[R] error using daisy() in library(cluster). Bug?

[Martin Maechler]

¡Hola Javier!

since I am the maintainer of the cluster  
*package* (not "library"), I'm interested to find out more about
this problem.  I assume, you now use R 1.9.1.

Can you give us an example we can reproduce?
Give the exact R commands you use and 
maybe attach the save()d data file (*.rda) in a private e-mail?

Or do this on R-help and give an URL where one can download the
data (you can't attach such binary files for R-help).

Thank you,
Martin Maechler

>>>>> "javier" == javier garcia <- CEBAS <rn001 at cebas.csic.es>>
>>>>>     on Thu, 12 Aug 2004 12:53:28 +0200 writes:

    javier> Hi, I'm using the cluster library to examine
    javier> multivariate data.  The data come from a connection
    javier> to a postgres database, and I did a short R script
    javier> to do the analisys. With the cluster version
    javier> included in R1.8.0, daisy worked well for my data,
    javier> but now, when I call daisy, I obtain the following
    javier> messages: --------- Error in if (any(sx == 0)) { :
    javier> missing value where TRUE/FALSE needed In addition:
    javier> Warning message: binary variable(s) 116 treated as
    javier> interval scaled in: daisy(concentracion.data.frame,
    javier> stand = TRUE) ---------

    javier> Al the variables in my dataframe are
    javier> numeric. Although I've got NA values, and I've seen
    javier> that if a do the analisys for a subset of the
    javier> dataframe, selecting just columns with no NA, the
    javier> result is good.  Could this be a bug?

    javier> Thanks, and best regards

    javier> Javier