sfpca: Compute rank 1 sparse and functional principal components

In alexpghayes/gsoc_moma_application: Application for GSOC MoMA Project

Description

Compute rank 1 sparse and functional principal components

Usage

sfpca(X, lambda_u = 0, lambda_v = 0, alpha_u = 0, alpha_v = 0,
  Omega_u = diag(nrow(X)), Omega_v = diag(ncol(X)))

Arguments

X	A data matrix. Considered n x p by convention.
lambda_u	Sparsity parameter for left singular vectors.
lambda_v	Sparsity parameter for right singular vectors.
alpha_u	Smoothness parameter for left singular vectors.
alpha_v	Smoothness parameter for right singular vectors.
Omega_u	Roughness penalty matrix (n x n) for left singular vectors. Must be positive semi-definite.
Omega_v	Roughness penalty matrix (p x p) for right singular vectors. Must be positive semi-definite.

Value

List with three named elements: left singular vector u, eigenval d, and right singular vector v.

alexpghayes/gsoc_moma_application documentation built on May 31, 2019, 2:21 p.m.