Project Meeting 2018.05.11

Stop models

• Stop frequency and trip purpose coding done
• Timing for full run on SD modeling server:
  • stop_frequency : 531.37 seconds (8.9 minutes)
  • trip_purpose : 99.178 seconds (1.7 minutes)
• Results look reasonable, but will check against MTC calibration/validation report
• Starting on trip destination choice
  • The three steps have to be combined into one and chunked as a trio, since we need to compute the trip dests in trip_num order, updating the orig of next trip based on the dest of the prior trip

Documentation / release

• I still need to update the documentation and do a release with joint models + stop models implemented to date

Project admin

• We've spent 2/3rd of the current amendment budget so about $50k remains
• We'll likely run out of funds in late June/early July
• I suggest we start drafting the next amendment
  • Here is the draft phase 4 amendment, which is around $200k of work
  • I think we're ready to tackle multi-processing now that the core model components are implemented. Multi-processing is the final significant task and everyone is interested in it.
• How much budget can we plan for starting July 1st?