# Groupies in multitype random graphs

- Yilun Shang
^{1}Email authorView ORCID ID profile

**Received: **5 April 2016

**Accepted: **28 June 2016

**Published: **7 July 2016

## Abstract

A groupie in a graph is a vertex whose degree is not less than the average degree of its neighbors. Under some mild conditions, we show that the proportion of groupies is very close to 1/2 in multitype random graphs (such as stochastic block models), which include Erdős-Rényi random graphs, random bipartite, and multipartite graphs as special examples. Numerical examples are provided to illustrate the theoretical results.

### Keywords

Random graph Degree Groupie Multitype### Mathematics Subject Classification

05C07 05C80## Background

A vertex in a graph *G* is said to be a *groupie* if its degree is not less than the average degree of its neighbors. Various properties of groupies have been investigated in deterministic graph theory (Ajtai et al. 1980; Bertram et al. 1994; Ho 2007; Mackey 1996; Poljak et al. 1995). For example, it was proved in Mackey (1996) that there are at least two groupies in any simple graphs with at least two vertices. Groupies were even found to be related to Ramsey numbers (Ajtai et al. 1980). More recently, Fernandez de la Vega and Tuza (2009) showed that, in Erdős-Rényi random graphs *G*(*n*, *p*), the proportion of vertices that are groupies is almost always very near to 1/2 as \(n\rightarrow \infty \). Later the author Shang (2010) obtained a result of similar flavor in random bipartite graphs \(G(n_1,n_2,p)\). It was shown that the proportion of groupies in each partite set is almost always very close to 1/2 if \(G(n_1,n_2,p)\) is balanced, namely, \(n_1=n_2\).

In this paper, we consider groupies in a more general random graph model, which we call *multitype random graphs*. Let *q* be a positive integer. Denote \([q]:=\{1,\ldots ,q\}\). Define the ‘gene’ for a multitype random graph as a weighted complete graph \(K_q\) (having a loop at each vertex) on the vertex set [*q*], with a weight \(\alpha _i>0\) associated to each vertex, and a weight \(0\le \beta _{ij}\le 1\) associate to each edge *ij*. Note that \(\beta _{ij}=\beta _{ji}\) since we deal with undirected graphs. We assume \(\sum _{i=1}^q\alpha _i=1\). The multitype random graph \(G(n,K_q)\) with gene \(K_q\) is generated as follows. Let *n* be much larger than *q*, and let [*n*] be its vertex set. We partition [*n*] into *q* sets \(V_1,\ldots ,V_q\) by putting vertex *v* in \(V_i\) with probability \(\alpha _i\) independently. Each pair of vertices \(v\in V_i\) and \(u\in V_j\) are connected with probability \(\beta _{ij}\) independently (all the decisions on vertices and edges are made independently).

For \(i=1,\ldots ,q\), let \(N_i\) represent the number of the groupies in \(V_i\). Thus, \(N:=\sum _{i=1}^qN_i\) is the number of groupies in the multitype random graph \(G(n,K_q)\). Denote by \(\alpha =(\alpha _i)\in \mathbb {R}^q\) and \(\beta =(\beta _{ij})\in \mathbb {R}^{q\times q}\). For generality, we will usually think of \(\beta \) and \(\alpha \) as functions of *n* in the same spirit of random graph theory (Bollobás 2001; Janson et al. 2000). Let \(\mathbf{1}=(1,\ldots ,1)^T\in \mathbb {R}^q\) be the all-one vector. All the asymptotic notations used in the paper such as *O*, *o*, and \(\Omega \) are standard, see e.g. Janson et al. (2000). Our first result is as follows.

###
**Theorem 1**

*Let*\(q\ge 2\).

*Assume that*\(\beta \alpha =(\theta +o(\sqrt{\ln n}/n))\mathbf{1}\),

*where*\(\theta >0\)

*is a constant. If*\(\min _{i\not =j}\{\alpha _i,\beta _{ij}\}>c\)

*for some constant*\(c>0\),

*and*\(\max _{i}\{\beta _{ii}\}=o(\sqrt{\ln n}/n)\),

*then*

*as*\(n\rightarrow \infty \),

*where*\(\omega (n)=\Omega (\ln n)\)

*is any function tending to infinity. Hence,*

*as*\(n\rightarrow \infty \),

*where*\(\omega (n)=\Omega (\ln n)\)

*is any function tending to infinity.*

When \(\beta \) and \(\alpha \) are independent of *n*, the following corollary is immediate.

###
**Corollary 1**

*Let*\(q\ge 2\).

*Assume that*\(\beta \alpha =\theta \mathbf{1}\)

*for*\(\theta >0\),

*and*\(\beta _{ii}=0\)

*for all*

*i*.

*Then*

*as*\(n\rightarrow \infty \),

*where*\(\omega (n)=\Omega (\ln n)\)

*is any function tending to infinity.*

Clearly, by taking \(q=2\), \(\alpha _1=\alpha _2=1/2\), and \(\beta _{11}=\beta _{22}=0\), we recover the result in Shang (2010, Thm. 1) for balanced random bipartite graphs.

Theorem 1 requires that the edges between sets \(V_i\), \(i=1,\ldots ,q\) are dense, namely, the multitype random graph \(G(n,K_q)\) in question resembles a dense ‘multipartite’ graph. For sparse random graphs on the other hand, we have the following result.

###
**Theorem 2**

*Let*\(q\ge 1\).

*Assume that*\(\beta \alpha =(\theta +o(\sqrt{\ln n}/n))\mathbf{1}\),

*where*\(\theta =\theta (n)\)

*is a function of*

*n*.

*If*\(\min _{i}\{\alpha _i\}>c\)

*for some constant*\(c>0\), \(\min _{i\not =j}\{\beta _{ij}\}\gg (\ln n)^2/n \),

*and*\(\max _{i}\{\beta _{ii}\}=o(\sqrt{\ln n}/n)\),

*then*

*as*\(n\rightarrow \infty \),

*where*\(\varepsilon (n)=\Omega (\ln n/\sqrt{n})\)

*is any function tending to zero. Hence,*

*as*\(n\rightarrow \infty \),

*where*\(\varepsilon (n)=\Omega (\ln n/\sqrt{n})\)

*is any function tending to zero.*

It follows from Theorem 2 that we may reproduce the result for sparse Erdős-Rényi random graphs Fernandez de la Vega and Tuza (2009, Thm. 2) by taking \(q=\alpha _1=1\), \(\beta _{11}=o(\sqrt{\ln n}/n)\); and the result for sparse balanced random bipartite graphs Shang (2010, Thm. 2) by taking \(q=2\), \(\alpha _1=\alpha _2=1/2\), \(\beta _{11}=\beta _{22}=0\) and \(\beta _{12}\gg (\ln n)^2/n\).

The multitype random graph \(G(n,K_q)\) is generated through a double random process. In the following, we will also consider a closely related ‘random-free’ model \(G'(n,K_q)\). Given a gene \(K_q\) defined as above, the *random-free multitype random graph*
\(G'(n,K_q)\) (a.k.a. stochastic block model Holland et al. 1983) is constructed by partitioning [*n*] into *q* sets \(V_1,\ldots ,V_q\) with \(|V_i|=\alpha _in\). Recall that \(\sum _{i=1}^q\alpha _i=1\). We draw an edge *vu* with probability \(\beta _{ij}\) independently for \(v\in V_i\) and \(u\in V_j\); thus the first random step in the original construction disappears, which explains the name ‘random-free’.

In "Proof of the main results" section, we will show Theorems 1 and 2 by first proving analogous results for the random-free version \(G'(n,K_q)\). To illustrate our theoretical results, a numerical example is presented in "Numerical simulations" section.

## Proof of the main results

###
**Proposition 1**

*Theorem* 1
*holds verbatim for the random-free model*
\(G'(n,K_q)\).

###
*Proof*

*i*being completely similar. Take vertex \(v\in V_1\) and denote by \(d_v\) the degree of

*v*in \(G'(n,K_q)\). Therefore, \(d_v=\sum _{i=1}^qd_i\), where \(d_i\) means the number of neighbors of

*v*in \(V_i\). Let \(S_v\) represent the sum of degrees of the neighbors of

*v*. Write \({\text {Bin}}(n,p)\) for a Binomial variable with parameters

*n*and

*p*. Assuming that

*v*has degree \(d_v\), we obtain

###
**Proposition 2**

*Theorem* 2
*holds verbatim for the random-free model*
\(G'(n,K_q)\), *except that we herein allow*
\(\varepsilon (n)=\Omega (\sqrt{\ln n/n})\)
*as any function tending to zero.*

###
*Proof*

## Numerical simulations

To illustrate our theoretical results, in this section we present a numerical example for the \(G(n,K_q)\) model with \(q=3\).

*n*, (i) with the above constant \(\beta \); and (ii) with perturbed \(\beta +\Delta \beta \), where \(\Delta \beta =(\ln ^{1/4}n)/n\mathbf{1}\mathbf{1}^T\). Clearly, the conditions in Theorem 1 hold for both situations (i) and (ii). Fig. 1 shows that the agreement between the simulations and the theoretical prediction of Theorem 1 is excellent.

## Conclusion

In this paper, we have studied the groupies in multitype random graphs. It is discovered that the proportion of groupies is very close to 1/2 in multitype random graphs, which include Erdős-Rényi random graphs, random bipartite, and multipartite graphs as special examples. We mention that there are several possibilities to continue this line of research, both by considering other more realistic random network models as well as by analyzing the limit distribution of groupies in random graphs. For example, a natural question could be to ask if there are similar results for \(q=q(n)\) or edge-independent random graphs (e.g. Shang 2016)?

## Declarations

### Acknowledgements

The author is thankful to the reviewers for careful reading and constructive suggestions. The work is supported in part by the National Natural Science Foundation of China (11505127), the Shanghai Pujiang Program (15PJ1408300), and the Program for Young Excellent Talents in Tongji University (2014KJ036).

### Competing interests

The author declares that he has no competing interests.

**Open Access**This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

## Authors’ Affiliations

## References

- Ajtai M, Komlós J, Szemerédi E (1980) A note on Ramsey numbers. J Comb Theory Ser A 29:354–360View ArticleGoogle Scholar
- Bertram E, Erdős P, Horák P, Širáň J, Tuza Zs (1994) Local and global average degree in graphs and multigraphs. J Graph Theory 18:647–661View ArticleGoogle Scholar
- Bollobás B (2001) Random graphs. Cambridge University Press, CambridgeView ArticleGoogle Scholar
- Butler K, Stephens M (1993) The distribution of a sum of binomial random variables. Technical Report No. 467. Department of Statistics, Stanford UniversityGoogle Scholar
- Drezner Z, Farnum N (2007) A generalized binomial distribution. Commun Stat Theory Methods 22:3051–3063View ArticleGoogle Scholar
- Fernandez de la Vega W, Tuza Zs (2009) Groupies in random graphs. Inform Process Lett 109:339–340View ArticleGoogle Scholar
- Ho PT (2007) On groupies in graphs. Aust J Combin 38:173–177Google Scholar
- Holland PW, Laskey KB, Leinhardt S (1983) Stochastic blockmodels: first steps. Soc Netw 5:109–137View ArticleGoogle Scholar
- Janson S, Luczak T, Ruciński A (2000) Random graphs. Wiley, New YorkView ArticleGoogle Scholar
- Mackey J (1996) A lower bound for groupies in graphs. J. Graph Theory 21:323–326View ArticleGoogle Scholar
- Pólya G, Szegö G (1972) Problems and theorems in analysis, vol I: series, integral calculus, theory of functions. Translated from the German by D. Aeppli Die Grundlehren der mathematischen Wissenschaften, Band 193. Springer, New YorkGoogle Scholar
- Poljak S, Szabó T, Tuza Zs (1995) Extremum and convergence of local average degrees in graphs. Congr Numer 112:191–198Google Scholar
- Shang Y (2010) Groupies in random bipartite graphs. Appl Anal Discrete Math 4:278–283View ArticleGoogle Scholar
- Shang Y (2016) Bounding extremal degrees of edge-independent random graphs using relative entropy. Entropy 18:art. 53Google Scholar