API Documentation

This is the API documentation for scikit-hubness.

Analysis: skhubness.analysis

The skhubness.analysis package provides methods for measuring hubness.

analysis.Hubness

Examine hubness characteristics of data.

Neighbors: skhubness.neighbors

The skhubness.neighbors package is a drop-in replacement for sklearn.neighbors, providing all of its features, while adding transparent support for hubness reduction and approximate nearest neighbor search.

neighbors.BallTree	BallTree for fast generalized N-point problems
neighbors.DistanceMetric	DistanceMetric class
neighbors.KDTree	KDTree for fast generalized N-point problems
neighbors.HNSW	Wrapper for using nmslib
neighbors.KNeighborsClassifier	Classifier implementing the k-nearest neighbors vote.
neighbors.KNeighborsRegressor	Regression based on k-nearest neighbors.
neighbors.FalconnLSH	Wrapper for using falconn LSH
neighbors.NearestCentroid	Nearest centroid classifier.
neighbors.NearestNeighbors	Unsupervised learner for implementing neighbor searches.
neighbors.ONNG	Wrapper for ngtpy and ONNG
neighbors.PuffinnLSH	Wrap Puffinn LSH for scikit-learn compatibility.
neighbors.RadiusNeighborsClassifier	Classifier implementing a vote among neighbors within a given radius
neighbors.RadiusNeighborsRegressor	Regression based on neighbors within a fixed radius.
neighbors.RandomProjectionTree	Wrapper for using annoy.AnnoyIndex
neighbors.kneighbors_graph	Computes the (weighted) graph of k-Neighbors for points in X
neighbors.radius_neighbors_graph	Computes the (weighted) graph of Neighbors for points in X
neighbors.KernelDensity	Kernel Density Estimation
neighbors.LocalOutlierFactor	Unsupervised Outlier Detection using Local Outlier Factor (LOF)
neighbors.NeighborhoodComponentsAnalysis	Neighborhood Components Analysis

Reduction: skhubness.reduction

The skhubness.reduction package provides methods for hubness reduction.

reduction.MutualProximity	Hubness reduction with Mutual Proximity [R5a390b9a9956-1].
reduction.LocalScaling	Hubness reduction with Local Scaling [Rf1f7cb70176a-1].
reduction.DisSimLocal	Hubness reduction with DisSimLocal [R3ede0f1b99b2-1].