R: Contrastive Principal Component Analysis in Parallel

bpFitCPCA {scPCA}

R Documentation

Contrastive Principal Component Analysis in Parallel

Description

Given target and background dataframes or matrices, cPCA will perform contrastive principal component analysis (cPCA) of the target data for a given number of eigenvectors and a vector of real valued contrast parameters. This is identical to the implementation of cPCA method by Abid et al. Abid et al. (2018). Analogous to fitCPCA, but replaces all lapply calls by bplapply.

Usage

bpFitCPCA(
  target,
  center,
  scale,
  c_contrasts,
  contrasts,
  n_eigen,
  n_medoids,
  eigdecomp_tol,
  eigdecomp_iter
)

Arguments

`target`	The target (experimental) data set, in a standard format such as a `data.frame` or `matrix`.
`center`	A `logical` indicating whether the target and background data sets should be centered to mean zero.
`scale`	A `logical` indicating whether the target and background data sets should be scaled to unit variance.
`c_contrasts`	A `list` of contrastive covariances.
`contrasts`	A `numeric` vector of the contrastive parameters used to compute the contrastive covariances.
`n_eigen`	A `numeric` indicating the number of eigenvectors to be computed.
`n_medoids`	A `numeric` indicating the number of medoids to consider.
`eigdecomp_tol`	A `numeric` providing the level of precision used by eigendecompositon calculations. Defaults to `1e-10`.
`eigdecomp_iter`	A `numeric` indicating the maximum number of interations performed by eigendecompositon calculations. Defaults to `1000`.

Value

A list of lists containing the cPCA results for each contrastive parameter deemed to be a medoid.

rotation - the list of matrices of variable loadings
x - the list of rotated data, centred and scaled if requested, multiplied by the rotation matrix
contrast - the list of contrastive parameters
penalty - set to zero, since loadings are not penalized in cPCA

References

Abid A, Zhang MJ, Bagaria VK, Zou J (2018). “Exploring patterns enriched in a dataset with contrastive principal component analysis.” Nature communications, 9.

[Package scPCA version 1.4.0 Index]