crossv : error when loading crossv

[EtienneFerrandi]

"Error in -c(train.samples) : invalid argument to unary operator" when I launch crossv... Here is my code :

library(stylo)
training.set=read.csv("C:/Users/etien/Downloads/two_subcorpora/freq_table_reference_set.csv",sep = " ")
###inversion of rows and columns required for crossv###
training.set=t(training.set)
results=crossv(training.set,cv.mode = "stratified",cv.folds=5,classification.method = "svm")

[computationalstylistics]

Two things:

(1) when loading the file from disk, use the parameter row.names = 1 to tell R where your rownames are, namely in the 1st column:

training.set=read.csv("C:/Users/etien/Downloads/two_subcorpora/freq_table_reference_set.csv", sep = " ", row.names=1)

(2) the stratified cross-validation is not fully implemented in the function crossv() yet. It's still pending. Try leave-one-out instead:

training.set=t(training.set)
results=crossv(training.set, classification.method = "svm")

I hope this helps!