A Note on Raising Operators of the Macdonald Polynomials

Jun'ichi Shiraishi Graduate School of Mathematical Science, University of Tokyo, Tokyo, Japan

Abstract
A multi variable hypergeometric-type formula for raising operators of the Macdonald polynomials is conjectured.
1. Introduction In this note, we present an observation that the raising operators for the Macdonald polynomials Q λ ( q , t )   [1can be written in a simple manner in terms of the q   -shifted factorials.
In the work by Lassalle and Schlosser [2, they have succeeded in obtaining a closed formulas of the raising operators for arbitrary Macdonald polynomials by inverting the Pieri formula, in terms of the q   -shifted factorials and a determinant. Hence the program initiated by Jing and Józefiak [3has been completed. In the papers [4, 5, the same object has been studied from a different point of view, and another kind of representation of the raising operators were obtained for some simple cases. Since there is no determinant appearing in the formulas given in [4, 5, it was expected that one may find a general formula for the raising operators without any determinant, if the advantage from the both approaches are taken. The aim of this paper is to present a conjectural formula for the raising operators of such kind. It has been checked for ( λ ) 4   and not very large | λ |   .
2. Basic hypergeometric-like series Let n   be a positive integer, and s 1 , s 2 , , s n   be indeterminates. Introduce c ~ n ( { i k , } 1 k < n ; s 1 , , s n )   recursively by c ~ 1 ( ; s 1 ) = 1   and
c ~ n ( { i k , } 1 k < n ; s 1 , , s n )
= c ~ n 1 ( { i k , } 1 k < n 1 ; q i 1 , n s 1 , , q i n 1 , n s n 1 ) (1)
× n 1 k = 1 t i k , n ( q t 1 ; q ) i k , n ( q ; q ) i k , n ( q t 1 s k / s n ; q ) i k , n ( q s k / s n ; q ) i k , n
× 1 k < n 1 ( q t 1 s k / s ; q ) i k , n ( q s k / s ; q ) i k , n ( q i , n t s k / s ; q ) i k , n ( q i , n s k / s ; q ) i k , n .
For example, we have
c ~ 2 ( { i 1 , 2 } ; s 1 , s 2 ) = t i 1 , 2 ( q t 1 ; q ) i 1 , 2 ( q ; q ) i 1 , 2 ( q t 1 s 1 / s 2 ; q ) i 1 , 2 ( q s 1 / s 2 ; q ) i 1 , 2 , (2)
c ~ 3 ( { i 1 , 2 , i 1 , 3 , i 2 , 3 } ; s 1 , s 2 , s 3 ) = t i 1 , 2 ( q t 1 ; q ) i 1 , 2 ( q ; q ) i 1 , 2 ( q i 1 , 3 i 2 , 3 q t 1 s 1 / s 2 ; q ) i 1 , 2 ( q i 1 , 3 i 2 , 3 q s 1 / s 2 ; q ) i 1 , 2
× t i 1 , 3 ( q t 1 ; q ) i 1 , 3 ( q ; q ) i 1 , 3 ( q t 1 s 1 / s 3 ; q ) i 1 , 3 ( q s 1 / s 3 ; q ) i 1 , 3 t i 2 , 3 ( q t 1 ; q ) i 2 , 3 ( q ; q ) i 2 , 3 ( q t 1 s 2 / s 3 ; q ) i 2 , 3 ( q s 2 / s 3 ; q ) i 2 , 3 (3)
× ( q t 1 s 1 / s 2 ; q ) i 1 , 3 ( q s 1 / s 2 ; q ) i 1 , 3 ( q i 2 , 3 t s 1 / s 2 ; q ) i 1 , 3 ( q i 2 , 3 s 1 / s 2 ; q ) i 1 , 3 ,
and so on. Note that c ~ n ( { i k , } 1 k < n ; s 1 , , s n )   is just given as a product of Lassalle and Schlosser's function C θ 1 , , θ n ( q , t ) ( u 1 , , u n )   (see §5 of [2) without taking into account the determinant factor.
In the paper [5, a Macdonald-type difference operator acting on the space of formal power series F [ [ x 2 / x 1 , x 3 / x 2 , , x n / x n 1 ] ]   was investigated, where F = Q ( q , t , s 1 , s 2 , , s n )   . It is defined by
D 1 ( s 1 , s 2 , , s n , t , q ) = i = 1 n s i j < i θ ( x i x j ) j > i θ + ( x j x i ) T q 1 , x i . (4)
Here T q , x i   denotes the q   -shift operator T q , x i g ( x 1 , , x n ) = g ( x 1 , , q x i , , x n )   and θ ± ( x )   are the series
θ ± ( x ) = 1 q ± 1 t 1 x 1 q ± 1 x = 1 + n = 1 ( 1 t 1 ) q ± n x n . (5)
Let us consider a basic hypergeometric-like series
f ( x 1 , x 2 , , x n ) = 1 k < n ( 1 x / x k ) (6)
× i k , 0 1 k < n c ~ n ( { i k , } 1 k < n ; s 1 , , s n ) 1 k < n ( x / x k ) i k , .
Then we observe Conjecture. The series f ( x 1 , x 2 , , x n )   in Eq.( 6 ) is an eigenfunction of the difference operator D 1  
D 1 ( s 1 , s 2 , , s n , t , q ) f ( x 1 , x 2 , , x n ) = i = 1 n s i f ( x 1 , x 2 , , x n ) . (7)
Note this conjecture implies that factoring out 1 k < n ( 1 x / x k )   from the series amounts to deal with the determinant in Lassalle and Schlosser's function C θ 1 , , θ n ( q , t ) ( u 1 , , u n )   .
3. Result From the conjecture above and Proposition A.6 in the paper [5, we obtain the raising operator for the Macdonald polynomials.
Proposition. Let λ = ( λ 1 , , λ n )   be a partition. Set s i = t n i q λ i   .
Assume that the above conjecture is true. Then the Macdonald polynomila Q λ = Q λ ( q , t )   can be represented by the raising operator R k   (see §1 of [1) as
Q ( λ 1 , , λ n )
= 1 k < n ( 1 R k ) (8)
× i k , 0 1 k < n c ~ n ( { i k , } 1 k < n ; s 1 , , s n ) 1 k < n R k i k , Q ( λ 1 ) Q ( λ n ) .
Acknowledgment. This work is supported by the Grant-in-Aid for Scientific Research (C) 16540183.
References

  1. I. G. Macdonald, Symmetric Functions and Hall Polynomials (2nd ed.), Oxford University Press, (1995).
  2. M. Lassalle and M. Schlosser, Inversion of the Pieri formula for Macdonald polynomials, math.CO/0402127.
  3. N. H. Jing and T. Joźefiak, A formula for two-row Macdonald functions, Duke Math. J. 67, 377-385 (1992).
  4. J. Shiraishi, A Commutative Family of Integral Transformations and Basic Hypergeometric Series. I. Eigenfunctions, math.QA/0501251.
  5. J. Shiraishi, A Commutative Family of Integral Transformations and Basic Hypergeometric Series. II. Eigenfunctions and Quasi-Eigenfunctions, math.QA/0502228.