corelay.processor.embedding

Embedding Processors

Classes

EigenDecomposition	Eigenvalue Decomposition
Embedding	Embedding Processor base class
LLEEmbedding	LocallyLinearEmbedding
PCAEmbedding	PCA Embedding
TSNEEmbedding	TSNE Embedding
UMAPEmbedding	UMAPEmbedding: https://umap-learn.readthedocs.io/en/latest/index.html

class corelay.processor.embedding.EigenDecomposition(*args, **kwargs)

Bases: Embedding

Eigenvalue Decomposition

function(data)

Compute spectral embedding of data

Parameters:

data (numpy.ndarray) – data with samples in rows

Returns:

• numpy.ndarray – Eigenvalues for spectral embedding

• numpy.ndarray – Spectral embedding (eigenvectors)

Note

We use the fact that (I-A)v = (1-λ)v and thus compute the largest eigenvalues of the identity minus the data and return one minus the eigenvalue.

class corelay.processor.embedding.Embedding(*args, **kwargs)

Bases: Processor

Embedding Processor base class

class corelay.processor.embedding.LLEEmbedding(*args, **kwargs)

Bases: Embedding

LocallyLinearEmbedding

function(data)

Abstract function this Processor should apply on input

Parameters:

data (object) – Input data to this Processor.

Raises:

NotImplementedError – Always, since this is an abstract function.

class corelay.processor.embedding.PCAEmbedding(*args, **kwargs)

Bases: Embedding

PCA Embedding

function(data)

Abstract function this Processor should apply on input

Parameters:

data (object) – Input data to this Processor.

Raises:

NotImplementedError – Always, since this is an abstract function.

class corelay.processor.embedding.TSNEEmbedding(*args, **kwargs)

Bases: Embedding

TSNE Embedding

function(data)

Abstract function this Processor should apply on input

Parameters:

data (object) – Input data to this Processor.

Raises:

NotImplementedError – Always, since this is an abstract function.

class corelay.processor.embedding.UMAPEmbedding(*args, **kwargs)

Bases: Embedding

UMAPEmbedding: https://umap-learn.readthedocs.io/en/latest/index.html

function(data)

Abstract function this Processor should apply on input

Parameters:

data (object) – Input data to this Processor.

Raises:

NotImplementedError – Always, since this is an abstract function.