Skip to contents

Build an analysis grid from a raster or vector source.

grid_from_src.Rd

Derives a GridSpec from any GDAL/OGR-readable file or URL. The two source kinds are treated differently, because they carry different information:

Usage

grid_from_src(
  x,
  res,
  projection = c("laea", "aeqd", "utm", "pconic", "eqdc"),
  ellps = "WGS84",
  buffer = 0,
  dtype = "f32"
)

Arguments

x

Path or URL to a raster or vector source.

res

Target resolution. For a raster, in the source's native CRS units; for a vector, in the chosen projected CRS units (see grid_from_bbox()).

projection, ellps

Vector sources only; ignored for rasters.

buffer, dtype

As in grid_from_bbox().

Value

A GridSpec.

Details

  • Raster (GeoTIFF, COG, ...): the source already has a projected CRS and footprint, so the grid keeps that native CRS and extent and is simply re-gridded to res (the extent is snapped out to whole multiples of res). projection/ellps are ignored: reprojecting a raster's own grid would distort it. Use this to coarsen a tile in place, e.g. read a 10 m embedding COG onto a 30 m grid while preserving its geometry.

  • Vector (GeoJSON, GeoPackage, shapefile, ...): only a footprint is known, so a projected CRS is chosen for it and the grid is built exactly as grid_from_bbox() does, honouring projection/ellps.

Remote sources are read via HTTP range requests (header only), so this does not download the whole file.

See also

grid_from_bbox() to build a grid from a lon/lat bounding box; grid_spec() to build one from an explicit extent.

Examples

if (FALSE) { # \dontrun{
# Vector footprint -> equal-area 30 m grid.
target <- grid_from_src("catchment.gpkg", res = 30)
# Raster -> its own CRS/extent at 30 m.
g <- grid_from_src("s3://.../embedding.tif", res = 30)
} # }