transformation of data

[Miguel Camacho]

Hi,

I have been reading in publications such as Lopes and  Boessenkool 2010 two steps related to data transformation for which I have the bellow questions:

1. It states all the statistics are normalised (mean = 0, var= 1) before estimation of euclidean distances. Should I do this by hand or the program rejection.exe does it automatically?

2. In the regression step it states that data is log or logit transformed. Again, does the function makepd4 transform (using the "trasnform" argument) and back-transform again or should I back transform by hand? In my experience, I run "transform="log"" and then loc1statsx to estimate the mode and the values seem already backtransformed.

Thanks,

Miguel