grid_indexing.RTree#

class grid_indexing.RTree(source_cells, shape=None)#

A geospatial RTree implementation

Based on the rstar rust crate, this class allows search for overlapping geometries (where “overlapping” refers to intersecting / containing but not touching).

Note that for now the search happens in the 2D cartesian space defined by the equirectangular (“PlateCarrée”) projection. As such, the search will not return the correct results around poles or the dateline.

Parameters:

  • source_cells (arrow-array) – The source geometries as a geoarrow polygon array, in EPSG:4326 coordinates. The longitude convention can be either 0° to 360° or -180° to 180°, and it is the user’s responsibility to ensure consistency.

  • shape (tuple of int, optional) – The shape of the input array. This is necessary to keep the shape of a 2D field of geometries, as geoarrow does not support 2D arrays (yet?).

    If omitted / None, a 1D array will be assumed.

__init__()#

Methods

__init__()

from_shapely(geoms)

construct a RTree from a array of shapely geometries

query(target_cells, *[, shape, method])

query for geometries

query_overlap(target_cells[, shape])

query for overlapping geometries
On this page

This Page