Format ov names in formatted objects #122
Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 755f0600f5
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| keepRn <- rep(TRUE, length(rn)) | ||
| keepRn[isTempRn][isDupTempRn] <- FALSE | ||
| mat <- mat[keepRn, , drop = FALSE] |
There was a problem hiding this comment.
In R, chained indexing like keepRn[isTempRn][isDupTempRn] <- FALSE updates a temporary vector, so keepRn is unchanged. When both .TEMP_OV__x and x rows exist, the temp rows are not dropped, and rnClean[keepRn] then collapses both to x, yielding duplicated row names/cols. That can lead to ambiguous indexing or downstream errors when consumers expect unique names (e.g., modsem_inspect/modsem_predict output). You’ll want to index keepRn directly (e.g., keepRn[which(isTempRn)[isDupTempRn]] <- FALSE), and the same pattern appears in the column and vector cases.
Useful? React with 👍 / 👎.
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 0fa75f8b14
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
R/inspect_da.R
Outdated
| partable = modsemParTable(model$parTable), | ||
| coefficients.all = .modsemVector(model$coefs.all), | ||
| coefficients.free = .modsemVector(model$coefs.free), | ||
| partable = parameter_estimates(model), |
There was a problem hiding this comment.
modsem_inspect_da() now builds partable via parameter_estimates(model) without forwarding is.public, but parameter_estimates.modsem_da defaults is.public = TRUE (R/generics_modsem_da.R:10–27), which formats names and drops temp OV rows. As a result, modsem_inspect(..., is.public = FALSE) returns an internal set of matrices/vectors alongside a public partable, so name-based joins (e.g., matching partable$lhs/rhs to lambda/theta from the same inspect call) can fail due to mismatched identifiers. Passing is.public = is.public here would keep the payload consistent.
Useful? React with 👍 / 👎.
No description provided.