperation with SEMCOG (milestone)

  • Joel/Albab working on person occupation/industry approach based on available data
  • I'll add more depth to the design page

Task 4 Transit Pass Ownership Model in Cooperation with SEMCOG (milestone)

  • No updates
  • I'll add more depth to the design page

Task 5 Performance Enhancements (milestone)

  • Presampling pushed to develop2 for PSRC testing
  • Stefan to do some testing, including shadow pricing convergence
  • Jeff finishing up tracing and then I'll review again for the merge to develop

Task 6 Prototype multiple models test system (milestone)

  • Jeff and I further discussed and I need to update the draft plan
  • Jeff continuing to get agency examples in order for the test system improvements
  • We need to continue to better define/differentiate test vs agency example...how do they blur? how are they different?
  • Let's inventory a list of issues/ideas
  • Every code change needs test backing
  • Regarding the frequency of updating agency registered models, how close do we want to tie the PSRC owned to day-to-day ActivitySim development?
  • Small/cropped examples tested by Travis - yes, Full scale examples - no, but what's in the middle? And its important we don't miss key aspects of the system such as scaling/performance by not adequately testing "complete" examples
  • I'll review ODOT's input checker stuff from a few months ago since its relevant (including the list of checks here)

Task 7 Input checking and error handling and pipeline improvements (milestone)

  • No updates

Task 8 Maintenance and Support (milestone)

  • Accessibility chunking pushed to develop2 for SANDAG testing
  • SANDAG has successfully ran 100k HHs all the way through using SANDAG three zone system, skims, and land use!
  • Took 560 minutes single processed
  • Plan to look at high level results and test multiprocessing
  • Representative logsums will help significantly with runtimes
Clone this wiki locally