sk_coords {snapKrig}

Return coordinates of a grid of points in column-vectorized order

Description

Expands a set of y and x grid line numbers in the column-vectorized order returned by sk. This is similar to base::expand.grid but with the first dimension (y) descending instead of ascending.

Usage

sk_coords(g, out = "matrix", corner = FALSE, na_omit = FALSE, quiet = FALSE)

Arguments

g

any object accepted by sk

out

character indicating return value type, either 'list', 'matrix', or 'sf'

corner

logical, indicates to return only the four corner points

na_omit

logical, indicates to return only locations of observed points

quiet

logical, suppresses console output

Details

out='sf' returns an sf simple features object containing points in the same order, with data (if any) copied from g[['gval']] into column 'gval'. Note that length(g) points are created, which can be slow for large grids.

If na_omit is TRUE the function omits points with NA data (in gval) and only returns the coordinates for observations. This argument is ignored when corners=TRUE (which always returns the four corner points) or when the grid contains no observations (all points returned).

Value

a matrix, list, or sf point collection in column vectorized order

See Also

sk sk_snap base::expand.grid

Other exporting functions: sk_export()

Examples

gdim = c(5,3)
g_complete = sk(gdim=gdim, gres=c(0.5, 0.7), gval=seq(prod(gdim)))
sk_coords(g_complete)
sk_coords(g_complete, out='list')

# missing data example
idx_obs =  sort(sample.int(length(g_complete), 5))
g = sk(gdim=gdim, gres=c(0.5, 0.7))
g[idx_obs] = g_complete[idx_obs]
all.equal(sk_coords(g, na_omit=TRUE), sk_coords(g_complete)[idx_obs,])

# corner points
sk_coords(g, corner=TRUE)
sk_coords(g, corner=TRUE, out='list')

# repeat with multi-layer example
g_multi = sk(utils::modifyList(g, list(gval = cbind(g[], 2*g[]))))
all.equal(sk_coords(g_multi, na_omit=TRUE), sk_coords(g_complete)[idx_obs,])
sk_coords(g_multi, corner=TRUE)

# sf output type
if( requireNamespace('sf') ) {

# gather all points but don't copy data
sf_coords_all = sk_coords(sk(g, vals=FALSE), out='sf')

# extract non-NA data
sf_coords = sk_coords(g, out='sf', na_omit=TRUE)

# plot everything together
plot(g, reset=FALSE)
plot(sf_coords_all, add=TRUE)
plot(sf_coords, pch=16, cex=2, add=TRUE)

}