Package 'Vicus'

Title: Exploiting Local Structures to Improve Network-Based Analysis of Biological Data
Description: Compared with the similar graph embedding method such as Laplacian Eigenmaps, 'Vicus' can exploit more local structures of graph data. For the details of the methods, see the reference section of 'GitHub' README.md <https://github.com/rikenbit/Vicus>.
Authors: Koki Tsuyuzaki [aut, cre]
Maintainer: Koki Tsuyuzaki <[email protected]>
License: MIT + file LICENSE
Version: 0.99.0
Built: 2024-11-05 04:28:07 UTC
Source: https://github.com/rikenbit/vicus

Help Index


Exploiting Local Structures to Improve Network-Based Analysis of Biological Data

Description

Compared with the similar graph embedding method such as Laplacian Eigenmaps, 'Vicus' can exploit more local structures of graph data. For the details of the methods, see the reference section of 'GitHub' README.md <https://github.com/rikenbit/Vicus>.

Details

The DESCRIPTION file:

Package: Vicus
Type: Package
Title: Exploiting Local Structures to Improve Network-Based Analysis of Biological Data
Version: 0.99.0
Authors@R: c(person("Koki", "Tsuyuzaki", role = c("aut", "cre"), email = "[email protected]"))
Depends: R (>= 3.4.0)
Imports: methods, utils, RANN, Matrix, RSpectra
Suggests: scatterplot3d, knitr, rmarkdown, testthat
Description: Compared with the similar graph embedding method such as Laplacian Eigenmaps, 'Vicus' can exploit more local structures of graph data. For the details of the methods, see the reference section of 'GitHub' README.md <https://github.com/rikenbit/Vicus>.
License: MIT + file LICENSE
URL: https://github.com/rikenbit/Vicus
VignetteBuilder: knitr
Repository: https://rikenbit.r-universe.dev
RemoteUrl: https://github.com/rikenbit/vicus
RemoteRef: HEAD
RemoteSha: 4be58d784ddf32d526fe4dd2bac40dda53d64efa
Author: Koki Tsuyuzaki [aut, cre]
Maintainer: Koki Tsuyuzaki <[email protected]>

Index of help topics:

Vicus-package           Exploiting Local Structures to Improve
                        Network-Based Analysis of Biological Data
embedding               Graph Embedding
graphMatrix             Graph Laplacian type matrix

Author(s)

Koki Tsuyuzaki [aut, cre]

Maintainer: Koki Tsuyuzaki <[email protected]>

References

Wang B, et al., (2017). Vicus: Exploiting local structures to improve network-based analysis of biological data. PLOS Computational Biology. 13(10), e1005621

See Also

graphMatrix,embedding

Examples

ls("package:Vicus")

Graph Embedding

Description

Lower dimensions are estimated from the object returned from graphMatrix function.

Usage

embedding(obj)

Arguments

obj

Object returned from graphMatrix function.

Value

A matrix (n times k)

Author(s)

Koki Tsuyuzaki

See Also

graphMatrix

Examples

X <- matrix(runif(10*20), nrow=10, ncol=20)
head(embedding(graphMatrix(X, K=2)))

Graph Laplacian type matrix

Description

A symmetric matrix is returned.

Usage

graphMatrix(X, algorithm=c("Vicus", "LEM", "HLLE"),
  K=10, alpha=0.9, ndim=2)

Arguments

X

A numeric matrix (n times p).

algorithm

Algorithm to construct a graph matrix. Vicus matrix (Vicus), Graph Laplacian matrix (LEM), and Hessian Locally Linear Embedding matrix (HLLE) are available (Default: "Vicus").

K

The number of neighborhoods to construct a graph matrix (Default: 10).

alpha

An optional parameter for Vicus (Default: 0.9).

ndim

The number of lower dimension to embed the graph (Default: 2).

Value

M: A symmetric matrix (n times n). This matrix can be applied to embedding function. algorithm: algorithm parameter specified by argument ndim: ndim parameter specified by argument

Author(s)

Koki Tsuyuzaki

See Also

embedding

Examples

X <- matrix(runif(10*20), nrow=10, ncol=20)
str(graphMatrix(X, K=2), 2)