apply.scaling {iSubGen}

Apply scaling factors

Description

Apply scaling factors prior to autoencoder

Usage

apply.scaling(data.matrices, scaling.factors);

Arguments

data.matrices

list, where each element is a matrix. The list has one matrix for each data type to be scaled

scaling.factors

list with two elements named: \"center\" and \"scale\", and each element is a named numerical vector or a list of named numerical vectors. If scaling.factors$center or scaling.factors$scale are a list then each element needs to correspond to a one of the data matrices. Finally, the named numerical vectors should match the row and rownames from the corresponding data matrix.

Details

The names for the data matrices and the center and scale lists all must match.

Value

A list of matrices of the same format as the data.matrices

Author(s)

Natalie Fox

Examples

# Load molecular profiles for three data types and calculate scaling for each
example.molecular.data.dir <- paste0(path.package('iSubGen'),'/exdata/');
molecular.data <- list();
scaling.factors <- list();
for(i in c('cna','snv','methy')) {
  # Load molecular profiles from example files saved 
  # in the package as <data type>_profiles.txt
  molecular.data[[i]] <- load.molecular.aberration.data(
    paste0(example.molecular.data.dir,i,'_profiles.txt'),
    patients = c(paste0('EP00',1:9), paste0('EP0',10:30))
    );

  scaling.factors[[i]] <- list();

  scaling.factors[[i]]$center <- apply(molecular.data[[i]], 1, mean);
  scaling.factors[[i]]$scale <- apply(molecular.data[[i]], 1, sd);
  }

# Example 1: Transform the molecular profiles by the scaling factors
scaled.molecular.data <- apply.scaling(molecular.data, scaling.factors);

# Example 2: Transform one of the data types based on the scaling factors
scaled.molecular.data2 <- apply.scaling(
  molecular.data[[1]],
  scaling.factors[[1]]
  );