Scaling mixed models for PCA using dudi.mix I am trying to do a kselect model from the adehabitatHS which uses commands from ade4 package.  I am trying to determine if I need to scale my variables.   My surface understanding is that the k-select is basically a fancy PCA.  In their example they scaled their variables, but their variables were only continuous measures.  I have mixed categorical and continuous variables.   In their example, they scale their variables before they use the command dudi.pca which from my understanding this is needed to set up the k-select.  I know from reading the dudi help in the ade4 vinette that I should use dudi.mix here instead of .pca, but what to do about scaling?  Do I need to scale my variables? Do I need to scale all variables? Do I need to scale all the variables EXCEPT the categorical variable?   I can't find reading material that explains what is happening in the dudi/pca mixed process in sufficient detail.
Below is the example code from the k-select help if you would like to see what I am referring to.
data(puechabonsp)
locs <- puechabonsp$relocs
    map <- puechabonsp$map
pc <- mcp(locs[,"Name"])
hr <- hr.rast(pc, map)
cp <- count.points(locs[,"Name"], map)

## prepares the data for the kselect analysis
x <- prepksel(map, hr, cp)
tab <- x$tab

## Example of analysis with two variables: the slope and the elevation.
tab <- tab[,((names(tab) == "Slope")|(names(tab) == "Elevation"))]
tab <- scale(tab)

## A K-select analysis
acp <- dudi.pca(tab, scannf = FALSE, nf = 2)
kn <- kselect(acp, x$factor, x$weight, scannf = FALSE, nf = 2)

Again, their example only used continuous variables.  Does the dudi.mix scale the variables appropriately in the code that we don't see?
 A: I found this as a partial explanation in http://cran.r-project.org/web/packages/adehabitatHS/vignettes/adehabitatHS.pdf. 

The function dudi.pca is to be used when all the variables present in
  the data.frame are numeric. The function dudi.acm is to be used when
  all the vari- ables present in the data.frame are factors. The
  function dudi.hillsmith (or, equivalently, dudi.mix) is to be used
  when the data.frame contains both types of variables. These functions,
  used as a preliminary to the GNESFA, are needed to scale the table
  suitably (so that all the variables have the same mean and the same
  variance), and to compute the weights of the variables in the
  analysis. For example, the use of dudi.hillsmith on a table containing
  a numeric variable and a factor with four levels ensures that the
  factor will have the same weight in the analysis as the numeric
  variable

I can only assume the dudi.hillsmith is scaling the data correctly and I do not need to scale my data before using the k-select.
