How to replace numbers with NAs with condition

Question

I have two matrices, X1 and X2, with the same dimensions.

X2 has NAs values.

How can I put NAs values in X1 in the same position of X2 (replacing the values in X1)?

Please read the info about [how to ask a good question](http://stackoverflow.com/help/how-to-ask) and how to give a [reproducible example](http://stackoverflow.com/questions/5963269). This will make it much easier for others to help you. — zx8754, Jul 28 '16 at 10:24

score 3 · Answer 1 · edited Jun 20 '20 at 09:12

3

We can use replace

replace(X1, is.na(X2), NA)
#     [,1] [,2] [,3]
#[1,]   NA    4    7
#[2,]    2    5    8
#[3,]    3   NA    9

Or

X1 * NA^is.na(X2)
#     [,1] [,2] [,3]
#[1,]   NA    4    7
#[2,]    2    5    8
#[3,]    3   NA    9

Or as @Roland mentioned in the comments

is.na(X1) <- is.na(X2)

BTW,

X1 + X2 - X2
#Error in X1 + X2 : non-numeric argument to binary operator

Bottomline is that both the solutions I posted is general and works for non-numeric matrices as well.

X1 <- matrix(1:9, 3, 3)
X2 <- matrix(c(NA, "a", "b", "c",  "a", NA, "c","f", "a"), 3, 3)

edited Jun 20 '20 at 09:12

Community

answered Jul 28 '16 at 09:31

akrun

6

Or `is.na(X1) <- is.na(X2)` because it looks nice. – Roland Jul 28 '16 at 10:07

Ronak Shah · Answer 2 · 2016-07-28T09:55:54.513

2

You can do

X1[is.na(X2)] <- NA

In case if the data is completely numeric,try

NA + any number is NA. You can add X2 and then subtract it.

X1 + X2 - X2

edited Jul 28 '16 at 09:55

answered Jul 28 '16 at 09:41

Ronak Shah

First of all that is not my data and secondly, it won't work always. See my edit. – akrun Jul 28 '16 at 09:49

2 Answers2