Discrete Entropy of Generalized Jacobi Polynomials

preprint2015arXivOpen access

Abstract

Given a sequence of orthonormal polynomials on R\Bbb R,{pn}n≥0\{p_n\}_{n\geq 0}, with pnp_n of degree nn, we define the discrete probability distribution Ψn(x)=(Ψn,1(x),…Ψn,n(x))Ψ_n(x) = \left(Ψ_{n,1}(x), \dots Ψ_{n,n}(x) \right) , with Ψn,j(x)=(∑j=0n−1pj2(x))−1pj−12(x)Ψ_{n,j}(x) = \big(\sum_{j=0}^{n-1} p_j^2(x)\big)^{-1} p_{j-1}^2(x), j=1,…,nj=1, \dots, n. In this paper, we study the asymptotic behavior as n→∞n\to \infty of the Shannon entropy S((Ψn(x))=−∑j=1nΨn,j(x)log⁡(Ψn,j(x))\mathcal S ((Ψ_n(x))= -\sum_{j=1}^n Ψ_{n,j}(x) \log (Ψ_{n,j}(x)), x∈(−1,1)x\in (-1,1), when the orthogonality weight is (1−x)α (1+x)β h(x) (1-x)^α\, (1+x)^β\, h(x) , α,β>−1α, β> -1, and where hh is real, analytic, and positive on [−1,1][-1,1]. We show that the limit lim⁡n→∞(S((Ψn(x))−log⁡n) \lim_{n \to \infty} \left(\mathcal{S} ((Ψ_n(x))- \log n\right) exists for all x∈(−1,1)x\in (-1,1), but its value depends on the rationality of arccos⁡(x)/π\arccos(x)/π. For the particular case of the Chebyshev polynomials of the first and second kinds, we compare our asymptotic result with the explicit formulas for S(Ψn(ζj(n)))\mathcal{S} (Ψ_n(ζ_j^{(n)})), where {ζj(n)}\{ζ_j^{(n)}\} are the zeros of pnp_n, obtained previously in [A.I. Aptekarev, J.S. Dehesa, A. Martinez-Finkelshtein, and R. Yañez, Constr. Approx., 30 (2009), pp. 93-119].

Explore this paper’s authors, topics and related work.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Reviews 0

Write a reviewWrite

No reviews yet.

Discussion 0

Add a commentWrite

No comments yet.