Update ETA shrinkage with subject-level filtering — update

Recomputes overall ETA shrinkage from stored subject-by-ETA data, applying threshold-based filtering to remove subjects whose individual shrinkage is too close to 1 (variance form). Returns an updated xpdb with the etashk summary row replaced.

Usage

update_etaShrinkageNlme(
  xpdb,
  threshold = 1e-06,
  eta_threshold = NULL,
  eta_name = NULL,
  .problem = 1
)

Arguments

xpdb: An xpose_data object created by xposeNlme or xposeNlmeModel.
threshold: Global default threshold applied to all ETAs (default 1e-6). A subject/ETA row is removed when its variance-form shrinkage $\ge (1 - \text{threshold})^2$.
eta_threshold: Optional named numeric vector overriding threshold for specific ETAs, e.g. c(nV = 1e-6, nCl = 0.05).
eta_name: Optional character vector selecting a subset of ETAs to recompute. ETAs not listed keep their current summary values.
.problem: Problem number (default 1).

Value

A new xpdb with the etashk value in xpdb$summary updated.