Skip to content

Project Meeting 2019.02.08

Ben Stabler edited this page Feb 8, 2019 · 18 revisions

Verification

  • We now have a complete TM1 setup and source for comparison, see the verification page
  • We re-built the asim example inputs using the same inputs
  • We ran asim with the same 10% sample of hhs above (using a new feature to specify the sample of hh ids)
  • We confirmed accessibility produces the same answers
  • It gets a lot more complicated with work/school location since it uses logsums and the random number sequence is different between TM1 and asim
  • Since it is difficult to control the random numbers, we're going to start by reviewing some aggregate summaries, similar to MTC's calibration scripts:
    • work/sch location choice - H to W/Sch flows by OD pair scatterplot for each market
    • auto own - hhs by num autos by taz
    • daily activity pattern - person type by CDAP choice
    • tour mode share by purpose, taz, OD pair, time period

Shadow pricing

  • I'll update the validation charts based on the example Joe and Guy shared - a scatterplot of estimated size to target by zone by iteration
  • Jeff will talk with Mark and Joel about the algorithm

Outreach

  • Update on various TRB outreach items
    • @Joe talking with Metro and ODOT
    • SEMCOG talked with MetCouncil and they are very interested. @SEMCOG to share MOU.
  • Update from AMPO on administrative issues
    • International payments - ok
    • MOU only needs to be signed by new partners - ok
    • Agency discretion if Executive Director or someone else can sign - ok
    • Next work order payments by April 30th, work order period July 1 to June 30 - ok
      • @Stefan to check on PSRC paying by deadline, and if not, next work order budget will be reduced and funds banked
      • SEMCOG should be able to pay by deadline
      • @Bill plan to send out invoices soon for next work order
    • @Ben draft updated DBE GFE
Clone this wiki locally