Programming interface

The interface of manifold learning methods in this packages is partially adopted from the packages StatsBase, MultivariateStats.jl and LightGraphs.jl. You can easily implement additional dimensionality reduction algorithms by implementing the following interface.

Types and functions

The following functions are currently available from the interface. AbstractDimensionalityReduction is a an abstract type required for all implemented algorithms models.

For performing the data dimensionality reduction procedure, a model of the data is constructed by calling fit method, and the transformation of the data given the model is done by transform method.

StatsBase.fitMethod
fit(AbstractDimensionalityReduction, X)

Perform model fitting given the data X

source
MultivariateStats.transformMethod
transfrom(R::AbstractDimensionalityReduction)

Returns a reduced space representation of the data given the model R inform of the projection matrix (of size $(d, n)$), where d is a dimension of the reduced space and n in the number of the observations. Each column of the projection matrix corresponds to an observation in projected reduced space.

source

There are auxiliary methods that allow to inspect properties of the constructed model.

LinearAlgebra.eigvalsMethod
eigvals(R::AbstractDimensionalityReduction)

Returns eignevalues of the reduced space reporesentation for the model R

source
LightGraphs.verticesMethod
vertices(R::AbstractDimensionalityReduction)

Returns vertices of largest connected component in the model R.

source
LightGraphs.neighborsMethod
neighbors(R::AbstractDimensionalityReduction)

Returns the number of nearest neighbors used for aproximate local subspace

source