Find and Replace number to NA in array R

Question

I have some NifTi files of brain images where there are lots of zeros that I want to replace with NAs but I'm not sure how to do it. I read in the description of the is.na() function that: "the generic function is.na<- sets elements to NA" so I thought I could use that, but it didn't work. Here is specifically what I tried:

library(RNifti)

in.path <- "R_Things/Voxel/"
out.path <- "R_Things/Outputs/"

ids <- c('100','101','102','103','104')

for (i in ids) {

a <- readNifti(paste0(in.path, i, ".nii.gz"))

is.na(a) <- 0


writeNifti(image=a, file=paste0(out.path, i, "_with_NAs.nii.gz"))

}

Any thoughts on what I could do differently would be very helpful, thanks!

Possible duplicate - [Replace all 0 values to NA](https://stackoverflow.com/questions/11036989/replace-all-0-values-to-na) — Ronak Shah, Jun 09 '21 at 03:32

score 0 · Answer 1 · answered Jun 08 '21 at 22:19

0

Here are some possible option to replace 0 by NA

> replace(a, a == 0, NA)
[1] NA  1  2 NA  3 -1

> a * NA^(a == 0)
[1] NA  1  2 NA  3 -1

Dummy Data

> a <- c(0, 1, 2, 0, 3, -1)

answered Jun 08 '21 at 22:19

ThomasIsCoding

96,636
9
24
81

The `a * NA^(a == 0)` is charming just for seeing a truth eval in an exponent. – Chris Jun 09 '21 at 01:31
Thanks @ThomasIsCoding but believe it or not neither of those worked. I guess it might have to do with the fact that the data is in an array? – beanboy Jun 09 '21 at 16:32

score 0 · Accepted Answer · answered Jun 09 '21 at 16:35

For anyone interested, here was my final code thanks to @Baroque

library(RNifti)

in.path <- "R_Things/Voxel/"
out.path <- "R_Things/Outputs/"

ids <- c('100','101','102','103','104')

for (i in ids) {

a <- readNifti(paste0(in.path, i, ".nii.gz"))

a[a==0] <- NA

writeNifti(image=a, file=paste0(out.path, i, "_with_NAs.nii.gz"))

}

Find and Replace number to NA in array R

2 Answers2