fs: Calculate an SVD on a matrix

In moosik/snorm: Supporting function for supervised exploration of data structure

Description

Use fast.svd from corpcor library to calculate an SVD of a matrix. The data before the calculation is centered. The D matrix is transformed to show percent variance explained. The missing or infinite numbers should be removed or imputed in the matrix prior to the calculation.

Usage

fs(x)

Details

The function is a slight modification of the function created by Brig Mecham.

@references mGenomics

@import corpcor

@param x data matrix

@return

u

a matrix with the corresponding left singular vectors. Column names of the matrix are the column names of the input data matrix x.

v

a matrix with the corresponding right singular vectors. We refer to them as eigen vectors in the documentation for all other functions. The matrix has column names in the following format "EigenN", where N is the number of the corresponding vector. The row names are the same as the column names of the input data matrix x.

d

a vector containing the values for the variance explained by each singular vector

@export

@seealso fast.svd

moosik/snorm documentation built on May 23, 2019, 6:11 a.m.