knn_label_count

knn_label_count(domain, label_name=None, objects_as_nodes=None, source_objects=None, include_boundaries=None, exclude_boundaries=None, boundary_exclude_distance=0, neighbourhood_size_KNN=10, max_distance=inf, min_distance=0, include_source_point=True)

Counts the number of each label within proximity of a source object using KNN networks. To be merged into a more general neighbourhood label count function.

Parameters:

domainmuspan domain

The domain containing the objects and labels.

label_namestr, optional

The name of the label to count, by default None.

source_objectsarray-like, optional

Indices of the source objects, by default None.

network_namestr, optional

The name of the network to use, by default None.

neighbourhood_size_KNNint, optional

The number of nearest neighbors to consider, by default 10.

max_distancefloat, optional

The maximum distance to consider for neighbors, by default np.inf.

min_distancefloat, optional

The minimum distance to consider for neighbors, by default 0.

include_source_pointbool, optional

Whether to include the source point in the count, by default True.

Returns:

ordered_labelslist

The ordered list of label categories.

count_dictdict

A dictionary where keys are object IDs and values are dictionaries with counts and neighbor objects.

Notes

To be merged into a more general neighbourhood label count function in V1.1.