Analyticity of the planar limit of a matrix model (Q1947175)

Matrix models are integrals of exponentiated potential functions over finite dimensional vector spaces. They come in two flavors: formal and analytic. When the potential is \(V_4=\frac 12 x^2-\frac 14 a_4 x^4\), \textit{D. Bessis} et al. [Adv. Appl. Math. 1, 109--157 (1980; Zbl 0453.05035)] gave an exact formula for the planar limit of the formal matrix model. Their method was justified by \textit{N. M. Ercolani} and \textit{K. D. T.-R. McLaughlin} [Int. Math. Res. Not. 2003, No. 14, 755--820 (2003; Zbl 1140.82307)] using potential theory and the Riemann-Hilbert method. Using Chebyshev polynomials combined with some mild combinatorics, the authors provide an alternative approach to the planar limits of the analytical and formal matrix models with a 1-cut potential. For potentials \(V=\frac 12 x^2-\sum_{n\geq 1}\frac 1n a_n x^n\), as a power series in all \(a_n\), the formal Taylor expansion of the analytic planar limit is exactly the formal planar limit. In the case \(V\) is analytic in infinitely many variables \(\{a_n \}_{n \geq 1}\), the planar limit is also an analytic function in infinitely many variables. These results confirm a conjecture of 't Hooft, which states that if the potential is analytic, the planar limit is also analytic. In enumerative combinatorics, two gradings of \(V\) are useful: the edge grading \(V_e=\frac 12 x^2-\sum_{n\geq 1}\frac 1n a_n t^{\frac{n}{2}} x^n\) and the face grading \(V_f=\frac 12 x^2-\sum_{n\geq 3}\frac 1n a_n t^{\frac{n}{2}-1} x^n\). The associated planar limits as functions of \(t\) count planar diagram sorted by the number of edges respectively faces. The authors point out a method of computing the asymptotic of the coefficients of the associated planar limits using the combination of the \(wzb\) method and the resolution of singularities, reestablishing some results of \textit{P. M. Bleher} and \textit{A. R. Its} [Ann. Inst. Fourier 55, No. 6, 1943--2000 (2005; Zbl 1135.82016)].