model_parameters() unrolls all pairwise correlations for AR(1) structures#1213
model_parameters() unrolls all pairwise correlations for AR(1) structures#1213strengejacke merged 16 commits intomainfrom
Conversation
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request updates the package version to 0.28.3.15 and introduces logic to handle special covariance structures, such as AR1 and compound symmetry, within the random variance extraction process. It also includes significant code reformatting for better readability and style consistency. Feedback highlights a critical indexing bug in the covariance structure loop and redundant logic in a switch statement.
| toep = , | ||
| cs = |
There was a problem hiding this comment.
These lines in the switch statement are redundant or incomplete. The cs case is already handled above (lines 222-225). Additionally, an empty case like cs = at the end of a switch can lead to errors or unexpected NULL returns in R. If toep was intended to fall through to the cs logic, it should be placed immediately before the first cs block. If not, these lines should be removed.
There was a problem hiding this comment.
Pull request overview
This PR addresses #1202 by revising how model_parameters() (via random-effects variance extraction) reports structured covariance parameters from glmmTMB models, aiming to avoid expanding AR(1)/OU/etc. structures into all implied pairwise correlations and instead returning the actual estimated structure parameters. It also updates documentation/examples gating and adds regression tests for several covariance structures.
Changes:
- Collapses structured glmmTMB covariance-structure outputs (e.g., AR1/OU/CS/Toeplitz) to more concise parameter rows aligned with
VarCorr(). - Adds a new test suite covering AR1, OU, CS, and Toeplitz covariance structures for glmmTMB random effects.
- Minor documentation/vignette maintenance (examplesIf checks via
insight::check_if_installed(), vignette theme initialization, NEWS + version bump).
Reviewed changes
Copilot reviewed 10 out of 10 changed files in this pull request and generated 7 comments.
Show a summary per file
| File | Description |
|---|---|
R/extract_random_variances.R |
Adds logic to detect glmmTMB covariance structures and reduce redundant random-effect rows/labels accordingly. |
tests/testthat/test-random_effects_covstruct.R |
New tests validating the revised random-effects output for several glmmTMB covariance structures. |
vignettes/demean.Rmd |
Only sets ggplot theme when required packages are available. |
R/4_standard_error.R |
Updates @examplesIf condition and applies formatting-only refactors. |
R/1_model_parameters.R |
Updates @examplesIf condition for examples. |
man/standard_error.Rd |
Syncs rendered examples condition with roxygen change. |
man/model_parameters.default.Rd |
Syncs rendered examples condition with roxygen change. |
man/model_parameters.brmsfit.Rd |
Documentation update expanding diagnostic option descriptions (ESS tail/bulk). |
NEWS.md |
Notes glmmTMB cov-structure output revision and BSDA htest support. |
DESCRIPTION |
Bumps package version. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| re_cor_slope[to_select] <- .update_indicator_rows(re_cor_slope[to_select]) | ||
| }, | ||
| toep = { | ||
| re_cor_slope[to_select][duplicated(re_data$var2[to_select])] <- FALSE |
| new_label <- paste(new_label, seq_along(which(re_cor_slope[to_select]))) | ||
| } | ||
| ) | ||
| # fix label for cov_structure correlation | ||
| re_data[which(re_cor_slope[to_select]), "var1"] <- new_label | ||
| re_data[which(re_cor_slope[to_select]), "var2"] <- "" |
| # keep all random effects parameters, because some are redundant | ||
| .update_indicator_rows <- function(indicator_rows) { | ||
| if (any(indicator_rows)) { | ||
| # we have multiple rows to select for unstructed ("us") covariance structure |
| out <- model_parameters(m_toep, effects = "random", verbose = FALSE) | ||
| expect_identical( | ||
| out$Parameter, | ||
| c( | ||
| "SD (time1)", | ||
| "SD (time2)", | ||
| "SD (time3)", | ||
| "SD (time4)", | ||
| "Cor (Toeplitz Lag 1)", | ||
| "Cor (Toeplitz Lag 2)", | ||
| "Cor (Toeplitz Lag 3)", | ||
| "SD (Observations)" | ||
| ) | ||
| ) |
| # Validated against output from VarCorr() | ||
| expect_equal( | ||
| out$Coefficient, | ||
| c(0, 1.0112, 0.9844, 1.0525, 1.0346, 0.3893, 0.2383, 0.1979, 0.0041), | ||
| tolerance = 1e-3 |
| "SD (time4)", | ||
| "Cor (Toeplitz Lag 1)", | ||
| "Cor (Toeplitz Lag 2)", | ||
| "Cor (Toeplitz Lag 3)", |
| # Validated against output from VarCorr() | ||
| expect_equal( | ||
| out$Coefficient, | ||
| c(0.131, 0.5664, 0.5578, 0.5678, 0.6301, 0.8596, -1, 0.0017, -0.0104, -0.0096, 1e-04), | ||
| tolerance = 1e-3 |
Fixes #1202