I wrote
> Hello
> 
> Using R 2.7.0 on Windows.
> 
> I am running a linear discriminant analysis as follows
> 
> <<<<
> discrim1 <- lda(normvar~ mafmahal+  mrfmahal+ mffmahal+ bafmahal+ brfmahal+ 
>       cofmahal+ bmfmahal+ cfmahal+ fractmahal+ antmahal+ absmifmahal+ absifmahal, subset = train)
> prediction <- predict(discrim1, traintest1[-train,])$class
> 
> When I do this, I get a warning message:
> 
> <<<
> Warning message:
> 'newdata' had 795 rows but variable(s) found have 796 rows 
> 
> However, when I look at 'newdata' (i.e. traintest1) with dim, I find it has 796 rows:
> 
> <<<
>> dim(traintest1)
> [1] 796 612
> 

What happens when you do

dim(traintest1[-train,])

?

> So, I do not understand the warning.
> 
> Thanks for any explanation or assistance
> 


I still don't understand this, as I thought I was following the example in  predict.lda;
but I did figure out how to avoid it:

<<<
test <- traintest1[train == 0,]
prediction <- predict(discrim1, test)$class
prediction <- prediction[train == 0]

table(prediction, normvar[train == 0])
>>>

Peter

	[[alternative HTML version deleted]]