Just move the sampling out of your extraction so you can refer to it again:
ky <- mtcars
## Here, I've moved the sampling out of your extraction
forty <- sample(1:nrow(ky), nrow(ky)*0.4,replace=FALSE)
## Now you can extract whatever you want
ky[ forty, ] # This will be the 40% of original dataset
ky[-forty, ] # This will be the remaining rows.