Stack overflow for a not so big task #1738
Description
I create a task with 298 samples and 22160 features, then do some resampling for CV. However, I got a stack overflow error. Is there any way to avoid this?
To reproduce the error (according to mb706's comment)
xt = cbind(as.data.frame(matrix(rnorm(298*22160), ncol=22160)), as.factor(sample(c('a', 'b'), 298, T)))
names(xt)[length(xt)] = 'x'
av_task = makeClassifTask("xt", xt, target='x')
r = resample(learner = makeLearner("classif.svm", predict.type = "prob"), task = av_task, resampling = makeResampleDesc(method = "CV", iters = 3), show.info = FALSE, measures = auc)Error: protect(): protection stack overflow