Caret mvrValstats Error Message

[jahneldardavik]

I try to run caret, but get an error messge:
Error in mvrValstats(object = object, estimate = "train") :
could not find function "mvrValstats"

Here is what I do:

arcene <- read.table('http://archive.ics.uci.edu/ml/machine-learning-databases/arcene/ARCENE/arcene_train.data', sep =' ',colClasses = c(rep('numeric',10000),'NULL'))
#head(arcene)
#dim(arcene)
#head[1:4, 1:20]
str(arcene)
arcene$class <- factor(scan('https://archive.ics.uci.edu/ml/machine-learning-databases/arcene/ARCENE/arcene_train.labels', sep = '\t'))

any(is.na(arcene))

set.seed(100)
myfolds <- createMultiFolds(arcene$class, k = 5, times = 10)
control <- trainControl('repeatedcv', index = myfolds, selectionFunction = 'oneSE')

mod1 <- train(class ~ ., data = arcene, method = 'pls', metric = 'Accuracy', tuneLength = 20, trControl = control, preProc = c('zv','center','scale'))
plot(mod1)
plot(varImp(mod1),10, main = 'PLS')

The last plot statement results in this error message:
Error in mvrValstats(object = object, estimate = "train") :
could not find function "mvrValstats"

Advice is greatly appreciated.

Thank you
jahn davik

[topepo]

It is a bug that I'll fix in the upcoming release. In the meantime, load the pls package before computing the importances.

[jahneldardavik]

OK. Thank you. Which release can I go back to in order to avoid the error? jahn Fra: Max Kuhn [mailto:notifications@github.com] Sendt: 1. mars 2018 23:48 Til: topepo/caret <caret@noreply.github.com> Kopi: Jahn Davik <jahn.davik@nibio.no>; Author <author@noreply.github.com> Emne: Re: [topepo/caret] Caret mvrValstats Error Message (#848) It is a bug that I'll fix in the upcoming release. In the meantime, load the pls package before computing the importances. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<#848 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AYkgZh_2Ss1cVcG26Snhjor9h5jPjgBGks5taHqJgaJpZM4SYPjc>.

[topepo]

No need to change versions. Just load the pls package before you do the importance calculations.