Replace all values that differ from the value "NA"

Question

I seem to have stumbled upon a predicament in my efforts to learn using R. I have a dataset that looks like this:

col1  /   col2  /   col3   

NA  / 1 /  2

1  /  NA  /2

What I need to do is replace everything that IS NOT NA with another value (see below)

Col1  /  col2  /    col 3

NA / Data / Data

Data / NA/ Data

I've looked everywhere on the web, it seems that everybody wants to replace NA's with other values, I on the other hand need the exact opposite but was unable to find any command. I even looked into "IF's" but was only able to find functions that replace values if they are eiter < or = or > to a certain number. I need something that will specify..."IF different from NA, then replace with...."

Sven Hohenstein Sven Hohenstein · Accepted Answer · 2012-11-15T10:29:33

The data:

dat <- data.frame(col1 = c(NA, 1), col2 = c(1, NA), col3 = c(2, 2))

To choose all values that are not NA, use the logical operator ! ("not") together with is.na:

dat[!is.na(dat)] <- "Data"

  col1 col2 col3
1 <NA> Data Data
2 Data <NA> Data

The same can also be achieved with the replace function:

replace(dat, !is.na(dat), "Data")

Another approach is needed if you want to replace the values of factors. This can be achieved with the following command:

replace(data.frame(lapply(dat, as.character), stringsAsFactors = FALSE),
        !is.na(dat), "Data")

Replace all values that differ from the value "NA"

1 Answers