Having Trouble Replacing NA

Question

I'm new to R and I'm using a data set from kaggle on bombing operations in WWII

I wanted to replace all the NA's in a column with "UKNOWN" but I get this

Warning message:
In `[<-.factor`(`*tmp*`, is.na(operations[,"theater.of.operations"]),  :
invalid factor level, NA generated

This is how I tried to do it

operations[, "theater.of.operations"][is.na(operations[, "theater.of.operations"])] <- "UNKNOWN"

How else should I try to do this?

Try using as.character() to cast your df column to a character first as Sotos mentions — Single Entity, Mar 31 '17 at 14:55
Thanks I didn't really know what was meant by "convert to character" but @SingleEntity made it more clear, so I used "operations$theater.of.operations <- as.character(operations$theater.of.operations) " and it worked. — Niall, Mar 31 '17 at 15:03
Possible duplicate of [invalid factor level, NA generated](http://stackoverflow.com/questions/16819956/invalid-factor-level-na-generated) — Ronak Shah, Mar 31 '17 at 15:06
Glad it worked for you, I shall add an answer to your question to make it official :) — Single Entity, Mar 31 '17 at 15:10
Another possible dupe: [replace in a factor column in R](http://stackoverflow.com/questions/39126537/replace-na-in-a-factor-column-in-r). — lmo, Mar 31 '17 at 15:11
Yes these are close duplicates, but not exact. I feel there has to be some understanding that the user may not have enough present knowledge to understand how those questions relate to their question, and that should be acceptable. — Single Entity, Mar 31 '17 at 15:14

score 2 · Accepted Answer · answered Mar 31 '17 at 15:12

To convert NA (which is not of string type) to a string such as "UNKNOWN" you need to cast your dataframe column to a string as follows.

operations$theater.of.operations <- as.character(operations$theater.of.operations)

The critical element being the function as.character() which casts to your required character variable type.

1 Answers1