Predict a mixed-effects panel score without reading test labels

Predict a mixed-effects panel score without reading test labels

Usage

predict_mixed_effects_scorer(
  fit,
  expr,
  meta,
  cohort_col = fit$cohort_col,
  block_col = fit$block_col,
  min_weighted_features = 3L
)

Arguments

fit

Object from `fit_mixed_effects_scorer()`.

expr

Numeric matrix, test samples x features.

meta

Data frame aligned to `expr`; labels are not used.

cohort_col

Test cohort column.

block_col

Test provenance-block column.

min_weighted_features

Minimum finite weights required for scoring.

Value

Numeric score vector, one value per sample.