R help - Jul 2003 - Help on NNET

Hi, Dear all,

I am just starting using R in my work and got some trouble to figure out some of
the errors. Can anybody help me?

The following is the script:

read.csv('pupil.txt',header=TRUE,sep='\t')->pupil
samp<-c(1:50, 112:162, 171:220, 228:278)
pupil.nn2 <- nnet(Type ~ ., data = pupil, subset = samp, size = 2, rang =
0.1, decay = 5e-4, maxit = 200)
table(pupil$Type[-samp], predict(pupil.nn2, pupil[-samp,], type =
"class"))

After running this, I got error information as

#Error in switch(type, raw = z, class = { : 
#        inappropriate fit for class

BTW, pupil.txt
X1   X2  Type
0.2  0.5  0
...........
..........
.........1
.........
.........
..........2
.......
........3
........
there are totally 351 records. 
My objective is to classify them into 4 classes.

Thanks a lot for your help!

I'm not a nnet() user, but after reading the help, there are
two things you might try, both in the model formula passed to
nnet() in the third line of code below.  Alternate version:

pupil.nn2 <- nnet( as.factor(Type) ~ X1 + X2, ...
	 	(with all the remaining arguments as before).

I'm not sure which (if either) of these two variants might do
the trick.

-  tom blackwell  -  u michigan medical school  -  ann arbor  -

On Thu, 17 Jul 2003, sbinny wrote:
> The following is the script:
>
> read.csv('pupil.txt',header=TRUE,sep='\t')->pupil
> samp<-c(1:50, 112:162, 171:220, 228:278)
> pupil.nn2 <- nnet(Type ~ ., data = pupil, subset = samp, size = 2, rang
= 0.1, decay = 5e-4, maxit = 200)
> table(pupil$Type[-samp], predict(pupil.nn2, pupil[-samp,], type =
"class"))
>
> There are totally 351 observations.
> My objective is to classify them into 4 classes.
>
> Thanks a lot for your help!