jillbo1000 / eztune Goto Github PK

If a matrix with one row is given as input to the predict function, with the trained model being an elasticnet tuned by eztune, the following error occurs

Error in rep(0, nrow(dat)) : invalid 'times' argument

This bug can be fixed by replacing line 78 of https://github.com/jillbo1000/EZtune/blob/master/R/predict_eztune.R with the following

x <- x[, colnames(x) %in% nms, drop=FALSE]

Without drop=FALSE x becomes a vector and the error occurs in line 79.