Transform.no {grandR}

Transformations for PlotHeatmap

Description

Functions to perform transformations on the matrix used for PlotHeatmap.

Usage

Transform.no(label = " ")

Transform.Z(label = "z score", center = TRUE, scale = TRUE)

Transform.VST(label = "VST")

Transform.logFC(label = "log2 FC", LFC.fun = NULL, columns = NULL, ...)

Arguments

label

label that is used for the heatmap legend

center

perform centering when computing Z scores (see scale)

scale

perform scaling when computing Z scores (see scale)

LFC.fun

function to compute log fold changes (default: PsiLFC, other viable option: NormLFC)

columns

which columns (i.e. samples or cells) to use as reference when computing log fold changes (see details)

...

further parameters passed down to LFC.fun

Details

These functions should be used as transform parameter to PlotHeatmap. Available data transformations are

• transform=Transform.Z(): compute z scores for each row; you can omit the usual centering or scaling by setting the respective parameters to false; see scale

• transform=Transform.VST(): do a variance stabilizing transformation using vst

• transform=Transform.logFC(): compute log2 fold changes to one or several reference columns; see below how to define them; fold changes are computed using the lfc package)

• transform=Transform.no(): do not transform

The label to be used in the heatmap legend can be changed by specifying the label parameter.

For Transform.logFC, columns can be given as a logical, integer or character vector representing a selection of the columns (samples or cells).

Value

A function that transforms a matrix.