preprocess_pca: Preprocess the input data by principal component analysis...

In UCLA-BD2K/CV.Signature.TCP: Cardiovascular Signature Temporal Clustering Platform

Description

The data is denoised and the missing values are imputed by using the top r eigenmatrices. After apply SVD on the centered and/or scaled data, the top r eigenmatrices are constructed where is r < min(n,m).

Usage

preprocess_pca(
  dat,
  r = NULL,
  method = c("em", "nipals"),
  center.dat = TRUE,
  scale.dat = FALSE,
  verbose = FALSE,
  seed = NULL,
  ...
)

Arguments

dat	a time-series data matrix with m biomarkers as rows, over n time points (columns).
r	the number of PCs or eigenmatrices to retain.
method	a method to perform singular-value decomposition when a dataset has missing values. See below for the explanation.
center.dat	a logical specifying to center the input and denoised data. By default, TRUE.
scale.dat	a logical specifying to scale the input and denoised data. By default, FALSE.
verbose	a logical specifying to print the computational progress. By default, FALSE.
seed	a seed for the random number generator.
...	optional arguments.

Details

To impute missing values with PCA/SVD, two approximation methods are provided. For method="nipals", a Non-linear Iterative Partial Least Squares (NIPALS) algorithm is used from nipals in the mixOmics package. For method="em", a low-rank SVD approximation by the EM algorithm is used from imputed.svd in the bcv package.

Value

preprocess_pca returns a matrix of imputed and/or denoised data.

Author(s)

Neo Christopher Chung [e-mail]

References

Troyanskaya, O., Cantor, M., Sherlock, G., Brown, P., Hastie, T., Tibshirani, R., Botstein, D. and Altman, R.B. (2001). Missing value estimation methods for DNA microarrays. Bioinformatics 17(6), 520–525.

Examples

## Not run: 
data(cys_optm)
meta <- cys_optm[,1:4]
optm <- log(cys_optm[meta$Select,5:10])
optm <- t(scale(t(optm), scale=TRUE, center=TRUE))
days <- as.numeric(colnames(optm))

preprocessed_optm <- preprocess_pca(optm, timepoints = days, r=3, method=c("nipals","em"))

## End(Not run)

UCLA-BD2K/CV.Signature.TCP documentation built on May 15, 2020, 11:27 p.m.