Problems coercing factors into intergers

Question

I am trying to convert a factor (tickets_other) in a data frame (p2) into an integer. Following the R help guide, as well as other advice from others, this code should work:

as.numeric(levels(p2$tickets_other))[p2$tickets_other]

The column does contain NAs, and so I get a warning:

Warning message:
NAs introduced by coercion

Which is fine, but after coercing it to numeric, it still reads as a factor:

class(p2$tickets_other)
[1] "factor"

The same result happens if I use as.numeric(as.character.()):

as.numeric(as.character(p2$tickets_other))
Warning message: 
NAs introduced by coercion 
class(p2$tickets_other)
[1] "factor"

Post output of command 'head(p2)' or 'head(p2$tickets_other)' here. — rnso, Sep 24 '14 at 17:34
Without looking at your data. I'm guessing you need `as.numeric(as.character(p2$tickets_other))` — Vlo, Sep 24 '14 at 17:53
doesn't look like you're resetting p2$tickets_other with the result of your conversion? nothing will change if you don't store the result — DMT, Sep 24 '14 at 18:01

score 0 · Answer 1 · edited Sep 24 '14 at 18:17

0

You're doing:

as.numeric(levels(p2$tickets_other))[p2$tickets_other]

You are reading the levels from p2$tickets_other (a vector), then converting them to numeric (still a vector), then accessing the indices of that vector according to the values in p2$tickets_other

I can't imagine this is what you really want to do.

Maybe just

as.numeric(p2$tickets_other)

is what you want?

edited Sep 24 '14 at 18:17

Ben Bolker

211,554
25
370
453

answered Sep 24 '14 at 17:53

rcorty

1,140
1
10
28

score 0 · Accepted Answer · answered Sep 25 '14 at 18:22

I fixed the problem. It was actually very simple. The command:

as.numeric(levels(p2$tickets_other))[p2$tickets_other]

is correct, but I failed to store the result:

p2$tickets_other <- as.numeric(levels(p2$tickets_other))[p2$tickets_other]

Simple mistake, it retrospect. Thanks to DMT for the suggestion.

Problems coercing factors into intergers

2 Answers2