"Error: Missing data in columns" due to fix.factors.prediction=TRUE #2611
Description
library(mlr)
train_data <- data.frame(
A = runif(100), B = factor(sample(c("A", "B"), 100, replace = T)))
test_data <- data.frame(
A = runif(100), B = factor(sample(c("A", "B", "C"), 100, replace = T)))
lrn <- makeLearner("regr.ranger", fix.factors.prediction = TRUE)
train_task <- makeRegrTask(
data = train_data,
target = "A"
)
model <- train(lrn, train_task)
predictions <- predict(model, newdata = test_data)Gives Error: Missing data in columns: B., although there is no missing data. Same for classification.
Kind regards