signedRescale {CNVScope}

Rescale positive and negative data, preserving sign information.

Description

Performs a signed rescale on the data, shrinking the negative and positive ranges into the [0,1] space, such that negative is always less than 0.5 and positive is always greater.

Usage

signedRescale(
  matrix,
  global_max = NULL,
  global_min = NULL,
  global_sigma = NULL,
  global_mu = NULL,
  max_cap = NULL,
  method = "minmax",
  tan_transform = F,
  global_sigma_pos = NULL,
  global_sigma_neg = NULL,
  asymptotic_max = T
)

Arguments

matrix

A matrix to be transformed

global_max

the global maximum (used if scaling using statistics from a large matrix upon a submatrix).

global_min

the global minimum

global_sigma

the global signma

global_mu

the global mu

max_cap

the maximum saturation– decreases the ceiling considered for the scaling function. Useful to see greater differences if an image is too white, increase it if there is too much color to tell apart domains.

method

method to perform the rescaling. Options are "minmax" (default), "tan" for tangent, and "sd" for standard devation

tan_transform

apply a tangent transformation?

global_sigma_pos

The positive global sigma. See getGlobalRescalingStats.

global_sigma_neg

The negative global sigma. See getGlobalRescalingStats.

asymptotic_max

make the maximum value in the matrix not 1, but rather something slightly below.

Value

transformedmatrix A transformed matrix.

Examples

mat<-matrix(c(5,10,15,20,0,40,-45,300,-50),byrow=TRUE,nrow=3)
rescaled_mat<-signedRescale(mat)
mat
rescaled_mat<-signedRescale(abs(mat))