What does as.numeric applied to a factor do?

Asked Apr 20 '17 at 13:32

Active Apr 20 '17 at 13:32

Viewed 29 times

So, I have been looking at some code in R where the person mistakenly put as.numeric to convert a factor to a numeric value. I have already fixed this issue, but it got me wondering what does as.numeric actually do to a factor. I have been looking through the code and it seems to randomly choose what numbers to leave out. Here is some example code:

    d <- c(10,10,2,3,5,29,1, 1.5, 0.5,3,4)
    max_value <- c(4,4,4,4,5,5,2,2,2,2,2)
    max_value <- as.factor(max_value)

    d <= as.numeric(max_value)
    d[d <= as.numeric(max_value)]

I would assume the as.numeric of a factor would give the number of levels. However, this code has a factor with 3 levels and only keeps 2, 1, and 0.5. If it did use the number 3 as the max level, then it should also keep 3, 1.5, and 3. What is going on?

asked Apr 20 '17 at 13:32

Katie

1

Basically, `as.numeric` applied to a factor will simply return the (arbitrary) integer codes used to store the underlying factor. – joran Apr 20 '17 at 13:39
Thank you! How does it decide the order? – Katie Apr 24 '17 at 16:52
In the order of the sorted unique values in the original vector, I believe. – joran Apr 24 '17 at 17:39

What does as.numeric applied to a factor do?

0 Answers0