Gauss-Kuzmin-Wirsing operator

From Wikipedia, the free encyclopedia

"GKW" redirects here. For the Indian engineering firm see Guest Keen Williams.

In mathematics, the Gauss-Kuzmin-Wirsing operator occurs in the study of continued fractions; it is also related to the Riemann zeta function.

1 Introduction
2 Relationship to the Riemann zeta
3 Matrix elements
4 Riemann zeta
5 References

[edit] Introduction

The Gauss-Kuzmin-Wirsing operator is the transfer operator of the Gauss map

$h(x)=1/x-\lfloor 1/x \rfloor.\,$

This operator acts on functions as

$[Gf](x) = \sum_{n=1}^\infty \frac {1}{(x+n)^2} f \left(\frac {1}{x+n}\right).$

The zeroth eigenfunction of this operator is

$\frac {\ln 2} {1+x}$

which corresponds to an eigenvalue of 1. This eigenfunction gives the probability of the occurrence of a given integer in a continued fraction expansion, and is known as the Gauss-Kuzmin distribution. This follows in part because the Gauss map acts as a truncating shift operator for the continued fractions: if $x=[0;a_1,a_2,a_3,\dots]$ is the continued fraction representation of a number 0 < x < 1, then $h(x)=[0;a_2,a_3,\dots]$ .

Additional eigenvalues can be computed numerically; the next eigenvalue is $λ 1 = 0.3036630029...$ and is known as the Gauss-Kuzmin-Wirsing constant. Analytic forms for additional eigenfunctions are not known. It is not known if the eigenvalues are irrational.

[edit] Relationship to the Riemann zeta

The GKW operator is related to the Riemann zeta function. Note that the zeta can be written as

$\zeta(s)=\frac{1}{s-1}-s\int_0^1 h(x) x^{s-1} \; dx$

which implies that

$\zeta(s)=\frac{s}{s-1}-s\int_0^1 dx\; x \left[Gx^{s-1} \right]$

by change-of-variable.

[edit] Matrix elements

Consider the Taylor series expansions at x=1 for a function f(x) and $g (x) = [G f](x)$ . That is, let

$f(1-x)=\sum_{n=0}^\infty (-x)^n \frac{f^{(n)}(1)}{n!}$

and write likewise for g(x). The expansion is made about x=1 because the GKW operator is poorly-behaved at x=0. The expansion is made about 1-x so that we can keep x a positive number, 0 ≤ x ≤ 1. Then the GKW operator acts on the Taylor coefficients as

$(-1)^m \frac{g^{(m)}(1)}{m!} = \sum_{n=0}^\infty G_{mn} (-1)^n \frac{f^{(n)}(1)}{n!},$

where the matrix elements of the GKW operator are given by

$G_{mn}=\sum_{k=0}^n (-1)^k {n \choose k} {k+m+1 \choose m} \left[ \zeta (k+m+2)- 1\right].$

This operator is extremely well-formed, and thus very numerically tractable. Note that each entry is a finite rational zeta series. The Gauss-Kuzmin constant is easily computed to high precision by numerically diagonalizing the upper-left n by n portion. There is no known closed-form expression that diagonalizes this operator; that is, there are no closed-form expressions known for the eigenvalues or eigenvectors.

[edit] Riemann zeta

The Riemann zeta can be written as

$\zeta(s)=\frac{s}{s-1}-s \sum_{n=0}^\infty (-1)^n {s-1 \choose n} t_n$

where the $t n$ are given by the matrix elements above:

$t_n=\sum_{m=0}^\infty \frac{G_{mn}} {(m+1)(m+2)}.$

Performing the summations, one gets:

$t_n=1-\gamma + \sum_{k=1}^n (-1)^n {n \choose k} \left[ \frac{1}{k} + \frac {\zeta(k+1)} {k+1} \right]$

where $γ$ is the Euler-Mascheroni constant. These $t n$ play the analog of the Stieltjes constants, but for the falling factorial expansion. By writing

$a_n=t_n - \frac{1}{2(n+1)}$

one gets: a₀ = −0.0772156... and a₁ = −0.00474863... and so on. The values get small quickly but are oscillatory. Some explicit sums on these values can be performed. They can be explicitly related to the Stieltjes constants by re-expressing the falling factorial as a polynomial with Stirling number coefficients, and then solving. More generally, the Riemann zeta can be re-expressed as an expansion in terms of Sheffer sequences of polynomials.

The expansion of Riemann zeta referred here as the "falling factorial" one, was introduced and thoroughly investigated in:

A. Yu. Eremin, I. E. Kaporin, and M. K. Kerimov, The calculation of the Riemann zeta-function in the complex domain, U.S.S.R. Comput. Math. and Math. Phys. 25 (1985), no. 2, 111--119

A. Yu. Yeremin, I. E. Kaporin, and M. K. Kerimov, Computation of the derivatives of the Riemann zeta-function in the complex domain, U.S.S.R. Comput. Math. and Math. Phys. 28 (1988), no. 4, 115--124

The latter paper contains a proof for the asymptotics of the coefficients of the expansion. The coefficients are decreasing nearly as

exp( - c n 1 / 2),

where $c$ is a positive constant.

[edit] References

A. Ya. Khinchin, Continued Fractions, 1935, English translation University of Chicago Press, 1961 ISBN 0-486-69630-8 (See section 15).
K. I. Babenko, On a Problem of Gauss, Soviet Mathematical Doklady 19:136-140 (1978) MR 57 #12436
K. I. Babenko and S. P. Jur'ev, On the Discretization of a Problem of Gauss, Soviet Mathematical Doklady 19:731-735 (1978). MR 81h:65015
Keith Briggs, A precise computation of the Gauss-Kuzmin-Wirsing constant (2003) (Contains a very extensive collection of references.)
A. Durner, On a Theorem of Gauss-Kuzmin-Lévy. Arch. Math. 58, 251-256, (1992). MR 93c:11056
Phillipe Flajolet and Brigitte Vallée, On the Gauss-Kuzmin-Wirsing Constant (1995).
A. J. MacLeod, High-Accuracy Numerical Values of the Gauss-Kuzmin Continued Fraction Problem. Computers Math. Appl. 26, 37-44, (1993).
E. Wirsing, On the Theorem of Gauss-Kuzmin-Lévy and a Frobenius-Type Theorem for Function Spaces. Acta Arith. 24, 507-528, (1974). MR 49 #2637
Linas Vepstas The Bernoulli Operator, the Gauss-Kuzmin-Wirsing Operator, and the Riemann Zeta (2004) (PDF)