2018
DOI: 10.1093/comnet/cny008
|View full text |Cite
|
Sign up to set email alerts
|

The connectivity of graphs of graphs with self-loops and a given degree sequence

Abstract: Double edge swaps' transform one graph into another while preserving the graph's degree sequence, and have thus been used in a number of popular Markov chain Monte Carlo (MCMC) sampling techniques. However, while double edge-swaps can transform, for any fixed degree sequence, any two graphs inside the classes of simple graphs, multigraphs, and pseudographs, this is not true for graphs which allow self-loops but not multiedges (loopy graphs). Indeed, we exactly characterize the degree sequences where double edg… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1

Citation Types

0
15
0

Year Published

2018
2018
2024
2024

Publication Types

Select...
5
2

Relationship

1
6

Authors

Journals

citations
Cited by 7 publications
(15 citation statements)
references
References 22 publications
0
15
0
Order By: Relevance
“…Thus the two graphs in the space are not connected by any sequence of double edge swaps that remain in the space of loopy graphs, and this lack of connectivity applies to both the stub-and vertex-labeled spaces. Generalizing this observation, it is the case that the space of loopy graphs is connected for any degree sequence that can wire a simple graph and is neither the degree sequence of a path, {2, 2, ..., 2}, nor that of a clique, {n − 1, n − 1, ..., n − 1} [111]. Alternatively, if the Markov chain is modified to occasionally employ a three-edge triangle-loop swap (the swap (u, u), (w, w)), a basic modification of Algorithm 1 and Algorithm 2 suffices to sample uniformly from these spaces; see [111] for more details.…”
Section: Markov Chains For Sampling Other Spacesmentioning
confidence: 89%
See 1 more Smart Citation
“…Thus the two graphs in the space are not connected by any sequence of double edge swaps that remain in the space of loopy graphs, and this lack of connectivity applies to both the stub-and vertex-labeled spaces. Generalizing this observation, it is the case that the space of loopy graphs is connected for any degree sequence that can wire a simple graph and is neither the degree sequence of a path, {2, 2, ..., 2}, nor that of a clique, {n − 1, n − 1, ..., n − 1} [111]. Alternatively, if the Markov chain is modified to occasionally employ a three-edge triangle-loop swap (the swap (u, u), (w, w)), a basic modification of Algorithm 1 and Algorithm 2 suffices to sample uniformly from these spaces; see [111] for more details.…”
Section: Markov Chains For Sampling Other Spacesmentioning
confidence: 89%
“…Although the space of loopy graphs is not necessarily connected under double-edge swaps, it can be shown to be connected for the degree sequence of this collaboration network, allowing the use of Algorithm 2 or 3 from Section 2; see[111] for details.…”
mentioning
confidence: 99%
“…We denote by G ⟳ the set of all graphs on n nodes, and by G ⊂ G ⟳ the set of graphs on n nodes without self-loops. Parallel edges are permitted in G. While it is indeed possible to define configuration models on G ⟳ , we do not do so here [16,39]. Rather we will formulate most of our results for elements of G, only discussing G ⟳ below in the context of certain technical issues.…”
Section: Definition 1 (Graph)mentioning
confidence: 99%
“…An alternative solution to drawing from the other seven graph space, e.g., that of simple graphs, is to sample them using a Markov chain Monte Carlo (MCMC) algorithm based on double-edge swaps. Although a number of such algorithms have been defined, the Fosdick et al MCMC is known to asymptotically converge on the target uniform distribution for each of the eight graph spaces (with rare exceptions in graph spaces that allow loops but not multi-edges [18]). However, practical guidance on the time required for convergence with finite-sized networks remains unknown.…”
mentioning
confidence: 99%