inducedSubgraphByPkgs: Return a minimal subgraph based on package name(s)

In seandavi/BiocPkgTools: Collection of simple tools for learning about Bioconductor Packages

Return a minimal subgraph based on package name(s)

Description

Find the subgraph induced by including specific packages. The induced subgraph is the graph that includes the named packages and all edges connecting them. This is useful for a developer, for example, to examine her packages and their intervening dependencies.

Usage

inducedSubgraphByPkgs(g, pkgs, pkg_color = "red")

Arguments

g	an igraph graph, typically created by buildPkgDependencyIgraph
pkgs	character() vector of packages to include. Package names not included in the graph are ignored.
pkg_color	character(1) giving color of named packages. Other packages in the graph that fall in connecting paths will be colored as the igraph default.

Examples

library(igraph)
g <- buildPkgDependencyIgraph(buildPkgDependencyDataFrame())
## subgraph of only the first 10 packages maintained by Bioconductor
biocmaintained <- head(biocMaintained()[["Package"]], 10L)
g2 <- inducedSubgraphByPkgs(g, pkgs = biocmaintained)
g2
V(g2)

plot(g2)

## subgraph of a package's strong Bioconductor package dependencies
maedeps <- unlist(pkgBiocDeps(
    "MultiAssayExperiment", which = "strong",
    recursive = TRUE, only.bioc = TRUE
), use.names = FALSE)
g3 <- inducedSubgraphByPkgs(g, pkgs = maedeps)
plot(g3)

## same subgraph with networkD3::forceNetwork
library(networkD3)
wt <- cluster_walktrap(g3)
members <- membership(wt)
ndg3 <- igraph_to_networkD3(g3, group = members)
forceNetwork(
    Links = ndg3$links, Nodes = ndg3$nodes, Source = 'source',
    Target = 'target', NodeID = 'name', Group = 'group', zoom = TRUE,
    linkDistance = 200, fontSize = 20, opacity = 0.9, opacityNoHover = 0.9
)

seandavi/BiocPkgTools documentation built on May 23, 2024, 1:53 p.m.