vol_preprocess: Preprocess the data for downstream volume analysis.

In vrnmf: Volume-Regularized Structured Matrix Factorization

Preprocess the data for downstream volume analysis.

Description

vol_preprocess Routine normalizes the data (as requested), estimates covariance and SVD decomposition.

Usage

vol_preprocess(X, col.norm = "sd", row.norm = NULL, pfactor = NULL)

Arguments

X	A numeric matrix. Covariance is estimated for column vectors of X.
col.norm	A character. Specifies column normalization strategy (by default "sd"). NULL to avoid normalization.
row.norm	A character. Specifies row normalization strategy (by default NULL).
pfactor	A numeric A factor to normalize co-occurence matrix (by default NULL). Row normalization follows column normalization. NULL to avoid normalization.

Value

A list of objects that include normalized matrix X.process, row and column normalization factors row.factors and col.factors, covariance matrix P0, covariance matrix P normalized to maximum value pfactor, orthonormal basis U and vector of eigenvalues eigens.

Examples

small_example <- sim_factors(5, 5, 5)
vol <- vol_preprocess(t(small_example$X))

vrnmf documentation built on March 18, 2022, 6:11 p.m.