Full_Name: Cam Webb
Version: 1.6.0 (fink X11 compile)
OS: Mac OS X (Jaguar)
Submission from: (NULL) (64.168.28.87)
This is an unpredictable, intermittant hang during cmdscale of the mva library.
Some data never cause a problem, other data always do, abut I can't track
down
the difference in the structure of the data. Sometimes the function will work
for `difficult' data after it has been run in the same session for
`behaved'
data (eurodist always works)! Here is a typical session:
$ R
> library(mva)
> x <- read.table("test.dat",row.names=1)
> x
V2 V3
A 1 2
B 2 4
C 5 7
D 1 1
> dist(x)
A B C
B 2.236068
C 6.403124 4.242641
D 1.000000 3.162278 7.211103
> cmdscale(dist(x))
^C
Here it just hangs, CPU whirring. Have to abort with ^C
> data(eurodist)
> cmdscale(eurodist)
[,1] [,2]
Athens 2290.274680 1798.80293
Barcelona -825.382790 546.81148
...
Vienna 911.230500 205.93020
> cmdscale(dist(x))
[,1] [,2]
A 1.9464399 0.1545044
B -0.2638877 0.4928116
C -4.4454618 -0.2244320
D 2.7629095 -0.4228840
Works now!
Not sure if this is a Mac problem, or a function problem. The intermittantness
is most puzzling.
Many thx,
Cam
