Description Usage Arguments Details Value Author(s) References Examples

Computation of different embeddings from a list of directed graphs

1 |

`gList` |
A |

`method` |
Embedding method. This should be (an unambiguous
abbreviation of) one of |

`groups` |
A named vector of class |

This function computes a graph embedding from a list of graph.

Comparing the architecture of interaction networks in space or time is essential for understanding the assembly, trajectory, functioning and persistence of species communities. Graph embedding methods, which position networks into a vector space where nearby networks have similar architectures, could be ideal tools for this purposes.

Here, the proposed embedding methods are:

-`"metric2vec"`

: each column is a network
metric. `connectance`

: network connectance,
`modularity`

: network modularity (undirected graph),
`rangeTL`

: trophic level range, (trophic level is computed as in Williams & Martinez, 2004),
`meanTL`

: mean trophic level,
`SWTL`

: shortest weighted trophic level,
`meanOmni`

: mean level of omnivory (standard deviation of the
trophic levels of the prey),
`propOmni`

: proportion of omnivorous species,
`propCanib`

: proportion of cannibal species,
`meanNbPrey`

: mean number of prey,
`sdNbPrey`

: standard deviation of the number of prey,
`skewNbPrey`

: skewness of the number of prey,
`meanNbPred`

: mean number of predators,
`sdNbPred`

: standard deviation of the number of predators,
`skewNbPred`

: skewness of the number of predators,
`propBasal`

: proportion of basal species (which have no prey),
`propTop`

: proportion of top species (which have no predator),
`propInter`

: proportion of intermediate species (which have prey and predator),
`sdVulnerability`

: standard deviation of the vulnerability (i.e. in-degree),
`sdGeneralism`

: standard deviation of the generalism (i.e. out-degree),
`transitivity`

: network transitivity,
`diameter`

: network diameter,
`meanSP`

: mean shortest path,
`assortativity`

: network assortativity.

-`"motif2vec"`

: Graph motifs are small connected subgraphs with a well-defined
structure. Each column is the proportion of each of the 13
connected motifs as enumerated by the 'motifs' function.

-`"group2vec"`

: Groups proportion, where the groups are defined
previously (e.g. taxonmic groups).

-`"shortestpath2vec"`

: Computing the proportion of shortest path
lengths. Can use the groups as well, path length proportion is decomposed
per source and target group in this case.

`embedding`

returns an object of class `matrix`

where rows
corresponds to the differents graphs and columns to the different
dimensions in the embedding space.

Authors: Christophe Botella, Stephane Dray, Catherine Matias, Vincent Miele, Wilfried Thuiller

Maintainer: Vincent Miele <vincent.miele@univ-lyon1.fr>

Christophe Botella, Stephane Dray, Catherine Matias, Vincent Miele & Wilfried Thuiller, An appraisal of graph embeddings for comparing trophic network architectures. Methods in Ecology and evolution (2021) <doi:10.1111/2041-210X.13738>

Williams, Richard J., and Neo D. Martinez. "Limits to trophic levels and omnivory in complex food webs: theory and data." The American Naturalist 163.3 (2004)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 | ```
library(igraph)
# Generating a set of graphs with 3 groups
groups = rep(-1,60)
groups[1:20] = 1
groups[21:40] = 2
groups[41:60] = 3
names(groups) = as.character(1:60)
# A first set with a few links between groups 1 and 3
gList = c()
for(i in 1:5){
pm <- rbind( c(0, .5, 0.05), c(0, 0, 0.5), c(0,0,0) )
graphLoc <- sample_sbm(60, pref.matrix=pm, block.sizes=c(20,20,20), directed=TRUE)
V(graphLoc)$name = as.character(1:60)
gList = c(gList,list(graphLoc))
}
# A second set with many links between groups 1 and 3
for(i in (6:10)){
pm <- rbind( c(0, .5, .5), c(0, 0, 0.5), c(0,0,0) )
graphLoc <- sample_sbm(60, pref.matrix=pm, block.sizes=c(20,20,20), directed=TRUE)
V(graphLoc)$name = as.character(1:60)
gList = c(gList,list(graphLoc))
}
names(gList) = LETTERS[1:10]
# Computing different embeddings and
par(mfrow=c(2,2))
embd <- embedding(gList, method="metric2vec")
pca <- prcomp(embd, rank=2)$x
plot(pca, main="metric2vec")
text(x=pca[,1], y=pca[,2], labels=rownames(embd), pos=2, col=c(rep("blue",5),rep("red",5)))
embd <- embedding(gList, method="motif2vec")
pca <- prcomp(embd, rank=2)$x
plot(pca, main="motif2vec")
text(x=pca[,1], y=pca[,2], labels=rownames(embd), pos=2, col=c(rep("blue",5),rep("red",5)))
embd <- embedding(gList, method="shortestpath2vec")
pca <- prcomp(embd, rank=2)$x
plot(pca, main="shortestpath2vec")
text(x=pca[,1], y=pca[,2], labels=rownames(embd), pos=2, col=c(rep("blue",5),rep("red",5)))
embd <- embedding(gList, method="shortestpath2vec", groups)
pca <- prcomp(embd, rank=2)$x
plot(pca, main="shortestpath2vec_group")
text(x=pca[,1], y=pca[,2], labels=rownames(embd), pos=2, col=c(rep("blue",5),rep("red",5)))
``` |

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.