esda.silhouettes.nearest_label¶

esda.silhouettes.nearest_label(data, labels, metric=<function euclidean_distances>, return_distance=False, keep_self=False)[source]¶

Find the nearest label in attribute space.

Given the data and a set of labels in labels, this finds the label whose mean center is closest to the observation in data.

Parameters:

data(N,P) array to cluster on or DataFrame indexed on the same values as: that in alist.focal/alist.neighbor
labels(N,) array containing classifications, indexed on the same values: as that in alist.focal/alist.neighbor
metriccallable(), array,: a function that takes an argument (data) and returns the all-pairs distances/dissimilarity between observations.
return_distance: bool: Whether to return the distance from the observation to its nearest cluster in feature space. If True, the tuple of (nearest_label, dissim) is returned. If False, only the nearest_label array is returned.
keep_self: bool: whether to allow observations to use their current cluster as their nearest label. If True, an observation’s existing cluster assignment can also be the cluster it is closest to. If False, an observation’s existing cluster assignment cannot be the cluster it is closest to. This would mean the function computes the nearest alternative cluster.

Returns:

(N_obs,) array of assignments reflect each observation’s nearest label.
If return_distance is True, a tuple of ((N,) and (N,)) where the first: array is the assignment, and the second is the distance to the centroid of that assignment.