mu_solver {SparseMDC}

Mu Solver

Description

Calculates the regularized center values for a cluster. This function runs inside sparse_mdc.

Usage

mu_solver(d, mu, v, EQ, dim, x, nk, p1, p2, delta)

Arguments

d

- The current condition.

mu

- A matrix of regularized mean values.

v

- Relationship matrix, describing relationship between d-1 and d.

EQ

- Equality matrix specifying the number of following terms to which mu_d is equal.

dim

- Total dimensions.

x

- Matrix of mean values.

nk

- Vector with the number of samples in each condition for this cluster.

p1

- Penalties on mean size.

p2

- Penalties on mean difference.

delta

Small term to ensure existance of solution.

Value

A matrix containing the regularized cluster means for each dimension.