xrspatial.polygon_clip.clip_polygon#

xrspatial.polygon_clip.clip_polygon(raster: DataArray, geometry, nodata: float = nan, crop: bool = True, all_touched: bool = False, rasterize_kw: dict | None = None, name: str | None = None) → DataArray[source]#

Clip a raster to an arbitrary polygon geometry.

Pixels outside the polygon are set to nodata. When crop is True (the default), the output is also trimmed to the polygon’s bounding box so the result is smaller than the input.

Parameters:

  • raster (xr.DataArray) – Input raster to clip. Must be at least 2-D with named y and x dimensions (last two dimensions).

  • geometry (shapely geometry, list of geometries, GeoDataFrame, or coordinate array) –

    The clipping polygon(s). Accepts:

    • A single shapely.geometry.Polygon or MultiPolygon.

    • An iterable of shapely polygon geometries (merged via unary_union).

    • A GeoDataFrame or GeoSeries (merged via unary_union).

    • A list of (x, y) coordinate pairs defining a single polygon ring.

  • nodata (float, default np.nan) – Value to assign to pixels outside the polygon.

  • crop (bool, default True) – If True, also trim the output to the bounding box of the polygon. This reduces memory usage for small clip regions within large rasters.

  • all_touched (bool, default False) – If True, all pixels touched by the polygon boundary are included. If False (default), only pixels whose centre falls inside the polygon are included.

  • rasterize_kw (dict, optional) – Extra keyword arguments forwarded to rasterize() when creating the polygon mask.

  • name (str, optional) – Name for the output DataArray. Defaults to the input name.

Returns:

Clipped raster with the same dtype and attributes as raster.

Return type:

xr.DataArray

Examples

>>> from shapely.geometry import Polygon
>>> import xrspatial
>>> poly = Polygon([(1, 1), (1, 3), (3, 3), (3, 1)])
>>> clipped = xrspatial.clip_polygon(raster, poly)