--- a
+++ b/OmicsFold/man/normalise.logit.empirical.Rd
@@ -0,0 +1,31 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/MixMC.R
+\name{normalise.logit.empirical}
+\alias{normalise.logit.empirical}
+\title{Normalise using the logit function in an empirical manner}
+\usage{
+normalise.logit.empirical(input)
+}
+\arguments{
+\item{input}{Data frame of input compositional data to normalise. Input data
+should be proportions 0-1.}
+}
+\value{
+Data normalised using empirical logit. Proportions below 0.5 will be
+negative, but output will not tend to infinity for zero or 1 input.
+}
+\description{
+Apply the empirical logit normalisation to a data frame of omics input data.
+This is intended to convert compositional data, e.g. proportional data in the
+range 0..1, to Euclidean space which is most appropriate for the linear
+models. The empirical logit function differs for standard logit normalisation
+in that an epsilon factor is added to ensure that function does not tend to
++/- infinity for input values close to 100\% and 0\% respectively. The logit
+or empirical logit function will be a more appropriate choice than centred
+log-ratio (CLR) for non-OTU data.
+}
+\examples{
+\dontrun{
+normalise.logit.empirical(data.proportional)
+}
+}