Fix CI #23

Activity

1. m-muecke commented on Mar 19, 2025

   

   m-muecke

   on Mar 19, 2025

   Contributor

   More actions

   Currently an old setup for Travis CI is used and hence automatic checks are broken.

   You open for using GitHub Actions? Then I could provide a PR.
2. davidruegamer commented on Mar 19, 2025

   

   davidruegamer

   on Mar 19, 2025

   MemberAuthor

   More actions

   Yes and thank you! That would be awesome!
3. m-muecke commented on Mar 20, 2025

   

   m-muecke

   on Mar 20, 2025

   Contributor

   More actions

   Yes and thank you! That would be awesome!

   Ok, great! You open for using testthat for testing? It's currently already in suggest, but afaik not really used.
4. davidruegamer commented on Mar 20, 2025

   

   davidruegamer

   on Mar 20, 2025

   MemberAuthor

   More actions

   IIRC, we didn't use testthat because our tests take a long time. I would have to think about how to simplify them to avoid long CI waiting times.
5. m-muecke commented on Mar 20, 2025

   

   m-muecke

   on Mar 20, 2025

   Contributor

   More actions

   IIRC, we didn't use testthat because our tests take a long time. I would have to think about how to simplify them to avoid long CI waiting times.

   What I generally do, when the tests are running too long for the CI is to add a skip call, e.g. skip_on_ci() and skip_on_cran()
6. davidruegamer commented on Mar 20, 2025

   

   davidruegamer

   on Mar 20, 2025

   MemberAuthor

   More actions

   Good idea. Another option would be to skip all the factorization tests from @Almond-S and let the boosting routines in general_tests only run for one iteration via control = boost_control(mstop = 1). Since we do not do any type checking in this file, I guess we could go with tests that simply check for no error