posthoc.cov_estimators.ShrinkageKernel¶

class posthoc.cov_estimators.ShrinkageKernel(alpha=0.5)¶

Shrinkage estimation of the covariance using the kernel formulation.

This approach is more efficient than Shrinkage if #features > #samples.

Parameters:	alpha : float The amount of shrinkage (>0 to 1) to apply to the empirical covariance matrix. Must be greater than zero. Defaults to 0.5

Methods

inv_dot(X, P)	Computes inv(cov(X)) @ P
loo_inv_dot(X, Ps[, remove_mean])	Computes inv(cov(X)) @ P in a leave-one-out scheme

compute_AB_parts
copy
fit
get_bounds
get_x0
update

__hash__($self, /)¶

Return hash(self).