Abstract
Using a 1 / n expansion, that is an expansion in descending powers of n, for the number of matchings in regular graphs with 2n vertices, we study the monomer-dimer entropy for two classes of graphs. We study the difference between the extensive monomer-dimer entropy of a random r-regular graph G (bipartite or not) with 2n vertices and the average extensive entropy of r-regular graphs with 2n vertices, in the limit \(n \rightarrow \infty \). We find a series expansion for it in the numbers of cycles; with probability 1 it converges for dimer density \(p < 1\) and, for G bipartite, it diverges as \(|\mathrm{ln}(1-p)|\) for \(p \rightarrow 1\). In the case of regular lattices, we similarly expand the difference between the specific monomer-dimer entropy on a lattice and the one on the Bethe lattice; we write down its Taylor expansion in powers of p through the order 10, expressed in terms of the number of totally reducible walks which are not tree-like. We prove through order 6 that its expansion coefficients in powers of p are non-negative.
Similar content being viewed by others
References
Bollobas, B.: A probabilistic proof of an asymptotic formula for the number of labelled regular graphs. Eur. J. Combin. 1, 311–316 (1980)
Bollobas, B., McKay, B.D.: The number of matching in random regular graphs and bipartite graphs. J. Combin. Theory 41, 80–91 (1986)
Butera, P., Federbush, P., Pernici, M.: Higher-order expansions for the entropy of a dimer or a monomer-dimer system on \(d\)-dimensional lattices. Phys. Rev. E 87, 062113 (2013)
Butera, P., Pernici, M.: High-temperature expansions of the higher susceptibilities for the Ising model in general dimension \(d\). Phys. Rev. E 86, 01113 (2012)
Butera, P., Pernici, M.: Yang-Lee edge singularities from extended activity expansions of the dimer density for bipartite lattices of dimensionality \(2 \le d \le 7\). Phys. Rev. E86, 011104
Federbush, P.: A Mysterious Cluster Expansion Associated to the Expectation Value of the Permanent of 0-1 Matrices. arXiv:1607.00885 [math-ph]
Federbush, P., Friedland, S.: An asymptotic expansion and recursive inequalities for the monomer-dimer problem. J. Stat. Phys. 143, 306–325 (2011)
Fisher, M.E.: Statistical mechanics of dimers on a plane lattice. Phys. Rev. 124, 1664–1672 (1961)
Friedland, S., Krop, E., Markström, K.: On the number of matchings in regular graphs. Electron. J. Combin. 15, R110 (2008)
Gaunt, D.S.: Exact series-expansion study of the monomer-dimer problem. Phys. Rev. 179, 174–179 (1969)
Godsil, C.G., McKay, B.D.: Asymptotic enumeration of Latin rectangles. J. Combin. Theory B48, 19–44 (1990)
Heilmann, O.J., Lieb, E.H.: Theory of monomer-dimer systems. Commun. Math. Phys. 25, 190–232 (1972)
Kasteleyn, P.W.: The statistics of dimers on a lattice. Physica 27, 1209–1225 (1961)
Kurtze, D.A., Fisher, M.E.: Yang-Lee edge singularities at high temperatures. Phys. Rev. B 20, 2785–2796 (1979)
McKay, B.D.: The expected eigenvalue distribution of a large regular graph. Linear Algebra Appl. 40, 203–216 (1981)
Nagle, J.F.: New series-expansion method for the dimer problem. Phys. Rev. 152, 190–197 (1966)
SageMath: the Sage Mathematics Software System (Version 7.3). The Sage Developers (2016). http://www.sagemath.org
Wanless, I.M.: Counting matchings and tree-like walks in regular graphs. Combin. Probab. Comput. 19, 463–480 (2010)
Wormald, N.C.: Models of random regular graphs. In: Lamb, J.D., Preece, D.A. eds.: Surveys in Combinatorics 1999, LMS Lecture Note Series, vol. 267, pp. 239–298. Cambridge University Press, Cambridge (1999)
Zdeborova, L., Mezard, M.: The number of matchings in random graphs. J. Stat. Mech. 2006.05, P05003 (2006)
Acknowledgements
I thank Paolo Butera for discussions.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A: \(a_h\) Coefficients of \(M_j\)
Since \(a_k\) has degree at most 2k in j, from Eq. (18) it follows that \(q_k\) has degree at most \(k-1\). Let
Examining the expressions computed recursively for \(a_k\) for generic r through \(k=20\), we find that the four highest coefficients of \(q_k\) have the form
where \(\alpha \equiv \frac{1}{2r} - 1\) and
Equation (43) has been tested till \(k=100\) for \(r=1,\ldots , 10\) using \(a_k\) computed with Eq. (19).
Here we write down the first five \(q_k\).
Appendix B: Coefficients \(d_h\) for the Entropy for Regular Lattice
Let \(z \equiv \frac{1}{r}\). For notational simplicity, here \(\epsilon _s\) is the constant \(\hat{\epsilon }_s\) in the text.
The coefficient \(d_h\) in Eq. (40) are given, through order 10, by
Appendix C: Proof of a Conjecture in [6] in Some Cases
The \(n\times n\) reduced adjacency matrix A of a bipartite r-regular graph G with 2n vertices has entries 0 and 1, and the sums of the rows and of the columns equals r.
The uniform distribution of the labelled graphs corresponds to the expectation E.
The number or i-matchings on G is \(m_i(G) = {{\mathrm{perm}}}_i(A)\), defined as the sum of the permanents of the \(i\times i\) minors of A.
\(A_i \equiv E({{\mathrm{perm}}}_i(A))\) is the average number of i-matchings.
Another expectation \(E_2\) is defined by the second measure in Section 4 of [9] for the non negative integer matrices, with
It is the average number of regular bipartite multigraphs under the expectation \(E_2\).
In [6] the following quantity is defined
with \(\mathcal{C}_0=1\) and \(\mathcal{C}_1 =0\).
From this the following cluster expansion is constructed
Similarly using Eq. (49), the quantities \(C_i^{E_2}\) are computed in terms of \(A_i^{E_2}\); from this one gets \(T_i^{E_2}\) using Eq. (50).
In [6] it is conjectured that
Using Eqs. (48, 49, 50) one can compute the first few \(Q_i\). In [6] the conjecture has been proven for \(i \le 3\); some numerical evidence has been given for \(4 \le i \le 25\).
We have proven this conjecture for \(i \le 7\).
For \(i \le 7\), \(m_i\) is linear in the \(\epsilon \)’s. Since \(\epsilon _s\) is linear in the contributors, \(E(\epsilon _s)\) is linear in the average of the contributors, so it can be computed modulo \(o(n^0)\) using \(E(C_s) = \frac{1}{s}(r-1)^s + o(n^0)\) and the fact that the polycyclic contributors are \(o(n^0)\) [19]. For \(i \le 7\), \(\epsilon _i\) is given in Eq. (6).
It is easy to see from Eqs. (49, 50) that \(T_i\) is linear in the average of contributors; expanding \(T_i\) in 1 / n expansion, keeping the average of the contributors as independent variables, it has the form
where \(Q_i\) is the one found in [6], and \(B_i\) depends on r and is linear in the average of contributors, plus \(o(n^0)\) terms. For instance
It follows that \(\lim _{n \rightarrow \infty }B_i\) is a function of r for \(i \le 7\).
Rights and permissions
About this article
Cite this article
Pernici, M. 1 / n Expansion for the Number of Matchings on Regular Graphs and Monomer-Dimer Entropy. J Stat Phys 168, 666–679 (2017). https://doi.org/10.1007/s10955-017-1819-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10955-017-1819-6