Conditional Logistic Regression
Conditional Logistic Regression
I don't see how to do conditional logistic regression in the TidyModels ecosystem. It is the workhorse for case/control studies in epidemiology. Google scholar finds ~75,000 articles using it. To learn more take a look at Agresti's Intro to Categorical Analysis chapter 8.2.4 Conditional Logistic Regression for Matched-Pairs*. I coded the book in R ... that chapter is here.
Somebody requested it six months back on SO and I have a pile of epidemiologist learning TM saying they need it.
Many/most people would use survival::clogit() to do conditional logistic.
We can make a new engine for logistic_reg() and put that in the censored package.
Would you like to make a PR for that?
There is info on how to build a parsnip model, and how to add an engine to an existing parsnip model, in this article.
As we haven't seen any further interest in the last few years, I'm going to go ahead and close. This does sound like a great idea for a parsnip extension package--feel free to holler in a new issue if you'd like to implement this in a new package and would appreciate support.
This issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.