Configuring random graph models with fixed degree sequences

Bailey K. Fosdick, Daniel B. Larremore, Joel Nishimura, Johan Ugander

Research output: Contribution to journalReview article

21 Citations (Scopus)

Abstract

Random graph null models have found widespread application in diverse research communities analyzing network datasets, including social, information, and economic networks, as well as food webs, protein-protein interactions, and neuronal networks. The most popular random graph null models, called configuration models, are defined as uniform distributions over a space of graphs with a fixed degree sequence. Commonly, properties of an empirical network are compared to properties of an ensemble of graphs from a configuration model in order to quantify whether empirical network properties are meaningful or whether they are instead a common consequence of the particular degree sequence. In this work we study the subtle but important decisions underlying the specification of a configuration model, and we investigate the role these choices play in graph sampling procedures and a suite of applications. We place particular emphasis on the importance of specifying the appropriate graph labeling-stub-labeled or vertex-labeled-under which to consider a null model, a choice that closely connects the study of random graphs to the study of random contingency tables. We show that the choice of graph labeling is inconsequential for studies of simple graphs, but can have a significant impact on analyses of multigraphs or graphs with self-loops. The importance of these choices is demonstrated through a series of three in-depth vignettes, analyzing three different network datasets under many different configuration models and observing substantial differences in study conclusions under different models. We argue that in each case, only one of the possible configuration models is appropriate. While our work focuses on undirected static networks, it aims to guide the study of directed networks, dynamic networks, and all other network contexts that are suitably studied through the lens of random graph null models.

Original languageEnglish (US)
Pages (from-to)315-355
Number of pages41
JournalSIAM Review
Volume60
Issue number2
DOIs
StatePublished - Jan 1 2018

Fingerprint

Degree Sequence
Graph Model
Random Graphs
Null
Configuration
Graph Labeling
Model
Graph in graph theory
Labeling
Proteins
Food Web
Neuronal Network
Directed Network
Network Dynamics
Multigraph
Protein Interaction Networks
Dynamic Networks
Contingency Table
Protein-protein Interaction
Simple Graph

Keywords

  • Complex networks
  • Configuration model
  • Graph enumeration
  • Graph sampling
  • Markov chain monte carlo
  • Permutation tests

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computational Mathematics
  • Applied Mathematics

Cite this

Configuring random graph models with fixed degree sequences. / Fosdick, Bailey K.; Larremore, Daniel B.; Nishimura, Joel; Ugander, Johan.

In: SIAM Review, Vol. 60, No. 2, 01.01.2018, p. 315-355.

Research output: Contribution to journalReview article

Fosdick, BK, Larremore, DB, Nishimura, J & Ugander, J 2018, 'Configuring random graph models with fixed degree sequences', SIAM Review, vol. 60, no. 2, pp. 315-355. https://doi.org/10.1137/16M1087175
Fosdick, Bailey K. ; Larremore, Daniel B. ; Nishimura, Joel ; Ugander, Johan. / Configuring random graph models with fixed degree sequences. In: SIAM Review. 2018 ; Vol. 60, No. 2. pp. 315-355.
@article{ff70c3074d2b41158ba56191ed44a1d0,
title = "Configuring random graph models with fixed degree sequences",
abstract = "Random graph null models have found widespread application in diverse research communities analyzing network datasets, including social, information, and economic networks, as well as food webs, protein-protein interactions, and neuronal networks. The most popular random graph null models, called configuration models, are defined as uniform distributions over a space of graphs with a fixed degree sequence. Commonly, properties of an empirical network are compared to properties of an ensemble of graphs from a configuration model in order to quantify whether empirical network properties are meaningful or whether they are instead a common consequence of the particular degree sequence. In this work we study the subtle but important decisions underlying the specification of a configuration model, and we investigate the role these choices play in graph sampling procedures and a suite of applications. We place particular emphasis on the importance of specifying the appropriate graph labeling-stub-labeled or vertex-labeled-under which to consider a null model, a choice that closely connects the study of random graphs to the study of random contingency tables. We show that the choice of graph labeling is inconsequential for studies of simple graphs, but can have a significant impact on analyses of multigraphs or graphs with self-loops. The importance of these choices is demonstrated through a series of three in-depth vignettes, analyzing three different network datasets under many different configuration models and observing substantial differences in study conclusions under different models. We argue that in each case, only one of the possible configuration models is appropriate. While our work focuses on undirected static networks, it aims to guide the study of directed networks, dynamic networks, and all other network contexts that are suitably studied through the lens of random graph null models.",
keywords = "Complex networks, Configuration model, Graph enumeration, Graph sampling, Markov chain monte carlo, Permutation tests",
author = "Fosdick, {Bailey K.} and Larremore, {Daniel B.} and Joel Nishimura and Johan Ugander",
year = "2018",
month = "1",
day = "1",
doi = "10.1137/16M1087175",
language = "English (US)",
volume = "60",
pages = "315--355",
journal = "SIAM Review",
issn = "0036-1445",
publisher = "Society for Industrial and Applied Mathematics Publications",
number = "2",

}

TY - JOUR

T1 - Configuring random graph models with fixed degree sequences

AU - Fosdick, Bailey K.

AU - Larremore, Daniel B.

AU - Nishimura, Joel

AU - Ugander, Johan

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Random graph null models have found widespread application in diverse research communities analyzing network datasets, including social, information, and economic networks, as well as food webs, protein-protein interactions, and neuronal networks. The most popular random graph null models, called configuration models, are defined as uniform distributions over a space of graphs with a fixed degree sequence. Commonly, properties of an empirical network are compared to properties of an ensemble of graphs from a configuration model in order to quantify whether empirical network properties are meaningful or whether they are instead a common consequence of the particular degree sequence. In this work we study the subtle but important decisions underlying the specification of a configuration model, and we investigate the role these choices play in graph sampling procedures and a suite of applications. We place particular emphasis on the importance of specifying the appropriate graph labeling-stub-labeled or vertex-labeled-under which to consider a null model, a choice that closely connects the study of random graphs to the study of random contingency tables. We show that the choice of graph labeling is inconsequential for studies of simple graphs, but can have a significant impact on analyses of multigraphs or graphs with self-loops. The importance of these choices is demonstrated through a series of three in-depth vignettes, analyzing three different network datasets under many different configuration models and observing substantial differences in study conclusions under different models. We argue that in each case, only one of the possible configuration models is appropriate. While our work focuses on undirected static networks, it aims to guide the study of directed networks, dynamic networks, and all other network contexts that are suitably studied through the lens of random graph null models.

AB - Random graph null models have found widespread application in diverse research communities analyzing network datasets, including social, information, and economic networks, as well as food webs, protein-protein interactions, and neuronal networks. The most popular random graph null models, called configuration models, are defined as uniform distributions over a space of graphs with a fixed degree sequence. Commonly, properties of an empirical network are compared to properties of an ensemble of graphs from a configuration model in order to quantify whether empirical network properties are meaningful or whether they are instead a common consequence of the particular degree sequence. In this work we study the subtle but important decisions underlying the specification of a configuration model, and we investigate the role these choices play in graph sampling procedures and a suite of applications. We place particular emphasis on the importance of specifying the appropriate graph labeling-stub-labeled or vertex-labeled-under which to consider a null model, a choice that closely connects the study of random graphs to the study of random contingency tables. We show that the choice of graph labeling is inconsequential for studies of simple graphs, but can have a significant impact on analyses of multigraphs or graphs with self-loops. The importance of these choices is demonstrated through a series of three in-depth vignettes, analyzing three different network datasets under many different configuration models and observing substantial differences in study conclusions under different models. We argue that in each case, only one of the possible configuration models is appropriate. While our work focuses on undirected static networks, it aims to guide the study of directed networks, dynamic networks, and all other network contexts that are suitably studied through the lens of random graph null models.

KW - Complex networks

KW - Configuration model

KW - Graph enumeration

KW - Graph sampling

KW - Markov chain monte carlo

KW - Permutation tests

UR - http://www.scopus.com/inward/record.url?scp=85046685956&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046685956&partnerID=8YFLogxK

U2 - 10.1137/16M1087175

DO - 10.1137/16M1087175

M3 - Review article

AN - SCOPUS:85046685956

VL - 60

SP - 315

EP - 355

JO - SIAM Review

JF - SIAM Review

SN - 0036-1445

IS - 2

ER -