0

I have a large raster stack that I need to subset by somewhat irregular time spans. I'd like to do something like this:

rstack[[c(1:3, 4:8, 9:15)]]

But there are 30 sequences like the ones above that I need to subset by, and I'd rather not type them all in manually.

I'm wondering if I can create a vector object, something like this:

vec <- c(1:3, 4:8, 9:15)
rstack[[vec]]

But based off of a data frame where I have the start, end values for the sequences to automate the process?:

df <- tibble(
start = c(1, 4, 9),
end = c(3, 8, 15)
)

# A tibble: 3 x 2
  start   end
  <dbl> <dbl>
1     1     3
2     4     8
3     9    15
  • How about raster::stack(rstack[[1:3]], rstack[[4:8]], rstack[[9:15]]) ? – Kazuhito Aug 31 at 6:48
0

With a data frame (not a tibble) like this:

> layers = data.frame(start=c(1,4,11),end=c(3,8,15))
> layers
  start end
1     1   3
2     4   8
3    11  15

Note I made it so that there's a gap in the total sequence so I can see it is doing something...

Do this to get the vector of merged sequences:

> unlist(apply(layers,1,function(s){s[1]:s[2]}))
 [1]  1  2  3  4  5  6  7  8 11 12 13 14 15

Assign that and use it to subset your stack:

> keep = unlist(apply(layers,1,function(s){s[1]:s[2]}))

s is my stack with 20 layers:

> dim(s)
[1]  3  4 20

subset to get 13 back:

> kept = s[[keep]]
> dim(kept)
[1]  3  4 13
> 

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.