Hannah Frick

We've released support for survival analysis in tidymodels. This methodology is useful for all kinds of time-to-event data: That event may be a customer churning, a machine needing repairs or replacement, a shelter animal being adopted, etc.

For a new case study on tidymodels.org, that event is a complaint to the NYC Departments of Buildings being dispositioned. The analysis shows how to use tidymodels for censored regression, covering the whole process. #RStats

https://www.tidymodels.org/learn/statistics/survival-case-study/

tidymodels - How long until building complaints are dispositioned? A survival analysis case study

Learn how to use tidymodels for survival analysis.

Apr 8, 2024 at 1:43pmWeb
Show threadbroccoliccoliApr 9, 2024
@hfrick I am a weirdo who grew into the tram package framework of regression, so I have to ask: How much has this been built with the assumption that censoring=time-to-event? Is clowning around with stuff like censored ordinal responses supported?
Show threadHannah FrickApr 9, 2024
@broccoliccoli for the initial round of development we’ve focused on what we perceive as the most common use cases, so indeed along your censoring = time-to-event line. We are happy to hear further suggestions and feature requests though. You are welcome to open an issue for this! Best place would be the repository for the censored package