[R] Need help with error

[Peter Ehlers]

On 2011-03-21 10:37, Savitri N Appana wrote:
> Thank you for your suggestion Allan.  I should have paid attention to
> the posting instructions.
>
>
> Pls find below the sample code from the ?splsda in the caret package.
> Note:  It used to work fine in R v2.8.1, but this error shows up now,
> given that I've modified the code on how the splsda or predict.splsda
> functions are called, i.e. as caret:::splsda and caret:::predict.splsda
> b/c I am running R v2.12.1 now.
>
>
> ######## sample code below..
> library(caret)
>
>
> data(mdrr)
> set.seed(1)
> inTrain<- sample(seq(along = mdrrClass), 450)
>
> nzv<- nearZeroVar(mdrrDescr)
> filteredDescr<- mdrrDescr[, -nzv]
>
>
> training<- filteredDescr[inTrain,]
> test<- filteredDescr[-inTrain,]
> trainMDRR<- mdrrClass[inTrain]
> testMDRR<- mdrrClass[-inTrain]
>
> preProcValues<- preProcess(training)
>
>
> trainDescr<- predict(preProcValues, training)
> testDescr<- predict(preProcValues, test)
>
>
> splsFit<- caret:::splsda(trainDescr, trainMDRR,
>                            K = 5, eta = .9,
>                            probMethod = "Bayes")
>> splsFit### ERROR is HERE!!!!
> Error in switch(classifier, logistic = { : EXPR must be a length 1
> vector
<snip>

This message comes from print.splsda() in the spls package.
As the caret:::splsda help page indicates, caret's splsda()
uses the spls:::spls() function. So, although your splsFit
object has class "splsda", you should print it with

  print.spls(splsdaFit)

Peter Ehlers