rtemis (version 0.79)

d.CUR: CUR Decomposition

Description

Performs CUR decomposition using rCUR::CUR

Usage

d.CUR(x, c = dim(x)[2], r = dim(x)[1], k = NULL, sv = NULL,
  method = "highest.ranks", matrix.return = TRUE,
  error.return = FALSE, scale = TRUE, center = TRUE,
  verbose = TRUE, ...)

Arguments

x

Numeric matrix / data.frame: Input data

c

Integer: Number of columns to be selected

r

Integer: Number of rows to be selected

k

Integer: Rank of decomposition (Creates k components)

sv

The SVD of x, if already available

method

String: "random", "top.scores", "ortho.top.scores", "exact.num.random", "highest.ranks" (Default). See rCUR::CUR("method")

matrix.return

Logical: if TRUE, matrices C, U, and R are returned. If FALSE, U is not computed, which can be expensive, if r and c are large. Default = TRUE

error.return

Logical: if TRUE, the Frobenius norm of the difference between the original matrix and the CUR approximation is returned. Effective only if matrix.return = TRUE. Default = FALSE

scale

Logical: If TRUE, scale input

center

Logical: If TRUE, center input

verbose

Logical: If TRUE, print messages to output

...

Additional parameters to be passed to rCUR::CUR

Value

rtDecom object

Details

Note that k here does not correspond with k in the other decomposition functions. Use c to determine dimensionality of resulting decomposition

See Also

Other Decomposition: d.H2OAE, d.H2OGLRM, d.ICA, d.ISOMAP, d.KPCA, d.LLE, d.MDS, d.NMF, d.PCA, d.SPCA, d.SVD, d.TSNE, d.UMAP