-
Notifications
You must be signed in to change notification settings - Fork 99
Project Meeting 2022.12.15
Vivek Yadav edited this page Dec 20, 2022
·
7 revisions
- Admin
- Security Changes
- Require 2FA for all GitHub collaborators
- Move wiki to a second location
- Security Changes
- Task and Code Review Updates
-
Disaggregate Accessibilities
- Comments appear to be addressed, ready to merge?
-
ARC Parking location choice model
- PR lacks testing to prove a bug is fixed. Existing tests
likelydon't actually exercise the buggy code. - Accept and defer proper testing to Phase 8, or do we demand testing to accept this bug fix?
- PR lacks testing to prove a bug is fixed. Existing tests
-
School Escorting
- Awaiting post-process to force all escorted trips to have same mode.
-
Shadow Pricing
- RSG testing a fix to the bug identified by CS
-
Estimation Fix
- RSG review ongoing
-
Sharrow
- Most RSG review comments have been addressed.
- Awaiting review/ok from WSP
-
Skim Wrapper
- Implementing the changes in ARC model. Needs corresponding changed config file for testing.
- CS to address the changes required to make it compatible with sharrow.
-
Operational Updates
- WSP review pending.
- Flexible Number of Tour and Trip IDs
- PTV's Window Installer
- Random Seed Generator
- Fix for Pandas 1.5 compatibility
- Error preventing profiling of MWCOG model on Windows has been addressed.
- CS requested everyone who has access to the activity sim code to activate 2FA
- CS suggested moving the activity sim wiki to a new repository to limit the access to people who can edit the activity sim code. Action Item: CS will move the activity sim wiki to the new repository.
- Disaggregate Accessibilities: Accessibilities: WSP comments addressed. WSP pointed out that disaggregate accessibilities is tested with only 1 zone MTC extended model but not the 2-zone model. RSG agreed to include SANDAG placeholder 2-zone model to test disaggregate accessibilities. Action Item: RSG will add the test case for 2-zone system.
- ARC Parking location choice model: CS pointed out that the test wasn’t actually testing the parking location choice fix. They suggested scripting the test as part of phase 8. WSP confirmed that the fix is working when they tested the model on their end with larger sample size. Consortium agreed to accept the pull request and leave the unit testing for the future but doesn’t have to address it instantly.
- School Escorting: RSG is still working on it.
- Shadow Pricing: RSG fixed the bug. CS review ongoing.
- Estimation Fix: RSG review ongoing
- Sharrow: Most RSG review comments have been addressed. Awaiting review/ok from WSP
- Skim Wrapper: CS review ongoing.
- Operational Updates: WSP review ongoing.
- WSP is able to run MWCOG model
- WSP did two runs – 5% and 25% household sample rate
- Both runs have very similar memory patterns throughout the runtime.
- The string data types in the choosers table are also using a lot of memory (mandatory tour scheduling)
- WSP will provide the memo by end of the calendar year.