Project Meeting 2016.12.16

#Task 1 Project Management

• @Elizabeth - discussed review and draft memo
  • She identified a number of areas of improvement:
    • a unified approach to data access, data I/O functions, the ability to restart model runs, output data, etc
    • simplified abstraction levels for different users - core developers, modelers with good programming skills, users
    • improved example and documentation
    • catching errors early and often, explicit error catching, improved error legibility, and a debug mode
  • There was much discussion about orca, both good and bad, especially legibility versus flexibility
  • We also discussed improvements to the data model / data access capabilities and the need for a clear set of wrapper functions around pandas, numpy, and orca for data I/O and data pipelining
  • @Everyone - comments due by end of year so @Elizabeth - can finalize the memo shortly after
• Next year's contract is still waiting on signatures from PSRC and SFCTA
  • Will hopefully be signed soon
  • @Ben to draft Phase 3 scope of work on wiki for discussion at TRB
    • Topics for the scope of work include:
      • logsums
      • intermediate stops
      • shadow pricing
      • threading
      • data access / pipelining
      • continued finishing up of phase 1 models
• SANDAG's supplemental contract has been reviewed by contracts and should be executed in early/mid January
• Contract ends Dec 31st 2016, so @Bill will extend a month or two so we can make sure to use up the hours we have
• @Bill to setup a meeting at TRB to go over the scope of work; the 8pm Sunday meetup will remain on the calendar too

#Task 10 Full Model Run

• Updated code due to update in OMX library #131
• @Ben and Jeff to spend the remaining contract hours on drafting next year's scope and documentation and usability improvements suggested by Elizabeth