Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation

We consider a q-Painlevé IV equation which is the A₄⁽¹⁾-surface type in the Sakai's classification. We find three distinct types of classical solutions with determinantal structures whose elements are basic hypergeometric functions. Two of them are expressed by ₂φ₁ basic hypergeometric series a...

Повний опис

Збережено в:
Бібліографічні деталі
Опубліковано в: :Symmetry, Integrability and Geometry: Methods and Applications
Дата:2014
Автор: Nakazono, N.
Формат: Стаття
Мова:Англійська
Опубліковано: Інститут математики НАН України 2014
Онлайн доступ:https://nasplib.isofts.kiev.ua/handle/123456789/146608
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Цитувати:Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation / N. Nakazono // Symmetry, Integrability and Geometry: Methods and Applications. — 2014. — Т. 10. — Бібліогр.: 41 назв. — англ.

Репозитарії

Digital Library of Periodicals of National Academy of Sciences of Ukraine
_version_ 1859817437277454336
author Nakazono, N.
author_facet Nakazono, N.
citation_txt Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation / N. Nakazono // Symmetry, Integrability and Geometry: Methods and Applications. — 2014. — Т. 10. — Бібліогр.: 41 назв. — англ.
collection DSpace DC
container_title Symmetry, Integrability and Geometry: Methods and Applications
description We consider a q-Painlevé IV equation which is the A₄⁽¹⁾-surface type in the Sakai's classification. We find three distinct types of classical solutions with determinantal structures whose elements are basic hypergeometric functions. Two of them are expressed by ₂φ₁ basic hypergeometric series and the other is given by ₂ψ₂ bilateral basic hypergeometric series.
first_indexed 2025-12-07T15:23:54Z
format Article
fulltext Symmetry, Integrability and Geometry: Methods and Applications SIGMA 10 (2014), 090, 23 pages Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation Nobutaka NAKAZONO School of Mathematics and Statistics, The University of Sydney, New South Wales 2006, Australia E-mail: nobua.n1222@gmail.com Received June 06, 2013, in final form August 14, 2014; Published online August 22, 2014 http://dx.doi.org/10.3842/SIGMA.2014.090 Abstract. We consider a q-Painlevé IV equation which is the A (1) 4 -surface type in the Sakai’s classification. We find three distinct types of classical solutions with determinantal structures whose elements are basic hypergeometric functions. Two of them are expressed by 2ϕ1 basic hypergeometric series and the other is given by 2ψ2 bilateral basic hypergeo- metric series. Key words: q-Painlevé equation; basic hypergeometric function; affine Weyl group; τ -func- tion; projective reduction; orthogonal polynomial 2010 Mathematics Subject Classification: 33D05; 33D15; 33D45; 33E17; 39A13 1 Introduction The focus of this paper is on the following single second-order ordinary difference equation: (Xn+1Xn − 1)(Xn−1Xn − 1) = q(−N+2n−m−1)/2a0a 3/2 1 a22 ( Xn + q(N−m)/2a 1/2 1 )( Xn + q(−N+m)/2a −1/2 1 ) Xn + q(−N+n−m)/2a 1/2 1 a2 , (1.1) where n ∈ Z is the independent variable, Xn = Xn(m,N) is the dependent variable and m,N ∈ Z and a0, a1, a2, q (|q| < 1) ∈ C∗ are parameters. Equation (1.1) is known as a q-discrete analog of the Painlevé IV equation (q-PIV) [35]. In 2001, Sakai introduced a geometric approach to the theory of the Painlevé and discrete Painlevé equations (Painlevé systems) and showed the classifications of Painlevé systems by the rational surface [37]. The rational surface can be identified with the space of initial condition, and the group of Cremona isometries associated with the surface generate the affine Weyl group. He also showed that the translation part of the affine Weyl group gives rise to various discrete Painlevé equations. Then, such discrete Painlevé equations are said to have the affine Weyl group symmetries. In 2004, q-PIV (1.1) was generalized to the following simultaneous first-order ordinary diffe- rence equations by the singularity confinement criterion [38]: (yk+1xk − 1)(ykxk − 1) = q(−N+4k−m+2l−2)/2a0a 3/2 1 a22a3 ( xk + q(N−m)/2a 1/2 1 )( xk + q(−N+m)/2a −1/2 1 ) xk + q(−N+2k−m+2l)/2a 1/2 1 a2 , (1.2a) (ykxk − 1)(ykxk−1 − 1) = q(−N+4k−m+2l−4)/2a0a 3/2 1 a22a3 ( yk + q(N−m)/2a 1/2 1 )( yk + q(−N+m)/2a −1/2 1 ) yk + q(−N+2k−m−2)/2a 1/2 1 a2a3 , (1.2b) mailto:nobua.n1222@gmail.com http://dx.doi.org/10.3842/SIGMA.2014.090 2 N. Nakazono where k ∈ Z is the independent variable, xk = xk(l,m,N) and yk = yk(l,m,N) are dependent variables and l,m,N ∈ Z and a0, a1, a2, a3, q (|q| < 1) ∈ C∗ are parameters. System (1.2) is known as a q-discrete analog of the Painlevé V equation (q-PV). It is also known that q-PV (1.2) is the A (1) 4 -surface type in the Sakai’s classification and has the affine Weyl group symmetry of type A (1) 4 . Conversely, q-PIV (1.1) can be recovered from q-PV (1.2) by putting a3 = q1/2, l = 0, and replacing the independent variable and the dependent variables by 2k = n, xk = Xn, yk = Xn−1. This procedure is referred to as “symmetrization” of q-PV (1.2), which comes from the ter- minology of the Quispel–Roberts–Thompson (QRT) mapping [33, 34]. After this terminology, q-PV (1.2) is sometimes called the “asymmetric discrete Painlevé equation”, and q-PIV (1.1) is called the “symmetric discrete Painlevé equation”. It appears as though the symmetrization is a simple specialization on the level of the equation, but the following problems were known: (i) According to Sakai’s theory [37], the discrete Painlevé equations arise as the birational mappings corresponding to the translations of the affine Weyl groups. The asymmetric discrete Painlevé equations are characterized in this manner, however, it was not known how to characterize the symmetric discrete Painlevé equations as the action of affine Weyl groups; (ii) Painlevé systems admit the particular solutions expressible in terms of the hypergeometric type functions (hypergeometric solutions) when some of the parameters take special values (see, for example, [13, 14] and references therein). However, the hypergeometric solutions to the symmetric discrete Painlevé equation cannot be obtained by the näıve specialization of those to the corresponding asymmetric equation. In [17], the mechanism of the symmetrization was investigated and the nontrivial incon- sistency among the hypergeometric solutions were explained in detail by taking an example of q-Painlevé equation with the affine Weyl group symmetry of type (A2 + A1) (1). The key to characterize the symmetric discrete Painlevé equation as the action of affine Weyl group is taking the half-step translation instead of a translation as a time evolution. In general, various discrete dynamical systems of Painlevé type can be obtained from elements of infinite order that are not necessarily translations in the affine Weyl group by taking the projection on appropriate subspaces of the parameter spaces. Such a procedure is called a “projective reduction”. It is well known that the τ -functions play a crucial role in the theory of integrable systems [25], and it is also possible to introduce them in the theory of Painlevé systems [6, 7, 8, 27, 29, 30, 31, 32]. A representation of the affine Weyl groups can be lifted on the level of the τ -functions [12, 15, 39], which gives rise to various bilinear equations of Hirota type satisfied the τ -functions. Usually, the hypergeometric solutions are derived by reducing the bilinear equations to the Plücker relations by using the contiguity relations satisfied by the entries of determinants [2, 3, 9, 10, 11, 18, 19, 20, 21, 28, 36]. This method is elementary, but it encounters technical difficulties for Painlevé systems with large symmetries. In order to overcome this difficulty, Masuda has proposed a method of constructing hypergeometric solutions under a certain boundary condition on the lattice where the τ -functions live (hypergeometric τ -functions), so that they are consistent with the action of the affine Weyl groups [23, 24, 26]. Although this requires somewhat complex calculations, the merit is that it is systematic and that it can be applied to the systems with large symmetries. Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 3 In [16], the list of the simplest hypergeometric solutions to the symmetric q-Painlevé equa- tions are shown. In general, hypergeometric solutions of Painlevé systems can be expressed by determinants whose entries are given by hypergeometric type functions. Therefore, it is natural to be curious about the determinant formulae of them. The purpose of this paper is to obtain the determinant formulae of the hypergeometric solutions to the q-PIV via the construction of the hypergeometric τ -functions and the theory of orthogonal polynomials. This paper is organized as follows: in Section 2, we first introduce a representation of the affine Weyl group of type A (1) 4 . Next, we show how q-PV (1.2) and q-PIV (1.1) can be de- rived from the representation. In Section 3, we construct the hypergeometric τ -functions for the q-PIV and obtain the hypergeometric solutions of the q-PIV which are expressed by basic hypergeometric series (see Theorems 1 and 2). In Section 4, we obtain the hypergeometric so- lutions of the q-PIV which are expressed by bilateral basic hypergeometric series via the theory of orthogonal polynomials (see Theorem 3). Some concluding remarks are given in Section 5. We use the following conventions of q-analysis with |q|, |p| < 1 throughout this paper [1, 22]: • q-shifted factorials: (a; q)∞ = ∞∏ i=1 ( 1− aqi−1 ) , (a; q)λ = (a; q)∞ (aqλ; q)∞ , where λ ∈ C; • Jacobi theta function: Θ(a; q) = (a; q)∞ ( qa−1; q ) ∞; • Elliptic gamma function: Γ(a; p, q) = ( pqa−1; p, q ) ∞ (a; p, q)∞ , where (a; p, q)k = k−1∏ i,j=0 ( 1− piqja ) ; • Basic hypergeometric series: sϕr ( a1, . . . , as b1, . . . , br ; q, z ) = ∞∑ n=0 (a1, . . . , as; q)n (b1, . . . , br; q)n(q; q)n [ (−1)nqn(n−1)/2 ]1+r−s zn, where (a1, . . . , as; q)n = s∏ j=1 (aj ; q)n; • Bilateral basic hypergeometric series: sψr ( a1, . . . , as b1, . . . , br ; q, z ) = ∞∑ n=−∞ (a1, . . . , as; q)n (b1, . . . , br; q)n [ (−1)nqn(n−1)/2 ]r−s zn; 4 N. Nakazono • Bilateral q-integral:∫ ∞ −∞ f(t)dqt = (1− q) ∞∑ n=−∞ ( f(qn) + f(−qn) ) qn. We note that the following formulae hold: (a; q)λ+1 (a; q)λ = 1− aqλ, Θ(qa; q) Θ(a; q) = −a−1, Γ(qa; q, q) Γ(a; q, q) = Θ(a; q). 2 Affine Weyl group of type A (1) 4 2.1 Birational representation of the affine Weyl group of type A (1) 4 In this section, we formulate the family of Bäcklund transformations of q-PV (1.2) as a birational representation of the affine Weyl group of type A (1) 4 . Let si (i = 0, 1, 2, 3, 4), σ and ι be transformations of the parameters ak (k = 0, 1, 2, 3, 4) and the variables fj (j = 0, 1, 2, 3, 4). The action of the transformations on the parameters is given by si(aj) = aja −aij i , σ(ai) = ai+1, ι : (a0, a1, a2, a3, a4) 7→ ( a−10 , a−14 , a−13 , a−12 , a−11 ) , where i, j ∈ Z/5Z and the symmetric 5× 5 matrix A = (aij) 4 i,j=0 =  2 −1 0 0 −1 −1 2 −1 0 0 0 −1 2 −1 0 0 0 −1 2 −1 −1 0 0 −1 2  is the Cartan matrix of type A (1) 4 . Moreover, the action on the variables is given by si(fi+2) = ai+3ai+4(aiai+1 + ai+3fi) a2i+1fi+3 , si(fi+4) = ai+4(ai+2 + ai+4aifi+1) aiai+1a2i+2fi+3 , si(fj) = fj , j 6= i+ 2, i+ 4, σ(fi) = fi+1, ι : (f0, f1, f2, f3, f4) 7→ (f1, f0, f4, f3, f2), where i ∈ Z/5Z. Note that the variables satisfy the following conditions: a2i+3ai+4fi = ai+1(aiai+1fi+2fi+3 − ai+3ai+4), where i ∈ Z/5Z. The conditions above look like five, but they are essentially three. Therefore, variables fi are essentially two. Proposition 1 ([2, 37, 39]). The group of birational transformations W̃ ( A (1) 4 ) = 〈s0, s1, s2, s3, s4, σ, ι〉 gives a representation of the (extended) affine Weyl group of type A (1) 4 . Namely, the transformations satisfy the fundamental relations s2i = 1, (sisi±1) 3 = 1, (sisj) 2 = 1, j 6= i± 1, σ5 = 1, σsi = si+1σ, ι2 = 1, ιs0 = s0ι, ιs1 = s4ι, ιs2 = s3ι, where i, j ∈ Z/5Z. Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 5 In general, for a function F = F (ai, fj), we let an element w ∈ W̃ ( A (1) 4 ) act as w.F (ai, fj) = F (w.ai, w.fj), that is, w acts on the arguments from the left. Note that q = a0a1a2a3a4 is in- variant under the action of 〈s0, s1, s2, s3, s4, σ〉. We define the translations Ti (i = 0, 1, 2, 3, 4) by T0 = σs4s3s2s1, T1 = σs0s4s3s2, T2 = σs1s0s4s3, T3 = σs2s1s0s4, T4 = σs3s2s1s0, (2.1) whose action on the parameters is given by T0 : (a0, a1, a2, a3, a4) 7→ ( qa0, q −1a1, a2, a3, a4 ) , T1 : (a0, a1, a2, a3, a4) 7→ ( a0, qa1, q −1a2, a3, a4 ) , T2 : (a0, a1, a2, a3, a4) 7→ ( a0, a1, qa2, q −1a3, a4 ) , T3 : (a0, a1, a2, a3, a4) 7→ ( a0, a1, a2, qa3, q −1a4 ) , T4 : (a0, a1, a2, a3, a4) 7→ ( q−1a0, a1, a2, a3, qa4 ) . Note that Ti (i = 0, 1, 2, 3, 4) commute with each other and T0T1T2T3T4 = 1. 2.2 Derivations of the q-Painlevé equations In this section, we derive the q-Painlevé equations from W̃ ( A (1) 4 ) . The action of T23 = T2T3 on f -variables can be expressed as (T23(y)x− 1)(yx− 1) = q−1a0a 3/2 1 a22a3 ( x+ a 1/2 1 )( x+ a −1/2 1 ) x+ a 1/2 1 a2 , (2.2) (yx− 1) ( yT−123 (x)− 1 ) = q−2a0a 3/2 1 a22a3 ( y + a 1/2 1 )( y + a −1/2 1 ) y + q−1a 1/2 1 a2a3 , (2.3) where x = a0a 1/2 1 a−13 f2, y = a −1/2 1 a−12 a−14 s4(f1). Applying T k23T l 2T m 0 T N 1 on equations (2.2) and (2.3) and putting xk(l,m,N) = T k23T l 2T m 0 T N 1 (x), yk(l,m,N) = T k23T l 2T m 0 T N 1 (y), we obtain q-PV (1.2). Then, we can regard T23 and Ti (i = 0, 1, 2, 3, 4) as the time evolution and the Bäcklund transformations of q-PV (1.2), respectively. We note that considering the action of T0: T0(g)g = ( f + a −3/4 0 a 1/4 1 a −1/4 3 )( f + a −3/4 0 a 1/4 1 a −1/4 3 a−14 ) 1 + a −1/4 0 a −1/4 1 a 1/4 3 f , T−10 (f)f = ( g + a −1/4 0 a 3/4 1 a 1/4 3 )( g + a −1/4 0 a 3/4 1 a2a 1/4 3 ) 1 + a 1/4 0 a 1/4 1 a −1/4 3 g , where f = a −3/4 0 a −3/4 1 a 3/4 3 f0, g = a 3/4 0 a 3/4 1 a −3/4 3 f2, we obtain another q-discrete analog of Painlevé V equation [37]: gn+1gn = ( fn+q−n+ k+l−m 4 a − 3 4 0 a 1 4 1 a − 1 4 3 )( fn+q−n+ k+l+3m 4 a − 3 4 0 a 1 4 1 a − 1 4 3 a−14 ) 1+q −k−l+m 4 a − 1 4 0 a − 1 4 1 a 1 4 3 fn , (2.4a) 6 N. Nakazono fn+1fn = ( gn+1+q−n−1+ 3k−l+m 4 a − 1 4 0 a 3 4 1 a 1 4 3 )( gn+1 + q−n−1+ −k+3l+m 4 a − 1 4 0 a 3 4 1 a2a 1 4 3 ) 1+q k+l−m 4 a 1 4 0 a 1 4 1 a − 1 4 3 gn+1 , (2.4b) where fn = fn(k, l,m) = Tn0 T k 1 T l 2T m 3 (f), gn = gn(k, l,m) = Tn0 T k 1 T l 2T m 3 (g). Thus, q-PV (1.2) and equation (2.4) are the Bäcklund transformations each other. In order to derive q-PIV (1.1), we factorize T23 as follows T23 = R2 23, where R23 is given by R23 = σs1s0s4. (2.5) The action of R23 on the parameters is given by R23 : (a0, a1, a2, a3, a4) 7→ ( a0, a1, a2a3, qa −1 3 , q−1a3a4 ) . Let us consider the projection of the action of R23 on the line a3 = q1/2. (2.6) Then, the action on the parameters becomes translational motion: R23 : (a0, a1, a2, a4) 7→ ( a0, a1, q 1/2a2, q −1/2a4 ) . Since the action of R23 on the variable f2 is given by R23(f2) = q a20a1f2 ( 1 + a0 ( a0f2 + q1/2 )( a0a1f2 + q1/2 ) q1/2 ( q1/2a2 + a0f2 ) f4 ) , R−123 (f2) = q a20a1f2 ( 1 + a1a 2 2 q1/2 f4 ) , we obtain ( R23(X)X − 1 )( R−123 (X)X − 1 ) = q−1/2a0a 3/2 1 a22 ( X + a 1/2 1 )( X + a −1/2 1 ) X + a 1/2 1 a2 , (2.7) where X = q−1/2a0a 1/2 1 f2. (2.8) Applying Rn23T m 0 T N 1 on equation (2.7) and putting Xn(m,N) = Rn23T m 0 T N 1 (X), we obtain q-PIV (1.1). Note that R23 commute with Ti (i = 0, 1, 4) and T0T1R 2 23T4 = 1. Then, R23 and Ti (i = 0, 1, 4) are regarded as the time evolution and the Bäcklund transformations of q-PIV (1.1), respectively. 3 Hypergeometric solutions of the q-PIV (I) In this section, we obtain the hypergeometric solutions of q-PIV (1.1) by constructing the hy- pergeometric τ -functions for the q-PIV. Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 7 3.1 τ -functions In this section, we define the τ -functions. We introduce the new variables τi (i = 1, 2, . . . , 7) with f2 = τ4τ5 τ6τ7 , f4 = τ1τ2 τ3τ7 , (3.1) and lift the representation of W̃ ( A (1) 4 ) on their level: s0(τ1) = a4(a0τ3τ4τ5 + a2a3τ1τ2τ6 + a0a3τ3τ6τ7) a20a1a2τ4τ7 , s0(τi) = τi, i = 2, 3, 5, 6, s0(τ4) = a0a4(a0τ3τ4τ5 + a2a3τ1τ2τ6 + a3τ3τ6τ7) a1a2τ1τ7 , s0(τ7) = a4 ( a20τ3τ4τ5 + a3a0τ3τ6τ7 + a2a3τ1τ2τ6 ) a0a1a2τ1τ4 , s1(τ1) = τ2, s1(τ2) = τ1, s1(τi) = τi, i = 3, . . . , 7, s2(τ1) = a0a1(a0τ4τ5 + a2a3τ6τ7) a23τ3 , s2(τ3) = a0a1(a0τ4τ5 + a3τ6τ7) a2a23τ1 , s2(τi) = τi, i = 2, 4, 5, 6, 7, s3(τ4) = a2(a2a3τ1τ2 + a0τ3τ7) a20a3a4τ6 , s3(τ6) = a2a3(a2τ1τ2 + a0τ3τ7) a20a4τ4 , s3(τi) = τi, i = 1, 2, 3, 5, 7, s4(τ4) = s4(τ5), s4(τ5) = s4(τ4), s4(τi) = τi, i = 1, 2, 3, 6, 7, ι : (τ1, τ2, τ3, τ4, τ5, τ6, τ7) = (τ4, τ5, τ6, τ1, τ2, τ3, τ7), σ(τ1) = a0a1 (a0τ4τ5 + a3τ6τ7) a2a23τ1 , σ(τ2) = τ3, σ(τ3) = τ6, σ(τ4) = a4 ( a20τ3τ4τ5 + a3a0τ3τ6τ7 + a2a3τ1τ2τ6 ) a0a1a2τ1τ4 , σ(τ5) = τ7, σ(τ6) = τ5, σ(τ7) = τ2. Then, we get the following proposition: Proposition 2 ([39]). The transformations: s0, s1, s2, s3, s4, σ, ι, on the variables τi (i = 1, 2, . . . , 7) also realize the (extended) affine Weyl group of type A (1) 4 . Let us define the τ -functions τn,mN (n,m,N ∈ Z) by τn,mN = Rn23T m 0 T N 1 (τ4). By definition, every τ -function can be determined by a rational function in 7 initial variables τi (i = 1, 2, . . . , 7). We note that the 7 initial variables are expressed by the τ -functions as the following (see Fig. 1): τ1 = τ1,01 , τ2 = τ1,10 , τ3 = τ1,11 , τ4 = τ0,00 , τ5 = τ3,11 , τ6 = τ2,11 , τ7 = τ1,00 . (3.2) 3.2 Hypergeometric τ -functions for the q-PIV The aim of this section is to construct the hypergeometric τ -functions for the q-PIV. We define the hypergeometric τ -functions for the q-PIV by τn,mN consistent with the action of 〈R23, T0〉. Here, we mean τ = τ(α) consistent with an action of transformation r as r.τ = τ(r.α). 8 N. Nakazono T1 T0 R23�4 = �0;00 �7 = �1;00 �2 = �1;10 �1 = �1;01 �3 = �1;11 �6 = �2;11 �5 = �3;11 Figure 1. Configuration of the τ -functions on the 3D-lattice. Hereinafter, we impose the condition (2.6), and then regard τ -functions τn,mN as the functions in a0 and a2 consistent with the action of 〈R23, T0〉, i.e., τn,mN = τ0,0N ( qma0, q n/2a2 ) . By definition, every τ -function τn,mN is determined by a rational function in τn,m0 and τn,m1 (or τ1, . . . , τ7). Thus, our purpose is determining τn,m0 and τn,m1 consistent with the action of 〈R23, T0〉 and constructing the closed-form expressions of τn,mN (N ≥ 2) under the condition a0a1 = q, (3.3) and the boundary condition τn,mN = 0, N < 0. (3.4) Henceforth, we construct the hypergeometric τ -functions for the q-PIV in the following four steps. Step 1. Conditions of τn,m 0 In the first step, we obtain the condition of τn,m0 , which follows from the boundary condi- tion (3.4). Lemma 1. The following bilinear equations hold: τn,mN+1τ n−1,m−1 N−1 − q(−n−4m+4N+7)/2a−20 a−12 τn,m−1N τn−1,mN − q(−n−2m+4N+4)/2a−10 a−12 τn,mN τn−1,m−1N = 0, (3.5) τn,mN+1τ n−1,m N−1 + q2N−n+1a−22 ( q(−n+2N+1)/2a−12 − 1 ) τn−1,mN τn,mN − q(−3n+6N+3)/2a−32 τn−2,mN τn+1,m N = 0, (3.6) τn,mN+1τ n,m N−1 + q(−2n+6N+1)/2a−22 ( 1− qN−m+1a−10 )( τn,mN )2 − q4N−4m+4a−40 τn,m−1N τn,m+1 N = 0. (3.7) Proof. Application of T1 on τ3 yields the following bilinear equations: T1(τ3)τ4 − q2a−10 a1a −1 2 τ1R −1 23 (τ3)− q3/2a1a−12 τ3R −1 23 (τ1) = 0, (3.8) T1(τ3)τ2 + q3/2a0a1a −2 2 (1− a1)(τ3)2 − a41τ1T0(τ3) = 0. (3.9) Applying Rn−123 Tm−10 TN−11 on equations (3.8) and (3.9) and substituting condition (3.3) in them, we obtain equations (3.5) and (3.7), respectively. Similarly, application of T1 on τ6 yields T1(τ6)τ2 + qa−22 ( q1/2a−12 − 1 ) τ3τ6 − q3/2a−32 R−123 (τ3)τ5 = 0. (3.10) Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 9 Then, applying Rn−223 Tm−10 TN−11 on equation (3.10) and substituting condition (3.3) in it, we obtain equation (3.6). Although we do not write the action of R23 and T0 on the variables τi here, it will be described in the next step. � Putting N = 0 in equations (3.5)–(3.7) and using the boundary condition (3.4), we get the following conditions: τn+1,m 0 τn,m+1 0 τn,m0 τn+1,m+1 0 = −q(2m−1)/2a0, (3.11) τn,m0 τn+3,m 0 τn+1,m 0 τn+2,m 0 = 1− q(n+1)/2a2, (3.12) τn,m0 τn,m+2 0( τn,m+1 0 )2 = q(−2n+8m+1)/2a40a −2 2 ( 1− q−ma−10 ) . (3.13) Step 2. Conditions of τn,m 1 In the second step, we shall get the conditions of τn,m1 from the consistency with the action of 〈R23, T0〉. By definitions (2.1) and (2.5) and Proposition 2, the action of T0 and R23 are given by the follows: T0(τ1) = τ3, T0(τ7) = τ2, R23(τ3) = τ6, R23(τ4) = τ7, R23(τ6) = τ5, T0(τ2) = q3/2a0a −1 1 a−12 τ2T0(τ6) + qa20a −1 2 T0(τ4)T0(τ5) τ6 , (3.14) T0(τ3) = a20a1T0(τ4)T0(τ5) + q−1/2a0a1a2T0(τ6)τ2 R23(τ2) , (3.15) T0(τ4) = a2τ2τ6 + qa−10 a−21 τ3R23(τ2) τ5 , (3.16) T0(τ5) = qa−20 a−21 a−12 τ2τ6 + q3/2a−20 a−31 a−12 τ3R23(τ2) τ4 , (3.17) T0(τ6) = qa−21 a−12 τ2τ6 + q3/2a−10 a−31 a−12 τ3R23(τ2) τ7 , (3.18) T−10 (τ1) = q−2a20a1a −1 2 T−10 (τ4)T −1 0 (τ5) + q−1/2a0a1a −1 2 T−10 (τ6)T −1 0 (τ7) R23(τ7) , (3.19) T−10 (τ4) = qa−20 a−21 a−12 τ7T −1 0 (τ6) + q1/2a−20 a−31 a−12 τ1R23(τ7) τ5 , (3.20) T−10 (τ5) = a2τ7T −1 0 (τ6) + a−10 a−21 τ1R23(τ7) τ4 , (3.21) T−10 (τ6) = q−1a20a −1 2 τ4τ5 + q−1/2a0a −1 1 a−12 τ6τ7 τ2 , (3.22) T−10 (τ7) = q−1a−21 a−12 τ7T −1 0 (τ6) + q−1/2a−10 a−31 a−12 τ1R23(τ7) τ6 , (3.23) R23(τ1) = q−1a20a −1 2 τ4τ5 + q−1/2a0a −1 1 a−12 τ6τ7 τ2 , (3.24) R23(τ2) = q−1a20a1a −1 2 τ4τ5 + q−1/2a0a1a −1 2 τ6τ7 τ1 , (3.25) R23(τ5) = q3/2a−20 a22R23(τ1)R23(τ2) + qa−20 a−11 τ6R23(τ7) τ4 , (3.26) 10 N. Nakazono R23(τ7) = q−1a20a1τ4τ5 + q−1/2a0a1a2τ6τ7 τ3 , (3.27) R−123 (τ1) = q−1/2a20a1a −1 2 R−123 (τ4)τ6 + a0a1a −1 2 τ3τ4 τ2 , (3.28) R−123 (τ2) = q−1/2a20a −1 2 R−123 (τ4)τ6 + a0a −1 1 a−12 τ3τ4 τ1 , (3.29) R−123 (τ3) = q−1a20a1R −1 23 (τ4)τ6 + q−1a0a1a2τ3τ4 τ7 , (3.30) R−123 (τ4) = q1/2a−20 a22τ1τ2 + qa−20 a−11 τ3τ7 τ5 . (3.31) Using notation (3.2) and condition (3.3), we can rewrite equations (3.14)–(3.31) as a2τ 2,1 1 τ1,20 = q1/2a20τ 1,1 0 τ2,21 + qa20τ 0,1 0 τ3,21 , (3.32) τ2,10 τ1,21 = qa0τ 0,1 0 τ3,21 + q1/2a2τ 2,2 1 τ1,10 , (3.33) qτ3,11 τ0,10 = qa2τ 1,1 0 τ2,11 + a0τ 1,1 1 τ2,10 , (3.34) q3/2a2τ 0,0 0 τ3,21 = q1/2τ1,10 τ2,11 + a0τ 1,1 1 τ2,10 , (3.35) q3/2a2τ 1,0 0 τ2,21 = q1/2a20τ 1,1 0 τ2,11 + a20τ 1,1 1 τ2,10 , (3.36) a2τ 2,0 0 τ1,−11 = q−1a0τ 0,−1 0 τ3,01 + q1/2τ2,01 τ1,−10 , (3.37) q5/2a2τ 3,1 1 τ0,−10 = q3/2τ1,00 τ2,01 + a0τ 1,0 1 τ2,00 , (3.38) q2τ0,00 τ3,01 = q2a2τ 1,0 0 τ2,01 + a0τ 1,0 1 τ2,00 , (3.39) q3/2a2τ 1,1 0 τ2,01 = q1/2a20τ 0,0 0 τ3,11 + a20τ 2,1 1 τ1,00 , (3.40) q7/2a2τ 2,1 1 τ1,−10 = q1/2a20τ 1,0 0 τ2,01 + a20τ 1,0 1 τ2,00 , (3.41) q3/2a2τ 1,1 0 τ2,01 = q1/2a20τ 0,0 0 τ3,11 + a20τ 2,1 1 τ1,00 , (3.42) a2τ 1,0 1 τ2,10 = a0τ 0,0 0 τ3,11 + q1/2τ2,11 τ1,00 , (3.43) a20τ 0,0 0 τ4,11 = q3/2a22τ 2,0 1 τ2,10 + a0τ 2,1 1 τ2,00 , (3.44) τ1,11 τ2,00 = a0τ 0,0 0 τ3,11 + q1/2a2τ 2,1 1 τ1,00 , (3.45) a2τ 1,1 0 τ0,01 = q1/2a0τ −1,0 0 τ2,11 + qτ1,11 τ0,00 , (3.46) qa2τ 1,0 1 τ0,10 = q1/2a20τ −1,0 0 τ2,11 + a20τ 1,1 1 τ0,00 , (3.47) τ1,00 τ0,11 = a0τ −1,0 0 τ2,11 + a2τ 1,1 1 τ0,00 , (3.48) a20τ 3,1 1 τ−1,00 = q1/2a22τ 1,0 1 τ1,10 + a0τ 1,1 1 τ1,00 , (3.49) respectively. By setting τn,m1 = ( q(n−1)/2a2; q 1/2 ) ∞τ n,m 0 Fn,m, (3.50) and using conditions (3.11)–(3.13), equations (3.32)–(3.49) can be reduced to the following contiguity relations: Fn+2,m − q(n−1)/2a2Fn+1,m − qm−2a0 ( 1− q(n−1)/2a2 ) Fn,m = 0, (3.51) Fn+1,m+1 − qm−1a0Fn,m+1 − q(n−2)/2a2Fn,m = 0, (3.52) q1/2Fn+2,m+1 − q1/2Fn+1,m+1 + q(n−1)/2a2 ( 1− q(n−1)/2a2 ) Fn,m = 0, (3.53)( 1− qm−1a0 ) Fn+1,m+1 − q(n−1)/2a2Fn+1,m − qn/2−1a2 ( 1− q(n−1)/2a2 ) Fn,m = 0, (3.54) Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 11 q3/2 ( 1− qm−1a0 ) Fn+2,m+1 − q(n+2)/2a2Fn+1,m − q(2m+n−1)/2a0a2 ( 1− q(n−1)/2a2 ) Fn,m = 0, (3.55) qFn+2,m+1 − qma0Fn,m+1 − q(2n−1)/2a22Fn,m = 0. (3.56) The correspondence between equations (3.32)–(3.49) and equations (3.51)–(3.56) is established as follows: (3.33), (3.34), (3.39), (3.45), (3.48)⇒ (3.51), (3.32), (3.40), (3.42), (3.47)⇒ (3.52), (3.37), (3.43), (3.46)⇒ (3.53), (3.36), (3.41)⇒ (3.54), (3.35), (3.38)⇒ (3.55), (3.44), (3.49)⇒ (3.56). Step 3. Determination of τn,m 0 and τn,m 1 In this step, we determine τn,m0 and τn,m1 , i.e., we solve equations (3.11)–(3.13) and equa- tions (3.51)–(3.56). It is easily verified that the function τn,m0 = ( qma0; q, q ) ∞ ( q(n+1)/2a2; q 1/2, q ) ∞Γ ( q(2n+2m−3)/4a 1/2 0 a2; q 1/2, q1/2 ) × Γ ( q(n−m+1)/4a −1/4 0 a 1/2 2 ; q1/4, q1/4 ) Γ ( q(n−m)/4a −1/4 0 a 1/2 2 ; q1/4, q1/4 ) Γ ( −q3m−1a30; q3, q3 ) Γ ( −q2na42; q2, q2 ) , (3.57) is the solution of equations (3.11)–(3.13). Therefore, the aim of this step is to solve the equa- tions (3.51)–(3.56). Since equations (3.51)–(3.56) are overdetermined system, let us first consider the essential conditions of Fn,m. Lemma 2. Equations (3.51) and (3.52) are essential conditions for Fn,m. Proof. Eliminating Fn,m+1 from equations (3.51)m→m+1 and (3.52), we obtain equation (3.53). In a similar manner, equations (3.54)–(3.56) can be proven by the following procedures: elimi- nating Fn+2,m+1 from equations (3.52)n→n+1 and (3.53), we obtain equation (3.54); elimina- ting Fn+1,m+1 from equations (3.52)n→n+1 and (3.53), we obtain equation (3.55); eliminating Fn+1,m+1 from equations (3.52) and (3.53), we obtain equation (3.56). These calculations mean that if Fn,m satisfies conditions (3.51) and (3.52), then it also satisfies conditions (3.53)–(3.56). Therefore we have completed the proof. � Next, we solve equations (3.51) and (3.52). Lemma 3. The general solution of contiguity relations (3.51) and (3.52) is given by Fn,m = An,m Θ ( qn/2a2; q 1/2 ) Θ ( q(2m−1)/2a0; q )( q(m−1)/2a 1/2 0 ; q1/2 ) ∞ Θ ( q(n+m−2)/2a 1/2 0 a2; q1/2 ) × 2ϕ1 ( 0, q(−m+2)/2a −1/2 0 −q1/2 ; q1/2, q(n−1)/2a2 ) +Bn,m Θ ( qn/2a2; q 1/2 ) Θ ( q(2m−1)/2a0; q )( −q(m−1)/2a1/20 ; q1/2 ) ∞ Θ ( −q(n+m−2)/2a1/20 a2; q1/2 ) × 2ϕ1 ( 0,−q(−m+2)/2a −1/2 0 −q1/2 ; q1/2, q(n−1)/2a2 ) , 12 N. Nakazono where An,m and Bn,m are periodic functions of period one for n and m, i.e., An,m = An+1,m = An,m+1, Bn,m = Bn+1,m = Bn,m+1. Proof. For convenience, we put t = qn/2a2, α = qma0, Fn,m = F (t, α). Then, equations (3.51) and (3.52) can be rewritten as F (qt, α)− q−1/2tF ( q1/2t, α ) − q−2α ( 1− q−1/2t ) F (t, α) = 0, (3.58) F ( q1/2t, qα ) − q−1αF (t, qα)− q−1tF (t, α) = 0, (3.59) respectively. Substituting F (t, α) = D(t, α) ∞∑ k=0 Ck(α)tk, in equation (3.58), we obtain D(qt, α) = q−2αD(t, α), (3.60) Ck(α) = ( q2D ( q1/2t, α ) D(t, α)−1α−1; q1/2 ) k qk/2 ( −q1/2, q1/2; q1/2 ) k C0(α). Therefore, we obtain the solution of equation (3.58): F (t, α) = D1(t, α) 2ϕ1 ( 0, qα−1/2 −q1/2 ; q1/2, q−1/2t ) +D2(t, α) 2ϕ1 ( 0,−qα−1/2 −q1/2 ; q1/2, q−1/2t ) , (3.61) where D1(t, α) and D2(t, α) are the solutions of (3.60) which satisfy D1 ( q1/2t, α ) = q−1α1/2D1(t, α), (3.62) D2 ( q1/2t, α ) = −q−1α1/2D2(t, α), (3.63) respectively. Substituting (3.61) in equation (3.59), we can obtain the following equations: q−1/2α1/2 2ϕ1 ( 0, q1/2α−1/2 −q1/2 ; q1/2, t ) − q−1α 2ϕ1 ( 0, q1/2α−1/2 −q1/2 ; q1/2, q−1/2t ) − q−1t D1(t, α) D1(t, qα) 2ϕ1 ( 0, qα−1/2 −q1/2 ; q1/2, q−1/2t ) = 0, (3.64) q−1/2α1/2 2ϕ1 ( 0,−q1/2α−1/2 −q1/2 ; q1/2, t ) + q−1α 2ϕ1 ( 0,−q1/2α−1/2 −q1/2 ; q1/2, q−1/2t ) + q−1t D2(t, α) D2(t, qα) 2ϕ1 ( 0,−qα−1/2 −q1/2 ; q1/2, q−1/2t ) = 0. (3.65) By the definition of basic hypergeometric series 2ϕ1, it is easily verified that 2ϕ1 ( 0, a c ; q1/2, z ) − a−1 2ϕ1 ( 0, a c ; q1/2, q−1/2z ) Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 13 − ( 1− a−1 ) 2ϕ1 ( 0, q1/2a c ; q1/2, q−1/2z ) = 0. (3.66) Therefore, we obtain the following conditions from equations (3.64) and (3.65) by using equa- tion (3.66): D1(t, qα) = − t α(1− q1/2α−1/2) D1(t, α), (3.67) D2(t, qα) = − t α(1 + q1/2α−1/2) D2(t, α). (3.68) By setting D1(t, α) = Θ ( t; q1/2 ) Θ ( q−1/2α; q )( q−1/2α1/2; q1/2 ) ∞ Θ ( q−1α1/2t; q1/2 ) A(t, α), D2(t, α) = Θ ( t; q1/2 ) Θ ( q−1/2α; q )( −q−1/2α1/2; q1/2 ) ∞ Θ ( −q−1α1/2t; q1/2 ) B(t, α), equations (3.62), (3.63), (3.67) and (3.68) can be rewritten as A ( q1/2t, α ) = A(t, α), B ( q1/2t, α ) = B(t, α), A(t, qα) = A(t, α), B(t, qα) = B(t, α), respectively. This completes the proof. � It was shown that hypergeometric solutions of a symmetric discrete Painlevé equation, which can be obtained by projective reduction, have two expressions and there are the following differences between the two expressions (see [17, Section 2.3]): (i) the bases of hypergeometric series appearing in the solutions are different; (ii) the periodicities of periodic functions appearing in the solutions are different. The differences between these two expressions can be explained by the factorization of the linear difference operators associated with the three-term relation of the hypergeometric functions (see [17, Section 3.2]). Namely, we can see these differences by comparing Lemmas 3 and 6. To get another expression, we first reselect essential conditions of Fn,m. Lemma 4. Equations (3.52) and qm−1a0 ( 1− qma0 ) Fn,m+2 − q(n−5)/2a2 (( 1 + q1/2 ) qma0 − qn/2a2 ) Fn,m+1 − q(2n−5)/2a22Fn,m = 0, (3.69) are essential conditions of Fn,m. Proof. Eliminating Fn,m from equations (3.52) and (3.69), we obtain qm−2a0 ( 1− qm−1a0 ) Fn,m+1 − q(n−3)/2a2Fn+1,m − q(n−5)/2a2 ( q(2m−1)/2a0 − qn/2a2 ) Fn,m = 0. (3.70) Similarly, eliminating Fn,m+1 from equations (3.52) and (3.70), we obtain( 1− qm−1a0 ) Fn+1,m+1 − q(n−1)/2a2Fn+1,m − q(n−2)/2a2 ( 1− q(n−1)/2a2 ) Fn,m = 0. (3.71) Finally, eliminating Fn+1,m+1 from equations (3.70)n→n+1 and (3.71), we obtain equation (3.51). This result together with Lemma 2 complete the proof. � 14 N. Nakazono By setting Fn,m = Θ ( qma0; q ) Θ ( qn/2a2; q 1/2 ) Θ ( −qn/2a2; q1/2 ) Gn−3,m−1, (3.72) equations (3.52) and (3.69) can be rewritten as q−m+1a−10 Gn−2,m +Gn−3,m − qn/2a2Gn−3,m−1 = 0, (3.73) Gn,m−2 + ( q−m+1a−10 − ( 1 + q1/2 ) q(−n−3)/2a−12 ) Gn,m−1 − q(−2n−5)/2a−22 ( q−m+1a−10 − 1 ) Gn,m = 0, (3.74) respectively. Before solving equations (3.73) and (3.74), we prepare the following lemma: Lemma 5. The following recurrence relations hold: 2ϕ1 ( a, b c ; q, z ) − 2ϕ1 ( a, b c ; q, qz ) = (1− a)(1− b)z 1− c 2ϕ1 ( qa, qb qc ; q, z ) , (3.75) ( q−1c− 1 ) 2ϕ1 ( a, b q−1c ; q, z ) + 2ϕ1 ( a, b c ; q, z ) − q−1c 2ϕ1 ( a, b c ; q, qz ) = 0. (3.76) Proof. Substituting 2ϕ1 ( a, b c ; q, z ) = 1 + ∞∑ n=0 (qa, qb; q)n (qc, q; q)n (1− a)(1− b) (1− c) ( 1− qn+1 )zn+1, in the left-hand side of equation (3.75), we obtain the right-hand side. Equation (3.76) can be easily verified as the following: 2ϕ1 ( a, b q−1c ; q, z ) = ∞∑ n=0 (a, b; q)n (c, q; q)n 1− qn−1c 1− q−1c zn = 1 1− q−1c 2ϕ1 ( a, b c ; q, z ) − q−1c 1− q−1c 2ϕ1 ( a, b c ; q, qz ) . Therefore we have completed the proof. � Using Lemma 5, we obtain the following lemma: Lemma 6. The general solution of equations (3.73) and (3.74) is given by Gn,m = Λn,m Θ ( q(n−2m+2)/2a−10 a2; q ) Θ ( q−ma−10 ; q ) Θ ( qn/2a2; q )( q(n+3)/2a2; q ) ∞ ( q−1/2; q ) ∞ × 2ϕ1 ( 0, q(n+3)/2a2 q3/2 ; q, q−m+1a−10 ) + Λn+1,m q1/2Θ ( q(n−2m+3)/2a−10 a2; q ) Θ ( q−ma−10 ; q ) Θ ( q(n+1)/2a2; q )( q(n+2)/2a2; q ) ∞ ( q1/2; q ) ∞ × 2ϕ1 ( 0, q(n+2)/2a2 q1/2 ; q, q−m+1a−10 ) , where Λn,m is a periodic function of period two for n and period one for m, i.e., Λn+2,m = Λn,m+1 = Λn,m. Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 15 Proof. For convenience, we put t = q−m+1a−10 , α = qn/2a2, Gn,m = G(t, α). Then, equations (3.73) and (3.74) can be rewritten as tG ( t, q−1α ) +G ( t, q−3/2α ) − αG ( qt, q−3/2α ) = 0, (3.77) G ( q2t, α ) + ( t− ( 1 + q1/2 ) q−3/2α−1 ) G(qt, α)− q−5/2α−2(t− 1)G(t, α) = 0, (3.78) respectively. Substituting G(t, α) = D(t, α) ∞∑ k=0 Ck(α)tk, in equation (3.78), we obtain G(t, α) = D1(t, α) 2ϕ1 ( 0, q3/2α q3/2 ; q, t ) +D2(t, α) 2ϕ1 ( 0, qα q1/2 ; q, t ) , (3.79) where D1(t, α) and D2(t, α) satisfy D1(qt, α) = q−1α−1D1(t, α), (3.80) D2(qt, α) = q−3/2α−1D2(t, α), (3.81) respectively. Substituting (3.79) in equation (3.77), we can obtain the following equations: 2ϕ1 ( 0, q−1/2α q1/2 ; q, t ) − 2ϕ1 ( 0, q−1/2α q1/2 ; q, qt ) = −t D1 ( t, q−1α ) D2 ( t, q−3/2α ) 2ϕ1 ( 0, q1/2α q3/2 ; q, t ) ,(3.82) t D2 ( t, q−1α ) D1 ( t, q−3/2α ) 2ϕ1 ( 0, α q1/2 ; q, t ) + 2ϕ1 ( 0, α q3/2 ; q, t ) − q1/22ϕ1 ( 0, α q3/2 ; q, qt ) = 0. (3.83) Therefore, we obtain D1 ( t, q1/2α ) = − 1− qα 1− q1/2 D2(t, α), (3.84) D2 ( t, q1/2α ) = −1− q1/2 t D1(t, α), (3.85) from equations (3.82) and (3.83) by using equations (3.75) and (3.76), respectively. By setting D1(t, α) = Θ(αt; q)( q−1/2; q ) ∞ ( q3/2α; q ) ∞Θ ( q−1t; q ) Θ(α; q) Λ(t, α), D2(t, α) = q1/2Θ(q1/2αt; q)( q1/2; q ) ∞(qα; q)∞Θ ( q−1t; q ) Θ ( q1/2α; q )Λ ( t, q1/2α ) , equations (3.80), (3.81), (3.84) and (3.85) can be reduced to Λ(t, qα) = Λ(qt, α) = Λ(t, α). This completes the proof. � 16 N. Nakazono Step 4. Constructing the closed-form expressions of τn,m N (N ≥ 2) In this final step, constructing the closed-form expressions of τn,mN (N ≥ 2), we obtain the hypergeometric τ -functions for the q-PIV. Let τn,mN = (−1)N(N−1)/2q3N(N−1)(N−n+1)/4a −3N(N−1)/2 2 ( N∏ k=1 ( q(n−2k+1)/2a2; q 1/2 ) ∞ ) × ( qma0; q, q ) ∞ ( q(n+1)/2a2; q 1/2, q ) ∞Γ ( q(2n+2m−3)/4a 1/2 0 a2; q 1/2, q1/2 ) × Γ ( q(n−m+1)/4a −1/4 0 a 1/2 2 ; q1/4, q1/4 ) Γ ( q(n−m)/4a −1/4 0 a 1/2 2 ; q1/4, q1/4 ) Γ ( −q3m−1a30; q3, q3 ) Γ ( −q2na42; q2, q2 ) φn,mN . From (3.4), (3.50) and (3.57), we get φn,mN = 0, N < 0, φn,m0 = 1, φn,m1 = Fn,m. Moreover, it is easily verified that equation (3.6) can be rewritten as φn,mN+1φ n−1,m N−1 − φ n−1,m N φn,mN + φn−2,mN φn+1,m N = 0. (3.86) In general, equation (3.86) admits a solution expressed in terms of Jacobi–Trudi type determi- nant φn,mN = det(cn−2i+j+1,m)i,j=1,...,N , N > 0, under the boundary conditions φn,mN = 0, N < 0, φn,m0 = 1, φn,m1 = cn,m, where cn,m is an arbitrary function. Therefore, we obtain the following lemma: Lemma 7. Under the assumptions a0a1 = q, τn,mN = 0, N < 0, the hypergeometric τ -functions for the q-PIV are given as τn,mN = (−1)N(N−1)/2q3N(N−1)(N−n+1)/4a −3N(N−1)/2 2 ( N∏ k=1 ( q(n−2k+1)/2a2; q 1/2 ) ∞ ) × ( qma0; q, q ) ∞ ( q(n+1)/2a2; q 1/2, q ) ∞Γ ( q(2n+2m−3)/4a 1/2 0 a2; q 1/2, q1/2 ) × Γ ( q(n−m+1)/4a −1/4 0 a 1/2 2 ; q1/4, q1/4 ) Γ ( q(n−m)/4a −1/4 0 a 1/2 2 ; q1/4, q1/4 ) Γ ( −q3m−1a30; q3, q3 ) Γ ( −q2na42; q2, q2 ) φn,mN , where φn,mN = ∣∣∣∣∣∣∣∣∣ Fn,m Fn+1,m . . . Fn+N−1,m Fn−2,m Fn−1,m . . . Fn+N−3,m ... ... . . . ... Fn−2N+2,m Fn−2N+3,m . . . Fn−N+1,m ∣∣∣∣∣∣∣∣∣ , φn,m0 = 1, φn,m−N = 0, N > 0. Here, Fn,m is given in Lemma 3. Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 17 We also show another expression of the hypergeometric τ -functions for the q-PIV. From relation (3.72), φn,mN can be rewritten as φn,mN = Θ ( qma0; q )N (N−1∏ k=0 Θ ( q(n+k)/2a2; q 1/2 ) Θ ( −q(n+k)/2a2; q1/2 ))ψn−3,m−1N , where ψn,mN = ∣∣∣∣∣∣∣∣∣ Gn,m Gn+1,m . . . Gn+N−1,m Gn−2,m Gn−1,m . . . Gn+N−3,m ... ... . . . ... Gn−2N+2,m Gn−2N+3,m . . . Gn−N+1,m ∣∣∣∣∣∣∣∣∣ , ψn,m0 = 1, ψn,m−N = 0, N > 0. This gives the following lemma: Lemma 8. Under the assumptions a0a1 = q, τn,mN = 0, N < 0, the hypergeometric τ -functions for the q-PIV are given as τn,mN = (−1)N(N−1)/2q3N(N−1)(N−n+1)/4a −3N(N−1)/2 2 ( N∏ k=1 ( q(n−2k+1)/2a2; q 1/2 ) ∞ ) × ( qma0; q, q ) ∞ ( q(n+1)/2a2; q 1/2, q ) ∞Γ ( q(2n+2m−3)/4a 1/2 0 a2; q 1/2, q1/2 ) × Γ ( q(n−m+1)/4a −1/4 0 a 1/2 2 ; q1/4, q1/4 ) Γ ( q(n−m)/4a −1/4 0 a 1/2 2 ; q1/4, q1/4 ) Γ ( −q3m−1a30; q3, q3 ) Γ ( −q2na42; q2, q2 ) ×Θ ( qma0; q )N (N−1∏ k=0 Θ ( q(n+k)/2a2; q 1/2 ) Θ ( −q(n+k)/2a2; q1/2 ))ψn−3,m−1N , where ψn,mN = ∣∣∣∣∣∣∣∣∣ Gn,m Gn+1,m . . . Gn+N−1,m Gn−2,m Gn−1,m . . . Gn+N−3,m ... ... . . . ... Gn−2N+2,m Gn−2N+3,m . . . Gn−N+1,m ∣∣∣∣∣∣∣∣∣ , ψn,m0 = 1, ψn,m−N = 0, N > 0. Here, Gn,m is given in Lemma 6. 3.3 Hypergeometric solutions of the q-PIV In this section, we show the hypergeometric solutions of q-PIV (1.1). From relations (2.8) and (3.1), the variable for q-PIV (1.1) is expressed by the τ -functions as the following: Xn(m,N) = q(m+N−1)/2a0a 1/2 1 τn,mN τn+3,m+1 N+1 τn+2,m+1 N+1 τn+1,m N . Therefore, from Lemmas 7 and 8, we obtain the following theorems: 18 N. Nakazono Theorem 1. The hypergeometric solutions of q-PIV (1.1) with a0a1 = q, N ≥ 0, (3.87) is given by Xn(m,N) = −q(−2N−m+1)/2a −1/2 0 φn,mN φn+3,m+1 N+1 φn+2,m+1 N+1 φn+1,m N , where φn,mN = ∣∣∣∣∣∣∣∣∣ Fn,m Fn+1,m . . . Fn+N−1,m Fn−2,m Fn−1,m . . . Fn+N−3,m ... ... . . . ... Fn−2N+2,m Fn−2N+3,m . . . Fn−N+1,m ∣∣∣∣∣∣∣∣∣ , φn,m0 = 1. Here, Fn,m is given in Lemma 3. Theorem 2. The hypergeometric solutions of q-PIV (1.1) with the condition (3.87) is given by Xn(m,N) = q(−2N−m+1)/2a −1/2 0 ψn−3,m−1N ψn,mN+1 ψn−1,mN+1 ψn−2,m−1N , where ψn,mN = ∣∣∣∣∣∣∣∣∣ Gn,m Gn+1,m . . . Gn+N−1,m Gn−2,m Gn−1,m . . . Gn+N−3,m ... ... . . . ... Gn−2N+2,m Gn−2N+3,m . . . Gn−N+1,m ∣∣∣∣∣∣∣∣∣ , ψn,m0 = 1. Here, Gn,m is given in Lemma 6. 4 Hypergeometric solutions of the q-PIV (II) In this section, we show that q-PIV (1.1) also has the hypergeometric solutions expressed by bilateral basic hypergeometric series. First, we recall the definitions of orthogonal polynomials. Definition 1. A polynomial sequence (Pn(t))∞n=0 which satisfies the following conditions is called an orthogonal polynomial sequence over the field K, and each term Pn(t) is called an orthogonal polynomial over the field K. (i) deg(Pn(t)) = n; (ii) there exists a linear functional L : K(t)→ K which holds the orthogonal condition: L[tkPn(t)] = hnδn,k, n ≥ k, where δn,k is Kronecker’s symbol. Here, hn and µn = L[tn] are called a normalization factor and a moment, respectively. Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 19 Definition 2. An orthogonal polynomial sequence whose leading coefficient is 1 is called a monic orthogonal polynomial sequence (MOPS). Let (Pn(t))∞n=0 be a MOPS. Then, polynomial Pn(t) and its normalization factor hn are expressed by the moment µn as the following: Pn = 1 Φn ∣∣∣∣∣∣∣∣∣∣∣ µ0 µ1 . . . µn−1 µn µ1 µ2 . . . µn µn+1 ... ... . . . ... ... µn−1 µn . . . µ2n−2 µ2n−1 1 t . . . tn−1 tn ∣∣∣∣∣∣∣∣∣∣∣ , P0 = 1, hn = Φn+1 Φn , h0 = µ0, (4.1) where Φn is the Hankel determinant given by Φn = ∣∣∣∣∣∣∣∣∣ µ0 µ1 . . . µn−1 µ1 µ2 . . . µn ... ... . . . ... µn−1 µn . . . µ2n−2 ∣∣∣∣∣∣∣∣∣ . (4.2) We assume that (Pn)∞n=0 = (Pn(t))∞n=0 and (P̂n)∞n=0 = (P̂n(t))∞n=0 are MOPSs which satisfy the following orthogonal conditions: L [ tkPn(t) ] = hnδn,k, n ≥ k, L̂ [ tkP̂n(t) ] = ĥnδn,k, n ≥ k, respectively. In addition, we put the case that Pn and P̂n are related by the Christoffel trans- formation (or Geronimus transformation), that is, the linear functionals satisfy the following relation for an arbitrary function f(t): L[f(t)] = L̂ [ f(t) t− c0 + δ(t− c0) ] , where δ(x) is the Dirac delta function and c0 ∈ C is an additional parameter. For these MOPSs, the following relations hold [4, 40, 41]: (t− c0)P̂n = Pn+1 + ĥn hn Pn, (4.3) Pn = P̂n + hn ĥn−1 P̂n−1. (4.4) Eliminating Pn from equation (4.3) by using equation (4.4), we obtain the following three-term recurrence relation: tP̂n = P̂n+1 + ( hn+1 ĥn + ĥn hn + c0 ) P̂n + ĥn ĥn−1 P̂n−1. (4.5) Let P̂n(t) = h̃n(c1t; p) cn1 , c1 > 0, where h̃n(t; q) is the discrete q-Hermite II polynomial: h̃n(t; q) = tn2ϕ1 ( q−n, q−n+1 0 ; q2,−q 2 t2 ) . 20 N. Nakazono Then, the linear functional, the normalization factor and the three-term recurrence relation for P̂n are given by L̂[f(t)] = ∫ ∞ −∞ f(t)( −c21t2; p2 ) ∞ dpt, ĥn = 2 pn2c2n1 (p; p)n ( p2; p2 ) ∞Θ ( −pc21; p2 )( p3; p2 ) ∞Θ ( −c21; p2 ) , (4.6) tP̂n = P̂n+1 + p−2n+1 ( 1− pn ) c−21 P̂n−1, (4.7) respectively. We note that these properties of q-Hermite II polynomials are given in [22]. We here impose the condition c0 6= pa for all a ∈ Z since the linear functional for Pn is given by L[f(t)] = ∫ ∞ −∞ f(t) (t− c0) ( −c21t2; p2 ) ∞ dpt. In addition, the moment µn can be obtained by µn = −1− p c20 ∞∑ k=−∞ ( 1− (−1)n ) pk + ( 1 + (−1)n ) c0( 1− p2kc−20 )( −c21p2k; p2 ) ∞ pk(n+1) = 2(1− p)( 1− c20 )( −c21; p2 ) ∞ ∞∑ k=−∞ ( −c21, c −2 0 ; p2 ) k( p2c−20 ; p2 ) k ( 1− (−1)n 2 pk + 1 + (−1)n 2 c0 ) pk(n+1) =  2(1− p)( 1− c20 )( −c21; p2 ) ∞ 2ψ2 ( −c21, c −2 0 0, p2c−20 ; p2, p2k+1 ) , n = 2k − 1, 2c0(1− p)( 1− c20 )( −c21; p2 ) ∞ 2ψ2 ( −c21, c −2 0 0, p2c−20 ; p2, p2k+1 ) , n = 2k. (4.8) Comparing the coefficients of equations (4.5) and (4.7), we obtain the following equations: hn+1 ĥn + ĥn hn + c0 = 0, (4.9) ĥn ĥn−1 = p−2n+1 ( 1− pn ) c−21 . (4.10) From equations (4.9) and (4.10), we obtain ĥn hn+1 = − hn p−2n+1(1− pn)c−21 ĥn−1 + c0hn . (4.11) By setting Xn = i 1− pn+1 pnc1 ĥn hn+1 , (4.12) equation (4.11) can be rewritten as the following discrete Riccati equation: Xn = 1− pn+1 Xn−1 + ipnc1c0 . (4.13) Since in the case of a 1/2 0 a 1/2 1 = q1/2, a2 = 1, N = −1, Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 21 q-PIV (1.1) admits the reduction to Xn = 1− q(n+1)/2 Xn−1 + q(n−m+1)/2a −1/2 0 , which is equivalent to equation (4.13) with the following correspondence: qm/2a 1/2 0 = −iq1/2c−10 c−11 , q1/2 = p, (4.1), (4.2), (4.6), (4.8) and (4.12) give the hypergeometric solutions of q-PIV (1.1). Therefore, we finally obtain the following theorem: Theorem 3. In the case of a 1/2 0 a 1/2 1 = q1/2, a2 = 1, N = −1, n ≥ 0, q-PIV (1.1) with qm/2a 1/2 0 = −iq1/2c−10 c−11 , admits the following hypergeometric solution: Xn = 2i ( 1− q(n+1)/2 )( q1/2; q1/2 ) n (q; q)∞Θ ( −q1/2c21; q ) qn(n+1)/2c2n+1 1 ( q3/2; q ) ∞Θ ( −c21; q ) Φn+1 Φn+2 , where Φn = ∣∣∣∣∣∣∣∣∣ µ0 µ1 . . . µn−1 µ1 µ2 . . . µn ... ... . . . ... µn−1 µn . . . µ2n−2 ∣∣∣∣∣∣∣∣∣ , µn =  2 ( 1− q1/2 )( 1− c20 )( −c21; q ) ∞ 2ψ2 ( −c21, c −2 0 0, qc−20 ; q, q(2k+1)/2 ) , n = 2k − 1, 2c0 ( 1− q1/2 )( 1− c20 )( −c21; q ) ∞ 2ψ2 ( −c21, c −2 0 0, qc−20 ; q, q(2k+1)/2 ) , n = 2k. Here, c0 6= qa/2 for all a ∈ Z. 5 Concluding remarks In this paper, we have constructed the hypergeometric solutions of q-PIV (1.1) via the construc- tion of the hypergeometric τ -functions and the theory of orthogonal polynomials. We showed that the hypergeometric solutions of the q-PIV can be expressed by the three expressions. We note that the hypergeometric solutions of Painlevé systems expressed by the determinants whose sizes do not depend on the independent variable are called the lattice type solutions, while those expressed by the determinants whose sizes depend on the independent variable are called molecule type solutions. Thus, the hypergeometric solutions given in Theorems 1 and 2 are lattice type solutions whereas those given in Theorem 3 are molecule type solutions. Before closing, we mention the bilateral type hypergeometric solutions here. It is well known that the coalescence cascade of hypergeometric functions, from the Gauss’s hypergeometric func- tion to the Airy function, corresponds to the diagram of degeneration of the Painlevé equations, 22 N. Nakazono from the Painlevé VI equation to the Painlevé II equation, in the sense of the hypergeometric solutions [5]: PVI → PV → PIII Gauss Kummer Bessel ↓ ↓ PIV → PII Weber Airy Similarly, the relations between basic hypergeometric series and q-Painlevé equations are also investigated [14, 16]. However, the hypergeometric solutions described by bilateral basic hyper- geometric series have not been considered. It might be an interesting future problem to make a list of the bilateral basic hypergeometric series that appear as the solutions of the q-Painlevé equations. Acknowledgments The author would like to thank Professors K. Kajiwara, S. Kakei, H. Miki, M. Noumi, and S. Tsujimoto for the useful comments. He also appreciates the valuable comments from the referees which have improved the quality of this paper. This work has been supported by JSPS Grant-in-Aid for Scientific Research No. 22·4366 and the Australian Research Council grant DP130100967. References [1] Gasper G., Rahman M., Basic hypergeometric series, Encyclopedia of Mathematics and its Applications, Vol. 96, 2nd ed., Cambridge University Press, Cambridge, 2004. [2] Hamamoto T., Kajiwara K., Hypergeometric solutions to the q-Painlevé equation of type A (1) 4 , J. Phys. A: Math. Gen. 40 (2007), 12509–12524, nlin.SI/0701001. [3] Hamamoto T., Kajiwara K., Witte N.S., Hypergeometric solutions to the q-Painlevé equation of type (A1 + A′1)(1), Int. Math. Res. Not. 2006 (2006), 84619, 26 pages, nlin.SI/0607065. [4] Ismail M.E.H., Classical and quantum orthogonal polynomials in one variable, Encyclopedia of Mathematics and its Applications, Vol. 98, Cambridge University Press, Cambridge, 2005. [5] Iwasaki K., Kimura H., Shimomura S., Yoshida M., From Gauss to Painlevé. A modern theory of special functions, Aspects of Mathematics, Vol. E16, Friedr. Vieweg & Sohn, Braunschweig, 1991. [6] Jimbo M., Miwa T., Monodromy preserving deformation of linear ordinary differential equations with ra- tional coefficients. II, Phys. D 2 (1981), 407–448. [7] Jimbo M., Miwa T., Monodromy preserving deformation of linear ordinary differential equations with ra- tional coefficients. III, Phys. D 4 (1981), 26–46. [8] Jimbo M., Miwa T., Ueno K., Monodromy preserving deformation of linear ordinary differential equations with rational coefficients. I. General theory and τ -function, Phys. D 2 (1981), 306–352. [9] Joshi N., Kajiwara K., Mazzocco M., Generating function associated with the Hankel determinant formula for the solutions of the Painlevé IV equation, Funkcial. Ekvac. 49 (2006), 451–468, nlin.SI/0512041. [10] Kajiwara K., Kimura K., On a q-difference Painlevé III equation. I. Derivation, symmetry and Riccati type solutions, J. Nonlinear Math. Phys. 10 (2003), 86–102, nlin.SI/0205019. [11] Kajiwara K., Masuda T., A generalization of determinant formulae for the solutions of Painlevé II and XXXIV equations, J. Phys. A: Math. Gen. 32 (1999), 3763–3778, solv-int/9903014. [12] Kajiwara K., Masuda T., Noumi M., Ohta Y., Yamada Y., 10E9 solution to the elliptic Painlevé equation, J. Phys. A: Math. Gen. 36 (2003), L263–L272, nlin.SI/0303032. [13] Kajiwara K., Masuda T., Noumi M., Ohta Y., Yamada Y., Hypergeometric solutions to the q-Painlevé equations, Int. Math. Res. Not. 2004 (2004), 2497–2521, nlin.SI/0403036. [14] Kajiwara K., Masuda T., Noumi M., Ohta Y., Yamada Y., Construction of hypergeometric solutions to the q-Painlevé equations, Int. Math. Res. Not. 2005 (2005), 1441–1463, nlin.SI/0501051. http://dx.doi.org/10.1017/CBO9780511526251 http://dx.doi.org/10.1088/1751-8113/40/42/S01 http://dx.doi.org/10.1088/1751-8113/40/42/S01 http://arxiv.org/abs/nlin.SI/0701001 http://dx.doi.org/10.1155/IMRN/2006/84619 http://arxiv.org/abs/nlin.SI/0607065 http://dx.doi.org/10.1017/CBO9781107325982 http://dx.doi.org/10.1017/CBO9781107325982 http://dx.doi.org/10.1007/978-3-322-90163-7 http://dx.doi.org/10.1016/0167-2789(81)90021-X http://dx.doi.org/10.1016/0167-2789(81)90003-8 http://dx.doi.org/10.1016/0167-2789(81)90013-0 http://dx.doi.org/10.1619/fesi.49.451 http://arxiv.org/abs/nlin.SI/0512041 http://dx.doi.org/10.2991/jnmp.2003.10.1.7 http://arxiv.org/abs/nlin.SI/0205019 http://dx.doi.org/10.1088/0305-4470/32/20/309 http://arxiv.org/abs/solv-int/9903014 http://dx.doi.org/10.1088/0305-4470/36/17/102 http://arxiv.org/abs/nlin.SI/0303032 http://dx.doi.org/10.1155/S1073792804140919 http://arxiv.org/abs/nlin.SI/0403036 http://dx.doi.org/10.1155/IMRN.2005.1439 http://arxiv.org/abs/nlin.SI/0501051 Hypergeometric Solutions of the A (1) 4 -Surface q-Painlevé IV Equation 23 [15] Kajiwara K., Masuda T., Noumi M., Ohta Y., Yamada Y., Point configurations, Cremona transformations and the elliptic difference Painlevé equation, in Théories asymptotiques et équations de Painlevé, Sémin. Congr., Vol. 14, Soc. Math. France, Paris, 2006, 169–198, nlin.SI/0411003. [16] Kajiwara K., Nakazono N., Hypergeometric solutions to the symmetric q-Painlevé equations, Int. Math. Res. Not., to appear, arXiv:1304.0858. [17] Kajiwara K., Nakazono N., Tsuda T., Projective reduction of the discrete Painlevé system of type (A2 + A1)(1), Int. Math. Res. Not. (2011), 930–966, arXiv:0910.4439. [18] Kajiwara K., Noumi M., Yamada Y., A study on the fourth q-Painlevé equation, J. Phys. A: Math. Gen. 34 (2001), 8563–8581, nlin.SI/0012063. [19] Kajiwara K., Ohta Y., Determinant structure of the rational solutions for the Painlevé IV equation, J. Phys. A: Math. Gen. 31 (1998), 2431–2446, solv-int/9709011. [20] Kajiwara K., Ohta Y., Satsuma J., Casorati determinant solutions for the discrete Painlevé III equation, J. Math. Phys. 36 (1995), 4162–4174, solv-int/9412004. [21] Kajiwara K., Ohta Y., Satsuma J., Grammaticos B., Ramani A., Casorati determinant solutions for the discrete Painlevé-II equation, J. Phys. A: Math. Gen. 27 (1994), 915–922, solv-int/9310002. [22] Koekoek R., Lesky P.A., Swarttouw R.F., Hypergeometric orthogonal polynomials and their q-analogues, Springer Monographs in Mathematics, Springer-Verlag, Berlin, 2010. [23] Masuda T., Hypergeometric τ -functions of the q-Painlevé system of type E (1) 7 , SIGMA 5 (2009), 035, 30 pages, arXiv:0903.4102. [24] Masuda T., Hypergeometric τ -functions of the q-Painlevé system of type E (1) 8 , Ramanujan J. 24 (2011), 1–31. [25] Miwa T., Jimbo M., Date E., Solitons. Differential equations, symmetries and infinite-dimensional algebras, Cambridge Tracts in Mathematics, Vol. 135, Cambridge University Press, Cambridge, 2000. [26] Nakazono N., Hypergeometric τ functions of the q-Painlevé systems of type (A2 +A1)(1), SIGMA 6 (2010), 084, 16 pages, arXiv:1008.2595. [27] Noumi M., Painlevé equations through symmetry, Translations of Mathematical Monographs, Vol. 223, Amer. Math. Soc., Providence, RI, 2004. [28] Ohta Y., Nakamura A., Similarity KP equation and various different representations of its solutions, J. Phys. Soc. Japan 61 (1992), 4295–4313. [29] Okamoto K., Studies on the Painlevé equations. III. Second and fourth Painlevé equations, PII and PIV, Math. Ann. 275 (1986), 221–255. [30] Okamoto K., Studies on the Painlevé equations. I. Sixth Painlevé equation PVI, Ann. Mat. Pura Appl. 146 (1987), 337–381. [31] Okamoto K., Studies on the Painlevé equations. II. Fifth Painlevé equation PV, Japan. J. Math. (N.S.) 13 (1987), 47–76. [32] Okamoto K., Studies on the Painlevé equations. IV. Third Painlevé equation PIII, Funkcial. Ekvac. 30 (1987), 305–332. [33] Quispel G.R.W., Roberts J.A.G., Thompson C.J., Integrable mappings and soliton equations, Phys. Lett. A 126 (1988), 419–421. [34] Quispel G.R.W., Roberts J.A.G., Thompson C.J., Integrable mappings and soliton equations. II, Phys. D 34 (1989), 183–192. [35] Ramani A., Grammaticos B., Discrete Painlevé equations: coalescences, limits and degeneracies, Phys. A 228 (1996), 160–171, solv-int/9510011. [36] Sakai H., Casorati determinant solutions for the q-difference sixth Painlevé equation, Nonlinearity 11 (1998), 823–833. [37] Sakai H., Rational surfaces associated with affine root systems and geometry of the Painlevé equations, Comm. Math. Phys. 220 (2001), 165–229. [38] Tamizhmani K.M., Grammaticos B., Carstea A.S., Ramani A., The q-discrete Painlevé IV equations and their properties, Regul. Chaotic Dyn. 9 (2004), 13–20. [39] Tsuda T., Tau functions of q-Painlevé III and IV equations, Lett. Math. Phys. 75 (2006), 39–47. [40] Uvarov V.B., The connection between systems of polynomials that are orthogonal with respect to different distribution functions, USSR Comput. Math. Math. Phys. 9 (1969), no. 6, 25–36. [41] Zhedanov A., Rational spectral transformations and orthogonal polynomials, J. Comput. Appl. Math. 85 (1997), 67–86. http://arxiv.org/abs/nlin.SI/0411003 http://dx.doi.org/10.1093/imrn/rnt237 http://dx.doi.org/10.1093/imrn/rnt237 http://arxiv.org/abs/1304.0858 http://dx.doi.org/10.1093/imrn/rnq089 http://arxiv.org/abs/0910.4439 http://dx.doi.org/10.1088/0305-4470/34/41/312 http://arxiv.org/abs/nlin.SI/0012063 http://dx.doi.org/10.1088/0305-4470/31/10/017 http://arxiv.org/abs/solv-int/9709011 http://dx.doi.org/10.1063/1.531353 http://arxiv.org/abs/solv-int/9412004 http://dx.doi.org/10.1088/0305-4470/27/3/030 http://arxiv.org/abs/solv-int/9310002 http://dx.doi.org/10.1007/978-3-642-05014-5 http://dx.doi.org/10.3842/SIGMA.2009.035 http://arxiv.org/abs/0903.4102 http://dx.doi.org/10.1007/s11139-010-9262-1 http://dx.doi.org/10.3842/SIGMA.2010.084 http://arxiv.org/abs/1008.2595 http://dx.doi.org/10.1143/JPSJ.61.4295 http://dx.doi.org/10.1143/JPSJ.61.4295 http://dx.doi.org/10.1007/BF01458459 http://dx.doi.org/10.1007/BF01762370 http://dx.doi.org/10.1016/0375-9601(88)90803-1 http://dx.doi.org/10.1016/0167-2789(89)90233-9 http://dx.doi.org/10.1016/0378-4371(95)00439-4 http://arxiv.org/abs/solv-int/9510011 http://dx.doi.org/10.1088/0951-7715/11/4/004 http://dx.doi.org/10.1007/s002200100446 http://dx.doi.org/10.1070/RD2004v009n01ABEH000260 http://dx.doi.org/10.1007/s11005-005-0037-3 http://dx.doi.org/10.1016/0041-5553(69)90124-4 http://dx.doi.org/10.1016/S0377-0427(97)00130-1 1 Introduction 2 Affine Weyl group of type A4(1) 2.1 Birational representation of the affine Weyl group of type A4(1) 2.2 Derivations of the q-Painlevé equations 3 Hypergeometric solutions of the q-PIV (I) 3.1 -functions 3.2 Hypergeometric -functions for the q-PIV 3.3 Hypergeometric solutions of the q-PIV 4 Hypergeometric solutions of the q-PIV (II) 5 Concluding remarks References
id nasplib_isofts_kiev_ua-123456789-146608
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
issn 1815-0659
language English
last_indexed 2025-12-07T15:23:54Z
publishDate 2014
publisher Інститут математики НАН України
record_format dspace
spelling Nakazono, N.
2019-02-10T09:57:23Z
2019-02-10T09:57:23Z
2014
Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation / N. Nakazono // Symmetry, Integrability and Geometry: Methods and Applications. — 2014. — Т. 10. — Бібліогр.: 41 назв. — англ.
1815-0659
2010 Mathematics Subject Classification: 33D05; 33D15; 33D45; 33E17; 39A13
DOI:10.3842/SIGMA.2014.090
https://nasplib.isofts.kiev.ua/handle/123456789/146608
We consider a q-Painlevé IV equation which is the A₄⁽¹⁾-surface type in the Sakai's classification. We find three distinct types of classical solutions with determinantal structures whose elements are basic hypergeometric functions. Two of them are expressed by ₂φ₁ basic hypergeometric series and the other is given by ₂ψ₂ bilateral basic hypergeometric series.
The author would like to thank Professors K. Kajiwara, S. Kakei, H. Miki, M. Noumi, and S. Tsujimoto for the useful comments. He also appreciates the valuable comments from the referees which have improved the quality of this paper. This work has been supported by JSPS Grant-in-Aid for Scientific Research No. 22·4366 and the Australian Research Council grant DP130100967.
en
Інститут математики НАН України
Symmetry, Integrability and Geometry: Methods and Applications
Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation
Article
published earlier
spellingShingle Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation
Nakazono, N.
title Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation
title_full Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation
title_fullStr Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation
title_full_unstemmed Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation
title_short Hypergeometric Solutions of the A₄⁽¹⁾-Surface q-Painlevé IV Equation
title_sort hypergeometric solutions of the a₄⁽¹⁾-surface q-painlevé iv equation
url https://nasplib.isofts.kiev.ua/handle/123456789/146608
work_keys_str_mv AT nakazonon hypergeometricsolutionsofthea41surfaceqpainleveivequation