Asymptotic behavior of spacing distributions for the eigenvalues of random matrices
J. des Cloizeaux, M. L. Mehta
Abstract
J. des Cloizeaux, M. L. Mehta
Abstract
It is known that the probability Eβ(0, S) that an arbitrary interval of length S contains none of the eigenvalues of a random matrix chosen from the orthogonal (β = 1), unitary (β = 2) or symplectic (β = 4) ensemble can be expressed in terms if infinite products ∏n=0∞[1−λ2n(S)] and ∏n=0∞[1=λ2n+1(S)], where λn(S) is an eigenvalue of a certain integral equation. Using values of λn(S), valid for S large, obtained in connection with a recent study of spheroidal functions, we derive asymptotic expressions (S ≫ 1) for Eβ(0, S).
OpenAlex reports 60 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
It is known that the probability Eβ(0, S) that an arbitrary interval of length S contains none of the eigenvalues of a random matrix chosen from the orthogonal (β = 1), unitary (β = 2) or symplectic (β = 4) ensemble can be expressed in terms if infinite products ∏n=0∞[1−λ2n(S)] and ∏n=0∞[1=λ2n+1(S)], where λn(S) is an eigenvalue of a certain integral equation. Using values of λn(S), valid for S large, obtained in connection with a recent study of spheroidal functions, we derive asymptotic expressions (S ≫ 1) for Eβ(0, S).
Key concepts: Eigenvalues and eigenvectors, Mathematics, Connection (principal bundle), Random matrix, Symplectic geometry, Matrix (chemical analysis), Unitary state, Circular ensemble