CombineSubspacesTest: Combine subspaces (i.e. using Grassmann Manifold Technique -...

In ncats/MultiOmicsGraphPrediction: Network-Based Ensemble Prediction from IntLIM Predictors

Combine subspaces (i.e. using Grassmann Manifold Technique - see PMID 30329022) given two different modalities of data (e.g. gene and metabolite) and the alpha value and the desired number of eigenvectors.

Description

Combine subspaces (i.e. using Grassmann Manifold Technique - see PMID 30329022) given two different modalities of data (e.g. gene and metabolite) and the alpha value and the desired number of eigenvectors.

Usage

CombineSubspacesTest(
  type1SimilarityTrain,
  type2SimilarityTrain,
  type1SimilarityTest,
  type2SimilarityTest,
  eigenCount,
  alpha
)

Arguments

type1SimilarityTrain	A cosine similarity matrix for the first data type, found using ComputeCosineSimilarity on the training data.
type2SimilarityTrain	A cosine similarity matrix for the second data type, found using ComputeCosineSimilarity on the training data.
type1SimilarityTest	A cosine similarity matrix for the first data type, found using ComputeCosineSimilarity on the testing data.
type2SimilarityTest	A cosine similarity matrix for the second data type, found using ComputeCosineSimilarity on the testing data.
eigenCount	The number of eigenvectors to use.
alpha	The value of alpha to use.

Value

A named list including the data projected onto the merged subspace, the optimal number of eigenvectors, the optimal alpha value, the clustering coefficient, and the dendrogram.

ncats/MultiOmicsGraphPrediction documentation built on Aug. 23, 2023, 9:19 a.m.