Re-factor a data frame after some values no longer exist

Question

I am trying to re-factor across all column factors in a data frame to reduce the number of factor levels. Example is in my case, after I removed about 50% of rows in my data frame, a lot of columns have values in their factor levels that no longer exist, and I want to remove them. The data frame contains quite a lot of columns which are factors, so re-factoring them one at a time was getting tedious for me.

There's another question in SO called refactor data.frame column values but it doesn't address my question, so I thought I'll put in the question here as reference.

@henrik, you really need to get your r gold badge already :) — David Arenburg, Jan 06 '15 at 08:12
Ah obviously I'm not good at searching for questions in SO :( To the veterans: should I delete this redundant question? — Ricky, Jan 06 '15 at 08:15
No need to delete it. Duplicates serve exactly the purpose of making the issue better to search. — Roland, Jan 06 '15 at 08:16

score 1 · Accepted Answer · answered Jan 06 '15 at 08:07

1

I found some answers here, and the one I chose for my purpose was from Prof Brian Ripley:

ind <- sapply(DF, is.factor)
DF[ind] <- lapply(DF[ind], "[", drop=TRUE)

where the DF is the dataframe.

answered Jan 06 '15 at 08:07

Ricky

4,616
6
42
72

Have you tried `droplevels` or call `factor` again – akrun Jan 06 '15 at 08:09

Re-factor a data frame after some values no longer exist

1 Answers1