Predict with a fitted cohort-z scorer — predict_cohort_z_then_pool_score • OmicSelector

Predict with a fitted cohort-z scorer

Usage

predict_cohort_z_then_pool_score(
  fit,
  expr,
  meta,
  cohort_col = fit$cohort_col,
  block_col = "provenance_block",
  allow_new_group_moments = TRUE
)

Arguments

fit: Optional object from `fit_cohort_z_then_pool_score()`.
expr: Numeric matrix, samples x features.
meta: Data frame aligned to `expr`.
cohort_col: Column in `meta` identifying cohorts.
block_col: Column in `meta` identifying provenance blocks; accepted for a common scorer signature but not used by this method.
allow_new_group_moments: Logical; if TRUE, new cohorts are centered on their own unlabeled samples.

Value

Numeric score vector.