tobler.dasymetric.extract_raster_features

tobler.dasymetric.extract_raster_features(gdf, raster_path, pixel_values=None, nodata=255, n_jobs=-1, collapse_values=False)

Generate a geodataframe from raster data by polygonizing contiguous pixels with the same value using rasterio’s features module.

Parameters:

gdfgeopandas.GeoDataFrame

geodataframe defining the area of interest. The input raster will be clipped to the extent of the geodataframe

raster_pathstr

path to raster file, such as downloaded from <https://lcviewer.vito.be/download>

pixel_valueslist-like, optional

subset of pixel values to extract, by default None. If None, this function may generate a very large geodataframe

nodataint, optional

pixel value denoting “no data” in input raster

n_jobsint

[Optional. Default=-1] Number of processes to run in parallel. If -1, this is set to the number of CPUs available

collapse_valuesbool, optional

If True, multiple values passed to the pixel_values argument are treated as a single type. I.e. polygons will be generated from any contiguous collection of values from pixel_types, instead of unique polygons generated for each value This can dramatically reduce the complexity of the resulting geodataframe a fewer polygons are required to represent the study area.

Returns:

geopandas.GeoDataFrame

geodataframe whose rows are the zones extracted by the rasterio.features module. The geometry of each zone is the boundary of a contiguous group of pixels with the same value; the value column contains the pixel value of each zone.