Skip to content

Pg simple update#1569

Merged
paciorek merged 4 commits into
develfrom
pg_simple_update
Jul 26, 2025
Merged

Pg simple update#1569
paciorek merged 4 commits into
develfrom
pg_simple_update

Conversation

@paul-vdb
Copy link
Copy Markdown
Contributor

@paul-vdb paul-vdb commented Jul 3, 2025

Three changes:

  1. Initiate inflationNodes to avoid a small error when check = FALSE.
  2. Ensure initializeX=FALSE if a design matrix is passed
  3. Add a warning if infinite values are found in the design matrix. Suggests to scale the design matrix in advance.

@paciorek Should we put a full stop if infinite values are detected in the design matrix instead of a warning?
Does it make sense to allow the user to override this warning? Currently infiniteOkay is a list input. Better name? I need to add it documentation still.

paul-vdb added 3 commits July 3, 2025 08:57
1. Fixed when Design Matrix is initialized
2. Small bug of an empty inflationNodes
3. Added a warning for when infinite design matrix values are constructed.
I forgot to actually switch this one.
@paciorek
Copy link
Copy Markdown
Contributor

Sorry about the delay in looking at this. I have some questions.

  1. Regarding the user providing inflationNodes if checking is not being done, I don't see why we need to have the user provide. Checking is just to make sure things are all valid and if the user doesn't want that done, I think we just find the inflationNodes using getParents and setDiff(inflationNodes, nonTarget). I.e., move those lines of code that initially determine inflationNodes outside of if(check). Am I missing something?
  2. Could we actually have valid computation in the sampling if there are Infs in X? I think we would get an invalid sample. So unless I'm missing something I think we should error out and not provide an override.

@paul-vdb
Copy link
Copy Markdown
Contributor Author

@paciorek Thanks. I'll make the suggested changes this afternoon and merge into devel if there are no conflicts etc.

-when inflation nodes are computed (outside of check)
-error out if an infinite design matrix value is found on the first round.
@paciorek paciorek merged commit cde9642 into devel Jul 26, 2025
8 checks passed
@paciorek paciorek deleted the pg_simple_update branch July 26, 2025 22:46
@paciorek paciorek mentioned this pull request Aug 14, 2025
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