glog: glog transformation
In FitAR: Subset AR Model Fitting

Description Usage Arguments Details Value Author(s) References See Also Examples

The glog is a better behaved log transformation when some data values are zero or just near zero.

1	glog(x, a = 1, InverseQ = FALSE)

`x`	numeric vector of data
`a`	additive constant, often 1
`InverseQ`	inverse glog

Basic properties of the glog transformation are illustrated in the Mathematica notebook glog.nb and its pdf version glog.pdf which are available in the package directory doc.

transformated data

A.I. McLeod

W. Huber, A. von Heydebreck, H. Sultmann, A. Poustka, and M. Vingron. Variance stablization applied to microarray data calibration and to quantification of differential expression. Bioinformatics, 18: S96-S10 2002.

bxcx

#usual log transformation doesn't work
all(is.finite(log(sunspot.month)))
#either shifted log
all(is.finite(log(sunspot.month+1)))
#or glog works
all(is.finite(glog(sunspot.month)))
#but glog may be better, especially for values <1 but >=0