[RsR] Box-Cox transformation and reverse Box-Cox transformation

[Mark Sharp]

I am wanting to use Box-Cox transformed values to develop a reference intervals for clinical chemistry tests given 25 observations per test. After I have used the transformed values to find the range as a function of the interquartile range, I want to remap the transformed variables back to the original scale using a reverse Box-Cox transformation. This will allow me to present the reference range in the original scale.

I have looked at the alr3 package, but do not see a way to get the reverse transformation and I have looked at boxcox (MASS), but it is dealing with linear models.

Any guidance will be appreciated.

Mark
R. Mark Sharp, Ph.D.
msharp using sfbr.org<mailto:msharp using sfbr.org>





	[[alternative HTML version deleted]]