specmine (version 4.0.0)

ica_pairs_kmeans_plot: ICA pairs k-means plot

Description

Creates a pairs plot for independent components coloured by k-means clusters.

Usage

ica_pairs_kmeans_plot(
  dataset,
  ica.result,
  num.clusters = 3,
  kmeans.result = NULL,
  use.embedding = TRUE,
  dims = 1:min(5, ncol(ica.result$embedding))
)

Value

A GGally pairs plot object.

Arguments

dataset

Dataset to cluster.

ica.result

Result returned by ica_analysis_dataset().

num.clusters

Number of k-means clusters.

kmeans.result

Optional k-means result containing a cluster vector.

use.embedding

Logical indicating whether k-means should be applied to the ICA embedding when kmeans.result is not supplied.

dims

Independent components to include in the pairs plot.

Examples

if (requireNamespace("fastICA", quietly = TRUE) &&
    requireNamespace("GGally", quietly = TRUE)) {
  datamat <- matrix(rnorm(240), nrow = 8)
  dataset <- list(
    data = datamat,
    metadata = data.frame(
      class = factor(rep(c("A", "B"), each = 15))
    )
  )

  ica.result <- ica_analysis_dataset(
    dataset,
    n_components = 3,
    maxit = 100,
    seed = 42
  )

  kmeans.result <- list(
    cluster = rep(1:2, each = 15)
  )

  ica_pairs_kmeans_plot(
    dataset,
    ica.result,
    num.clusters = 2,
    kmeans.result = kmeans.result,
    dims = 1:3
  )
}