Skip to main content

OAS

Oracle Approximating Shrinkage covariance — a shrinkage estimator whose intensity targets the oracle under Gaussian assumptions, often improving on Ledoit-Wolf for Gaussian data.

Algorithm

Uses the OAS closed-form shrinkage coefficient (the scikit-learn-corrected formula) to shrink the sample covariance toward a scaled identity.

Constructor

Skigen::OAS<Scalar> model(bool assume_centered = false);

Parameters

ParameterDefaultDescription
assume_centeredfalseSkip mean subtraction.

Methods

MethodDescription
fit(X)Estimate the shrunk covariance.
score(X)Gaussian log-likelihood.

Fitted Attributes

AccessorDescription
covariance()Shrunk covariance.
shrinkage()OAS shrinkage intensity.

Example

Skigen::OAS<double> oas;
oas.fit(X);
auto C = oas.covariance();
Verified against scikit-learn

This estimator is checked by the parity suite. See the generator tests/parity/generate_covariance_reference.py and the reference fixtures in tests/parity/data/oas/, exercised by tests/parity/parity_covariance.cpp.

API Reference

For full signatures see the OAS API Reference.