[R] idea for (further) debugging ace ()?

Lorenz Gygax lgygax at access.unizh.ch
Tue Mar 30 15:25:41 CEST 1999

Dear all,

I am trying to use ace on a 446 x 14 Matrix, with two categorical
variables in the first two columns. If I try to run it I get:

>   aceiH <- ace (data.mat, resp, cat= 1:length(x.cat))
Error: NAs in foreign function call (arg 3)

There is only one .Fortran call in ace () with the third argument being 
x = t(x). Thus I tried to debug ace, i.e. I went through ace until just
before the call to .Fortran. I can play with all the variables in the call
and all seem to be ok, specifically there are no NAs in t(x) but the call
fails also while debugging:

> debug (ace)
>   aceiH <- ace (data.mat, resp, cat= 1:length(x.cat))
debugging in: ace(data.mat, resp, cat = 1:length(x.cat))
debug: x <- as.matrix(x)
... ## go through ace until before .Fortran call
Browse[1]> table (is.na (t(x)))
debug: junk <- .Fortran("mace", p = ncol(x), n = nrow(x), x = t(x), 
    y = y, w = wt, l = l, delrsq = delrsq, ns = ns, tx = tx, 
    ty = ty, rsq = double(1), ierr = integer(1), m = m, z = z)
Browse[1]> junk
Error: Object "junk" not found

I am trying to find a smaller example that has the same effect so I could
send it to the list but was not successful until now, especially as I do
not really have any idea in what direction to search.

Any pointers or ideas are appreciated. Regards, Lorenz
--			~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Lorenz Gygax		 LGygax at amath.unizh.ch;       room: 36-L-40
			 Department of Applied Mathematics
			 University of Zuerich-Irchel
			 Winterthurerstr. 190; CH-8057 Zurich
			 voice: 41-1-635-58-52  fax: 41-1-635-57-05

r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list