-
Notifications
You must be signed in to change notification settings - Fork 11
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Scoping problem with "combined_risk" in nc_mboostLSS #49
Comments
@ja-thomas can you please look at this? |
Problem is here: Line 286 in a462d46
I'm not sure why I had to write this (horrible) line... I need to check |
Hmmmm, I played around with this some more. I kind of remember why I did it. Since assignment for the This is pretty bad design from my part, but I didn't find a solution at the time and unfortunately I'm still not sure how to find the correct environment to write the tldr.: This sucks, but I have no idea how to fix it :( |
Thanks @ja-thomas for digging into this. Without having had another look I think the issue is that some functions are defined outside of the relevant environment and hence do not share the same parent environments. Thus, the Of note, I do not think that gamboostLSS writes to the global environment in other places. We define global variables but only ro get rid of NOTES in R CMD check. Actually, these variables should be defined at the time they are used inside another environment than the global environment. @Almond-S I am not sure if we should really close this issue. It still isn't solved. Of course it has not the highest priority but it should stay on the list of open issues. Perhaps I'll have some spare time on a train ride can look into this (or someone else has a good idea how to fix this). |
With @ja-thomas hints that was easy to fix. |
Thank you both so much! I had all my simulation studies showing apparently corrupt crossvalidation results, which did not fit to the results with prior gamboostLSS versions, and think that this should have been the problem. And this was kind of the last issue prohibiting me from finishing a project on FDboostLSS for functional responses. So many thanks again!! |
Thanks @hofnerb you're a genius :) |
You are more than welcome. Thanks for locating the reason in the first place! @Almond-S Is it currently sufficient for you to have the bug fixed on github or should we make a quick release to CRAN as well? |
I think, having it fixed on gitHub should be completely fine. But if it won't, for whatever reason, I'd get back to you. Thanks again! |
attr(model1, "combined_risk")
seems to always relate to the last fitted model instead ofmodel1
.And perhaps related to this issue, the combined_risk cannot be returned when fitting multiple models with
mclapply
:The text was updated successfully, but these errors were encountered: