Fit a Group-DRO kit x biofluid logistic scorer.

Usage

fit_group_dro_scorer(
  expr_matrix,
  y,
  sample_meta,
  kit_col = "kit_family",
  biofluid_col = "biofluid",
  panel_features = NULL,
  panel_size = 20L,
  eta_grid = c(0.01, 0.05, 0.1),
  l2_grid = c(0.001, 0.01, 0.1),
  inner_folds = 3L,
  learning_rate = 0.05,
  epochs = 200L,
  tune = TRUE,
  seed = 42L,
  ...,
  group_train = NULL
)

Arguments

expr_matrix: Numeric training matrix, samples x miRNA features.
y: Binary 0/1 training labels.
sample_meta: Metadata aligned to rows of `expr_matrix`.
kit_col, biofluid_col: Metadata columns used for train-only group construction. Groups are `kit x biofluid`.
panel_features: Optional training-selected feature panel. If omitted, a train-only CLR panel is selected inside this function.
panel_size: Number of features selected when `panel_features` is NULL.
eta_grid, l2_grid: Hyperparameter grids for train-only nested CV.
inner_folds: Number of inner folds for hyperparameter tuning.
learning_rate: Gradient-descent learning rate.
epochs: Number of optimization epochs.
tune: Logical. If TRUE, tune `eta` and `l2` on training data.
seed: Integer random seed.
...: Reserved for scorer-interface compatibility.
group_train: Optional biological-group identifiers, one per training row, without missing values. Whole groups stay together in inner resampling. `NULL` preserves the original row-based fit. Invalid or infeasible groups raise `singlesample_input_error` with code `INVALID_GROUPS`.

Value

A frozen Group-DRO fit object consumed by `score_group_dro_scorer()`.