Linking focal cell ids with adjacent cell ids using terra in R

Question

I have a large raster and I need to link the ids of each of the focal cells with the ids of each focal's adjacent cell (in my case 'queen' neighbours), as shown below

finding raster neighbours, including the id of the focal cell

In my case, each focal cell is a 5km square representing use by an animal, and each neighbouring cell represents an available domain for that location (but the focal cell is also available and must be included). Within this larger 5km grid I am extracting covariates at 30 m resolution for each of those 5km squares. I hope that makes sense!

I tried this solution: (Get values from adjacent cells raster R)

But, while close to what I need, it doesn't produce the data in a one-to-many format that I think I need.

Here is my code to generate an example raster, and generate those adjacent cells, but I'm just not sure how to link the id of each of the adjacents to the id of the focal. Importantly, the focal must also be considered as part of the available domain, thus an adjacent.

create example raster

library(terra)
d = rast(nrow = 5, ncol = 5)
values(d) <- 1:ncell(d)
#select some cell ids to test
cell_ids = c(3, 14, 22)

#I'm not sure this next line is required
dvec = as.data.frame(d)
colnames(dvec) = "focal_id"

run terra::adjacent

dadj = as.vector(terra::adjacent(d, cells = cell_ids, include = TRUE, directions = "queen"))

#cell id vec to
cell.rep = rep(c(3,14,22),each=9)

#and this is not correct
df = as.data.frame(cbind(cell.rep, dadj))

What I want is an output as such: formatted output I'm after

#EDIT: I just found that:

cell.rep = rep(c(3,14,22),each=9)


dadj = c()
for(i in cell_ids){
   adel = as.vector(terra::adjacent(d, cells = i, include = TRUE,
   directions = "queen"))
   dadj = c(dadj, adel)
}

df = as.data.frame(cbind(cell.rep, dadj))

Produces what I'm after, but is there a faster way?

try rep(c(3,14,22),times=9) instead of each=9. This would preserve the order of your vector. — one
– one, Commented Oct 9, 2023 at 19:40

Robert Hijmans · Accepted Answer · 2023-10-10 14:29:22Z

0

You can do this in one call to terra::adjacent with argument "pairs=TRUE"

# example data 
library(terra)
d <- rast(nrow = 5, ncol = 5)
cell_ids <- c(3, 14, 22)

# solution
da <- adjacent(d, cell_ids, directions="queen", pairs=TRUE, include=TRUE)

head(da)
#    from to
#[1,]    3  3
#[2,]    3  2
#[3,]    3  4
#[4,]    3  7
#[5,]    3  8
#[6,]    3  9

edited Oct 10, 2023 at 14:29

answered Oct 10, 2023 at 8:21

Robert Hijmans

49.5k4 gold badges62 silver badges77 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

rgill Over a year ago

Excellent, thanks! I know this wasn't part of my original question, but ultimately I want to extract values from those adjacent cells and associate those values with the focal cell. Does that functionality exist in adjacent() or do I need to just go the route of linking values via cell id?

Robert Hijmans Over a year ago

You could do d[ da[ ,2] ] to extract from d (or use focal instead). Extracting values from another, higher resolution, raster is more involved.

rgill Over a year ago

Apologies if this should be a new question, if so I can re-submit it as such. I'm getting some strange behaviour when using adjacent() on a large vector of focal cells (100,000). I'm only getting 2 neighbours for each focal cell, and one of those values is 1, even though the minimum index value of my raster is 2,567,191. I must be doing something wrong, but I've followed your code above exactly. Any idea what might be causing that incorrect adjacent value?

Collectives™ on Stack Overflow

Linking focal cell ids with adjacent cell ids using terra in R

1 Answer 1

3 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related