log_sparse_prior: The *un-normalised* prior over W, sigma^2 in SpStPCA. A...

Description Usage Arguments Value

Description

The *un-normalised* prior over W, sigma^2 in SpStPCA. A product of the proper, un-normalised p(W | \beta), and the improper p(sigma^2).

Usage

1
log_sparse_prior(K, W, sigSq, b)

Arguments

K

Prior covariance matrix

W

Loadings matrix

sigSq

variance of noise added to 'signal' Wv_i

b

Variance on Laplace part of prior

Value

un-normalised log prior over W


JimSkinner/spca documentation built on May 7, 2019, 10:52 a.m.