Magnify.binary.features.norm-methods {NetPreProc}

Normalization of binary matrices

Description

Normalization of binary matrices according to the procedure described in Mostafavi et al. 2008. Having a binary matrix M, for each feature, if b is the proportion of 1, then ones are replaced with -log(b) and zeros with log(1-b).

Usage

Magnify.binary.features.norm(M)

Arguments

M

an object representing the matrix to be normalized

Value

The normalized matrix

Methods

signature(M = "matrix")

Input binary matrix. Rows are examples, columns features

References

Mostafavi, S., Ray, D., Warde-Farley, D., Grouios, C., & Morris, Q. GeneMANIA: a real-time multiple association network integration algorithm for predicting gene function. Genome Biology, 9, 2008.

Examples

D <- matrix(ifelse(runif(40000)>0.9,1,0),nrow=100);
M <- Magnify.binary.features.norm(D);