Skip to content

Conversation

@jpn--
Copy link
Member

@jpn-- jpn-- commented Dec 11, 2025

This pull request fixes an error in Larch estimation of the CDAP model, reported in #595. There was a bug in the code, where interaction terms for 3, 4, and 5 person households adopting an all-M, all-N, or all-H patterns were all mapped to the same coefficient; the original model structure has unique parameters for each household size. [1] [2]

Testing and validation

  • Expanded the test_cdap_model test to include assertions verifying the correct assignment of interaction parameters, specifically checking the presence and absence of wildcard coefficients for different model sizes.
  • Updated the regression test data in test_cdap_model_loglike.csv to reflect changes in log-likelihood values resulting from the repaired coefficient mapping logic.

Copy link
Contributor

@dhensle dhensle left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Fix looks good. Suggest we add more detail to the change log, and one minor typo to fix.

coef_N_88,-0.05952606764773704,-0.05952606764773704,0.87709999999999999,0
coef_N_xxx,-0.89914972777069246,-0.89914972777069246,-0.36530000000000001,0
coef_N_xxxx,-0.72130413279127681,-0.72130413279127681,-1.3460000000000001,0
coef_N_xxxxx,-6.6290578146794195,-6.6290578146794195,-3.4529999999999998,0
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Happy to see the additional coef_xxx and coef_xxxx coefficients. All these coefficients across M,N, and H make a lot more sense now with a strong negative coefficient for everyone in a 5-person household doing the same thing.

@jpn-- jpn-- merged commit 0811191 into ActivitySim:main Jan 8, 2026
19 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants