Fast ICA via Rust from processed data

This functions is a wrapper over the Rust implementation of fastICA expected already the pre-processed matrix and the pre-whitening matrix. It has the same two options c("logcosh", "exp") to run ICA in parallel modus. You can control the parameters of ICA via ica_params.

Usage

fast_ica_rust_helper(
  X,
  K,
  n_icas,
  ica_fun = c("logcosh", "exp"),
  seed = NULL,
  ica_params = params_ica_general()
)

Arguments

X

Numeric matrix. Processed data. Output of rs_prepare_whitening().

K

The K matrix. Pre-whitening matrix. Output of rs_prepare_whitening().

n_icas

Integer. Number of independent components to recover.

ica_fun

String, element of c("logcosh", "exp").

seed

Integer. Seed to ensure reproducible results.

ica_params

List. The ICA parameters, see params_ica_general() wrapper function. This function generates a list containing:

• maxit - Integer. Maximum number of iterations for ICA.

• alpha - Float. The alpha parameter for the logcosh version of ICA. Should be between 1 to 2.

• max_tol - Maximum tolerance of the algorithm.

• verbose - Controls verbosity of the function.

Value

A list containing:

• w The mixing matrix w.

• A ICA results matrix A.

• S ICA results matrix S.

• converged Boolean indicating if algorithm converged.

Examples

# same run, but with the whitening done up front
sources <- cbind(sin((1:1000) / 20), rep(((1:200) - 100) / 100, 5))
mixed <- sources %*% matrix(c(0.291, 0.6557, -0.5439, 0.5572), 2, 2)
whitened <- rs_prepare_whitening(
  x = mixed,
  fast_svd = TRUE,
  seed = 42L,
  rank = NULL,
  oversampling = NULL,
  n_power_iter = NULL
)
ica_res <- fast_ica_rust_helper(
  X = whitened$x,
  K = whitened$k,
  n_icas = 2L,
  ica_fun = "logcosh",
  seed = 42L
)
dim(ica_res$S)
#> [1]    2 1000