arcTrans: Efficient inverse hyperbolic cosine transformation

In jtheorell/flowSpecs: Tools for processing of high-dimensional cytometry data

Efficient inverse hyperbolic cosine transformation

Description

This is a simple wrapper function for the base asinh function, that is useful for flowFrames and flowSets. It also allows for reversing the transformation with the argument "unTrans".

Usage

arcTrans(flowObj, transNames, transCoFacs = "default", unTrans = FALSE)

Arguments

flowObj	The fcs object to be transformed. Both flowFrames and flowSets are accepted.
transNames	The variables that should be normalized.
transCoFacs	This value or vector of values define the values for the transformation during the normalization. In the "default" case, the function defines the object as a CyTOF object if >5 percent of the values are 0, and applies the transformation value 5. Otherwise, the value 400 is applied.
unTrans	If the reverse action should be taken, i.e. if an already transformed dataset should be un-transformed. NB! It is of great importance that the same transformation factors are used!

Value

A flow object containing the transformed data, and with all metadata left untouched.

Examples

# Import some data and the spectral matrix. The latter can be generated using
# specMatCalc
data(fullPanel)
data(specMat)

fullPanelUnmixed <- specUnmix(fullPanel, specMat)

# Identify the columns that should be transformed
colnames(fullPanelUnmixed)
# The time and scatter parameters should not, but apart from that, all should
# be included.
transNames <- colnames(fullPanelUnmixed)[seq(6,18)]

# ow, transform this file, with the default transformation factor of 400.
# NB! It is alway advisable to visually (or computationally) check the data
# for the most optimal transformation factors. These often vary from marker
# to marker.

fullPanelTrans <- arcTrans(fullPanelUnmixed, transNames)

jtheorell/flowSpecs documentation built on April 24, 2023, 2:59 a.m.