Fit the ECOD + COPOD novelty discriminator

Fits the bundled ECOD + COPOD novelty discriminator. The training matrix is mapped to the per-sample robust CLR (over each row's own positive support), and the in-distribution reference is taken to be the training CONTROL rows (y_train == 0). pyod's ECOD and COPOD are fitted on that control rCLR (leakage-safe: training only) and used to (a) source / verify the per-feature skewness signs and (b) provide the decision_scores_ anchor that the frozen pure-R reimplementation must reproduce. The frozen model stores only R-side state – the per-feature control reference columns, the skew signs, the per-detector control-aggregate mean / sd, and the orientation scalar – so scoring is pure R and never calls Python.

Usage

fit_ecod_copod(X_train, y_train, meta_train = NULL, hp = list())

Arguments

X_train

Numeric matrix (samples \(\times\) features) of non-negative abundances with unique, non-empty feature names (colnames).

y_train

Numeric / integer 0/1 labels (1 = case), length nrow(X_train), with at least one case and one control.

meta_train

Optional per-sample metadata. Accepted for interface uniformity and otherwise ignored.

hp

Optional list of hyperparameters. Allowed fields: min_features (positive integer feature-overlap floor at scoring; default 3L), eps (positive finite floor for the standardization sd and a numeric guard; default 1e-8), and verify_pyod (single logical; when TRUE (default) the frozen pure-R reimplementation is checked against pyod's decision_scores_ on the control rows at fit and the fit fails if the maximum difference exceeds 1e-6).

Value

Object of class ecod_copod_model: a list with feature_universe, ref_cols (frozen per-feature control reference columns), skew_sign, score_mean / score_sd (per-detector control-aggregate standardization constants), orientation, n_control, and hp.

References

Li Z, Zhao Y, Hu X, Botta N, Ionescu C, Chen G. (2022) ECOD. IEEE TKDE.

Li Z, Zhao Y, Botta N, Ionescu C, Hu X. (2020) COPOD. ICDM.

Examples

if (FALSE) { # \dontrun{
set.seed(1)
n <- 120; p <- 30; k <- 8
L <- matrix(stats::rnorm(n * p, 4, 0.5), nrow = n,
            dimnames = list(NULL, paste0("miR-", seq_len(p))))
y <- rep(c(0, 1), each = n / 2)
L[y == 1, seq_len(k)] <- L[y == 1, seq_len(k)] + 1.2
X <- exp(L)
model <- fit_ecod_copod(X, y)
score_ecod_copod(model, X[1, , drop = FALSE])
} # }