Research article Special Issues

Yabu's formulae for hypergeometric 3F2-series through Whipple's quadratic transformations

  • By means of Whipple's quadratic transformations, two classes of hypergeometric 3F2-series are expressed in terms of the Lerch transcendent function. Several difficult series with a free variable are explicitly evaluated in closed form, including Yabu's three remarkable identities.

    Citation: Marta Na Chen, Wenchang Chu. Yabu's formulae for hypergeometric 3F2-series through Whipple's quadratic transformations[J]. AIMS Mathematics, 2024, 9(8): 21799-21815. doi: 10.3934/math.20241060

    Related Papers:

    [1] A. Belafhal, N. Nossir, L. Dalil-Essakali, T. Usman . Integral transforms involving the product of Humbert and Bessel functions and its application. AIMS Mathematics, 2020, 5(2): 1260-1274. doi: 10.3934/math.2020086
    [2] Abdelhamid Zaidi, Saleh Almuthaybiri . Explicit evaluations of subfamilies of the hypergeometric function 3F2(1) along with specific fractional integrals. AIMS Mathematics, 2025, 10(3): 5731-5761. doi: 10.3934/math.2025264
    [3] D. L. Suthar, D. Baleanu, S. D. Purohit, F. Uçar . Certain k-fractional calculus operators and image formulas of k-Struve function. AIMS Mathematics, 2020, 5(3): 1706-1719. doi: 10.3934/math.2020115
    [4] Gauhar Rahman, Shahid Mubeen, Kottakkaran Sooppy Nisar . On generalized k-fractional derivative operator. AIMS Mathematics, 2020, 5(3): 1936-1945. doi: 10.3934/math.2020129
    [5] Chuanan Wei, Chun Li . New q-supercongruences arising from a summation of basic hypergeometric series. AIMS Mathematics, 2022, 7(3): 4125-4136. doi: 10.3934/math.2022228
    [6] Ye-Cong Han, Chuan-Yu Cai, Ti-Ren Huang . Monotonicity, convexity properties and inequalities involving Gaussian hypergeometric functions with applications. AIMS Mathematics, 2022, 7(4): 4974-4991. doi: 10.3934/math.2022277
    [7] Min Wang, Jiao Teng, Lei Wang, Junmei Wu . Application of ADMM to robust model predictive control problems for the turbofan aero-engine with external disturbances. AIMS Mathematics, 2022, 7(6): 10759-10777. doi: 10.3934/math.2022601
    [8] Robert Reynolds, Allan Stauffer . A quintuple integral involving the product of Hermite polynomial Hn(βx) and parabolic cylinder function Dv(αt): derivation and evaluation. AIMS Mathematics, 2022, 7(5): 7464-7470. doi: 10.3934/math.2022418
    [9] Mahmoud Abul-Ez, Mohra Zayed, Ali Youssef . Further study on the conformable fractional Gauss hypergeometric function. AIMS Mathematics, 2021, 6(9): 10130-10163. doi: 10.3934/math.2021588
    [10] Li Gong, Bo Jiang . A matrix analysis of BLMBPs under a general linear model and its transformation. AIMS Mathematics, 2024, 9(1): 1840-1860. doi: 10.3934/math.2024090
  • By means of Whipple's quadratic transformations, two classes of hypergeometric 3F2-series are expressed in terms of the Lerch transcendent function. Several difficult series with a free variable are explicitly evaluated in closed form, including Yabu's three remarkable identities.



    Denote by Z and N, respectively, the sets of integers and natural numbers. For an indeterminate x, define the rising factorials by

    (x)0=1and(x)n=x(x+1)(x+n1)fornN.

    According to Bailey [3,§2.1], the classical hypergeometric series, for mN and an indeterminate z, reads as

    1+mFm[a0,a1,,amb1,,bm|z]=k=0(a0)k(a1)k(am)kk!(b1)k(bm)kzk.

    When |z|<1 and none of the numerator and denominator parameters results in a non-positive integer, the corresponding series is not only convergent, but also well-defined and nonterminating.

    There exist numerous hypergeometric series identities in the literature (see [4, Chapter 8] and [6,7]). Recently, algebraic expressions for certain classes of 3F2-series arose much attention (see [1,2,5]). In particular, Yabu [9] succeeded in evaluating explicitly the following series with a free variable x in terms of the logarithmic function:

    3F2[1,1,1243,53|x],3F2[1,1,1254,74|x],3F2[1,1,1276,116|x].

    The formulae for these series are remarkable, since it is rare that a hypergeometric series of higher order beyond Gauss' classical 2F1-series with a free variable turns into a closed algebraic expression. Motivated by Yabu's formulae, we shall investigate two general classes of the 3F2-series as below:

    F(m,λ,y):=3F2[1,m2,1+m21+λ,1+mλ|y],G(m,λ,y):=3F2[1,1+m2,1+m21+λ,1+mλ|y],

    where λ(0,1) and mN, with y being a free variable subject to |y|<1 such that the series are convergent. Instead of algebraic-geometric approach employed in [1,2,9], we find that the quadratic transformations due to Whipple [8] (cf. Bailey [3, page 97]) are more efficient. To facilitate their subsequent use, they are reproduced as follows:

    3F2[a2,1+a2,1+abc1+ab,1+ac|y]=(1x)a3F2[a,b,c1+ab,1+ac|x], (1)
    3F2[1+a2,1+a2,1+abc1+ab,1+ac|y]=(1x)1+a1+x4F3[a,1+a2,b,ca2,1+ab,1+ac|x], (2)

    where the two variables are related by equations

    y=4x(1x)2x=(11y)2y, (3)

    with the domain y(1,1) and the codomain x(1,322), respectively.

    In the next section, we shall first reformulate F(m,λ,y) by means of (1) and then evaluate the resulting series by the Lerch transcendent function. Then, in Section 3, the series G(m,λ,y) will be treated analogously via the second quadratic transformation (2). The two main theorems (Theorems 2.1 and 3.1) state that the series F(m,λ,y) (also G(m,λ,y)) results in a two-term linear combination of the Lerch transcendent function plus a remainder polynomial. Finally, the paper will end in Section 4, where several difficult series are explicitly evaluated in closed form as applications. Compared with the algebraic method adopted by Yabu [9], the authors believe that the approach presented in this paper is simpler and more accessible.

    In Whipple's first transformation (1), by specifying the parameters

    a=m,c=λ,b=ac,

    we can reformulate the series F(m,λ,y) as

    3F2[1,m2,1+m21+λ,1+mλ|y]=(1x)m3F2[m,λ,mλ1+λ,1+mλ|x]. (4)

    The rightmost 3F2(x)-series can be explicitly expressed as

    3F2[m,λ,mλ1+λ,1+mλ|x]=λ(mλ)(m1)!n=0(n+1)m1xn(n+λ)(n+mλ). (5)

    Keeping in mind that λ(0,1) and mN, it suffices to examine the case "m2λ". Otherwise, the only case exists for "m=2λ=1", in which we have a simpler series

    3F2[1,1,1232,32|y]=(1x)×3F2[1,12,1232,32|x]=n=0(1x)xn(2n+1)2=1x4Φ(x,2,12),

    where Φ() stands for the Lerch transcendent function:

    Φ(z,σ,α)=n=0zn(n+α)σ, for |z|<1,(σ)>0 and αZN.

    Now rewrite the rational function by

    R(n)=(n+1)m1(n+λ)(n+mλ)=(n+1)m1m2λ{1n+λ1n+mλ}. (6)

    According to the Chu-Vandermonde convolution formula

    (n+1)m1=mi=1(m1i1)(n+λ)i1(1λ)mi                   =mi=1(m1i1)(n+mλ)i1(1m+λ)mi,

    we can express

    R(n)=1m2λ{(1λ)m1n+λ+mi=2(m1i1)(1+n+λ)i2(1λ)mi         (1m+λ)m1n+mλmi=2(m1i1)(1+n+mλ)i2(1m+λ)mi}.

    Rewrite further the shifted factorials

    (1+n+λ)i2=ik=2(i2k2)(n+1)k2(λ)ik,(1+n+mλ)i2=ik=2(i2k2)(n+1)k2(mλ)ik,

    and then making substitutions, we can manipulate the double series

    3F2[m,λ,mλ1+λ,1+mλ|x]=λ(mλ)(m1)!n=0R(n)xn  =λ(mλ)(m1)!(m2λ)n=0{(1λ)m1n+λxn(1m+λ)m1n+mλxn}  +λ(mλ)(m1)!(m2λ)n=0xnmi=2(m1i1)(1λ)miik=2(i2k2)(n+1)k2(λ)ik  λ(mλ)(m1)!(m2λ)n=0xnmi=2(m1i1)(1m+λ)miik=2(i2k2)(n+1)k2(mλ)ik.

    ● The sum in the first line results in

    λ(1λ)m(m1)!(m2λ)Φ(x,1,λ)(mλ)(1m+λ)m(m1)!(m2λ)Φ(x,1,mλ).

    ● The double sum in the middle line can be simplified into a finite sum

        λ(mλ)m2λmi=2(miλmi)ik=2(1)ik(λik)n=0xni1(n+k2k2)=λ(mλ)m2λmi=2ik=2(1)ik(miλmi)(λik)(1x)1ki1.

    ● The double sum in the ultimate line can be reduced analogously to a finite sum

        λ(mλ)m2λmi=2(λimi)ik=2(1)ik(λmik)n=0xni1(n+k2k2)=λ(mλ)m2λmi=2ik=2(1)ik(λimi)(λmik)(1x)1ki1.

    Summing up, we have established the following theorem:

    Theorem 1 (m2λ). For two variables x and y related by (3), we have

    3F2[1,m2,1+m21+λ,1+mλ|y]=(1x)m3F2[m,λ,mλ1+λ,1+mλ|x]=(1x)mΔ(m,λ;x)+λ(1λ)m(m1)!(m2λ)(1x)mΦ(x,1,λ)(mλ)(1m+λ)m(m1)!(m2λ)(1x)mΦ(x,1,mλ),

    where the remainder term is given by two finite sums

    Δ(m,λ;x)=λ(mλ)m2λmi=2ik=2(1)ik(miλmi)(λik)(1x)1ki1                 λ(mλ)m2λmi=2ik=2(1)ik(λimi)(λmik)(1x)1ki1.

    Alternatively, by specifying the parameters in Whipple's second transformation (2)

    a=m,c=λ,b=ac,

    we can transform the series G(m,λ,y) as

    3F2[1,1+m2,1+m21+λ,1+mλ|y]=(1x)m+11+x4F3[m,1+m2,λ,mλm2,1+λ,1+mλ|x]. (7)

    The 4F3(x)-series on the right can be explicitly restated as

    4F3[m,1+m2,λ,mλm2,1+λ,1+mλ|x]=λ(mλ)m!n=0(n+1)m1(m+2n)(n+λ)(n+mλ)xn. (8)

    Analogously, the only series with "m=2λ=1" is the following reduced one:

    2F1[1,132|y]=(1x)21+x×2F1[1,1232|x]=(1x)2arctanhx(1+x)x.

    Let R(n) be a rational function subject to with "m2λ"

    R(n):=(n+1)m1(n+m2)(n+λ)(n+mλ)=(n+1)m12{1n+λ+1n+mλ}.

    Observe that the above R(n) resembles almost identically that R(n) in (6) under replacements "m2λ2" for denominator factors and "+" inside braces "{}". By applying the same procedure used to prove Theorem 1, we derive the formula presented in the following theorem.

    Theorem 2 (m2λ). For two variables x and y related by (3), we have

    3F2[1,1+m2,1+m21+λ,1+mλ|y]=(1x)m+11+x4F3[m,1+m2,λ,mλm2,1+λ,1+mλ|x]                                           =(1x)m+11+x(m,λ;x)+λ(1λ)mm!(1x)m+11+xΦ(x,1,λ)                                           +(mλ)(1m+λ)mm!(1x)m+11+xΦ(x,1,mλ),

    where the remainder term is given by two finite sums

    (m,λ;x)=λ(mλ)mmi=2ik=2(1)ik(miλmi)(λik)(1x)1ki1                 +λ(mλ)mmi=2ik=2(1)ik(λimi)(λmik)(1x)1ki1.

    According to Theorems 1 and 2, both series F(m,λ,y) and G(m,λ,y) can be expressed in terms of the Lerch transcendent function Φ(x,m,λ) plus a remainder polynomial. When the involved Φ(x,m,λ) admit explicit expressions in terms of logarithmic and arctan functions, we then find closed formulae for the corresponding series F(m,λ,y) and G(m,λ,y).

    Throughout this section, x and y are two variables related by (3). For m=1 and λ{13,14,16}, we are first going to review three formulae due to Yabu [9]. Then, for m=1 and irreducible rational numbers λ=p/qQ with q{5,8,10,12}, several closed formulae will be shown in pairs for series F(1,p/q,y) and G(1,p/q,y). Finally, when m1, we shall record a few expressions, as examples, for F(m,p/q,y) and G(m,p/q,y) in terms of the Lerch transcendent function.

    We first review the explicit formulae for three particular 3F2-series in terms of the logarithmic function, obtained by Yabu in his thesis [9].

    ● Yabu's first formula (cf. [9, Theorem 1.4]) can be reproduced as below:

    3F2[1,1,1243,53|t6]=4i33t3{A(eπi3t)ln(1+3t32B(t))+A(t)ln(13t32B(eπi3t))},

    where

    A(t)=t(1+1t6)13(1+1t6)13t,
    B(t)=t(1+1t6)13+(1+1t6)13t.

    By making use of the trisection series (or Mathematica command "FunctionExpand"), we have the explicit expressions

    Φ(x,1,13)=123x{3ln(31x13x)+23arctan(33x2+3x)},Φ(x,1,23)=123x2{3ln(31x13x)23arctan(33x2+3x)}.

    According to Theorems 1 and 2, we obtain the following two closed formulae:

    F(1,13,y)=3F2[1,1,1243,53|y]=23(1x){Φ(x,1,13)Φ(x,1,23)}                =1x33x2{3(13x)ln(13x31x)+23(1+3x)arctan(33x2+3x)},G(1,13,y)=3F2[1,1,3243,53|y]=2(1x)29(1+x){Φ(x,1,13)+Φ(x,1,23)}                =(1x)293x2{31+3x1+xln(31x13x)2313x1+xarctan(33x2+3x)}.

    Without involving the imaginary root of unity, these expressions have advantages over Yabu's.

    ● Yabu's second formula reads as (see [9, Theorem 1.5])

    3F2[1,1,1254,74|t4]=3i1t22t3ln(1t2it)31+t22t3ln(1+t2t).

    Recalling Theorems 1 and 2, and then applying two equalities:

    Φ(x,1,14)=14x{ln(1+x141x14)+2arctan(x14)},Φ(x,1,34)=14x3{ln(1+x141x14)2arctan(x14)},

    we can directly write down two elegant closed formulae:

    F(1,14,y)=3F2[1,1,1254,74|y]=38(1x){Φ(x,1,14)Φ(x,1,34)}                =3(1x)84x3{(1x)ln(1x141+x14)+2(1+x)arctan(x14)},G(1,14,y)=3F2[1,1,3254,74|y]=3(1x)216(1+x){Φ(x,1,14)+Φ(x,1,34)}                =3(1x)2164x3{1+x1+xln(1+x141x14)21x1+xarctan(x14)}.

    These formulae look more transparent than Yabu's formula.

    ● The third formula due to Yabu [9, Theorem 1.6] is given by

    3F2[1,1,1276,116|t6]=5i12t3{A(eπi3t)ln(23t3(B(t)1)2+3t3(B(t)1))+A(t)ln(2+3t3(B(eπi3t)1)23t3(B(eπi3t)1))},

    where

    A(t)=t2(1+1t6)23(1+1t6)23t2,
    B(t)=t2(1+1t6)23+(1+1t6)23t2.

    By employing the two explicit expressions:

    Φ(x,1,16)=126x{ln((1x)(1+x16)3(1+x)(1x16)3)+23arctan(3x161x13)},Φ(x,1,56)=126x5{ln((1x)(1+x16)3(1+x)(1x16)3)23arctan(3x161x13)},

    and then, from Theorems 1 and 2, we derive the following closed formulae:

    F(1,16,y)=3F2[1,1,1276,116|y]=524(1x){Φ(x,1,16)Φ(x,1,56)}                =5(1x)486x5{(1x23)ln((1+x)(1x16)3(1x)(1+x16)3)+23(1+x23)arctan(3x161x13)},G(1,16,y)=3F2[1,1,3276,116|y]=5(1x)236(1+x){Φ(x,1,16)+Φ(x,1,56)}                =5(1x)2726x5{1+x231+xln((1x)(1+x16)3(1+x)(1x16)3)231x231+xarctan(3x161x13)}.

    They look simpler than Yabu's original formula.

    By carrying out the same procedure as in §4.1, we can establish further closed formulae for series F(1,λ,y) and G(1,λ,y).

    F(1,15,y) and G(1,15,y). Applying the explicit expressions:

    Φ(x,1,15)=145x{5ln((1x)151x15)+210+25arctan(10+25x154+(15)x15)                +5ln(2+(1+5)x15+2x252+(15)x15+2x25)+21025arctan(1025x154+(1+5)x15)},Φ(x,1,45)=145x4{5ln((1x)151x15)210+25arctan(10+25x154+(15)x15)                +5ln(2+(1+5)x15+2x252+(15)x15+2x25)21025arctan(1025x154+(1+5)x15)},

    we derive the following two closed formulae:

    F(1,15,y)=3F2[1,1,1265,95|y]=415(1x){Φ(x,1,15)Φ(x,1,45)}=1x155x4{5(1x35)ln(1x15(1x)15)+210+25(1+x35)arctan(10+25x154+(15)x15)+5(1x35)ln(2+(15)x15+2x252+(1+5)x15+2x25)+21025(1+x35)arctan(1025x154+(1+5)x15)},G(1,15,y)=3F2[1,1,3265,95|y]=4(1x)225(1+x){Φ(x,1,15)+Φ(x,1,45)}=(1x)2255x4{51+x351+xln((1x)151x15)210+251x351+xarctan(10+25x154+(15)x15)+51+x351+xln(2+(1+5)x15+2x252+(15)x15+2x25)210251x351+xarctan(1025x154+(1+5)x15)}.

    F(1,25,y) and G(1,25,y). By employing the two equalities:

    Φ(x,1,25)=145x2{5ln((1x)151x15)+21025arctan(10+25x154+(15)x15)                +5ln(2+(15)x15+2x252+(1+5)x15+2x25)210+25arctan(1025x154+(1+5)x15)},Φ(x,1,35)=145x3{5ln((1x)151x15)21025arctan(10+25x154+(15)x15)                +5ln(2+(15)x15+2x252+(1+5)x15+2x25)+210+25arctan(1025x154+(1+5)x15)},

    we can establish the following two closed formulae:

    F(1,25,y)=3F2[1,1,1275,85|y]=65(1x){Φ(x,1,25)Φ(x,1,35)}=3(1x)105x3{5(1x15)ln(1x15(1x)15)+21025(1+x15)arctan(10+25x154+(15)x15)+5(1x15)ln(2+(1+5)x15+2x252+(15)x15+2x25)210+25(1+x15)arctan(1025x154+(1+5)x15)},G(1,25,y)=3F2[1,1,3275,85|y]=6(1x)225(1+x){Φ(x,1,25)+Φ(x,1,35)}=3(1x)2505x3{51+x151+xln((1x)151x15)210251x151+xarctan(10+25x154+(15)x15)+51+x151+xln(2+(15)x15+2x252+(1+5)x15+2x25)+210+251x151+xarctan(1025x154+(1+5)x15)}.

    F(1,18,y) and G(1,18,y). By utilizing the two explicit expressions:

    Φ(x,1,18)=128x{2ln(1+x181x18)+2ln(1+2x18+x1412x18+x14)+4arctan(x18)+22arctan(2x181x14)},Φ(x,1,78)=128x7{2ln(1+x181x18)+2ln(1+2x18+x1412x18+x14)4arctan(x18)22arctan(2x181x14)},

    we find the following two closed formulae:

    F(1,18,y)=3F2[1,1,1298,158|y]=748(1x){Φ(x,1,18)Φ(x,1,78)}                =7(1x)968x7{2(1x34)ln(1x181+x18)+4(1+x34)arctan(x18)                +2(1x34)ln(12x18+x141+2x18+x14)+22(1+x34)arctan(2x181x14)},G(1,18,y)=3F2[1,1,3298,158|y]=7(1x)264(1+x){Φ(x,1,18)+Φ(x,1,78)}                =7(1x)21288x7{21+x341+xln(1+x181x18)41x341+xarctan(x18)                +21+x341+xln(1+2x18+x1412x18+x14)221x341+xarctan(2x181x14)}.

    F(1,38,y) and G(1,38,y). By employing the two equalities:

    Φ(x,1,38)=128x3{2ln(1+x181x18)+2ln(12x18+x141+2x18+x14)4arctan(x18)+22arctan(2x181x14)},Φ(x,1,58)=128x5{2ln(1+x181x18)+2ln(12x18+x141+2x18+x14)+4arctan(x18)22arctan(2x181x14)},

    we deduce the following two closed formulae:

    F(1,38,y)=3F2[1,1,12118,138|y]=1516(1x){Φ(x,1,38)Φ(x,1,58)}                =15(1x)328x5{2(1x14)ln(1x181+x18)4(1+x14)arctan(x18)                +2(1x14)ln(1+2x18+x1412x18+x14)+22(1+x14)arctan(2x181x14)},G(1,38,y)=3F2[1,1,32118,138|y]=15(1x)264(1+x){Φ(x,1,38)+Φ(x,1,58)}                =15(1x)21288x5{21+x141+xln(1+x181x18)+41x141+xarctan(x18)                +21+x141+xln(12x18+x141+2x18+x14)221x141+xarctan(2x181x14)}.

    F(1,110,y) and G(1,110,y). By utilizing the two explicit expressions:

    Φ(x,1,110)=1410x{(55)ln(1+x1101x110)+(15)ln(1x1+x)+25ln(2+(1+5)x110+2x152(1+5)x110+2x15)                +210+25arctan(10+25x11022x15)+21025arctan(1025x11022x15)},Φ(x,1,910)=1410x9{(55)ln(1+x1101x110)+(15)ln(1x1+x)+25ln(2+(1+5)x110+2x152(1+5)x110+2x15)                210+25arctan(10+25x11022x15)21025arctan(1025x11022x15)};

    we establish the following closed formulae:

    F(1,110,y)=3F2[1,1,121110,1910|y]=980(1x){Φ(x,1,110)Φ(x,1,910)}                =9(1x)32010x9{25(1x45)ln(2(1+5)x110+2x152+(1+5)x110+2x15)                +(55)(1x45)ln(1x1101+x110)+210+25(1+x45)arctan(10+25x11022x15)                +(15)(1x45)ln(1+x1x)+21025(1+x45)arctan(1025x11022x15)},G(1,110,y)=3F2[1,1,321110,1910|y]=9(1x)2100(1+x){Φ(x,1,110)+Φ(x,1,910)}                =9(1x)240010x9{251+x451+xln(2+(1+5)x110+2x152(1+5)x110+2x15)                +(55)1+x451+xln(1+x1101x110)210+251x451+xarctan(10+25x11022x15)                +(15)1+x451+xln(1x1+x)210251x451+xarctan(1025x11022x15)}.

    F(1,310,y) and G(1,310,y). By employing the two equalities:

    Φ(x,1,310)=1410x3{(5+5)ln(1+x1101x110)+(1+5)ln(1x1+x)+25ln(2(1+5)x110+2x152+(1+5)x110+2x15)                21025arctan(10+25x11022x15)+210+25arctan(1025x11022x15)},Φ(x,1,710)=1410x7{(5+5)ln(1+x1101x110)+(1+5)ln(1x1+x)+25ln(2(1+5)x110+2x152+(1+5)x110+2x15)                +21025arctan(10+25x11022x15)210+25arctan(1025x11022x15)},

    we find the following two closed formulae:

    F(1,310,y)=3F2[1,1,121310,1710|y]=2140(1x){Φ(x,1,310)Φ(x,1,710)}                =21(1x)16010x7{25(1x25)ln(2+(1+5)x110+2x152(1+5)x110+2x15)                +(5+5)(1x25)ln(1x1101+x110)21025(1+x25)arctan(10+25x11022x15)                +(1+5)(1x25)ln(1+x1x)+210+25(1+x25)arctan(1025x11022x15)},G(1,310,y)=3F2[1,1,321310,1710|y]=21(1x)2100(1+x){Φ(x,1,310)+Φ(x,1,710)}                =21(1x)240010x7{251+x251+xln(2(1+5)x110+2x152+(1+5)x110+2x15)                +(5+5)1+x251+xln(1+x1101x110)+210251x251+xarctan(10+25x11022x15)                +(1+5)1+x251+xln(1x1+x)210+251x251+xarctan(1025x11022x15)}.

    F(1,112,y) and G(1,112,y). By utilizing the two explicit expressions:

    Φ(x,1,112)=1212x{ln((1x14)(1+x112)3(1+x14)(1x112)3)+3ln(1+3x112+x1613x112+x16)                +23arctan(3x1121x16)+2π+2arctan(3x1123x1414x16+x13)},Φ(x,1,1112)=1212x11{ln((1x14)(1+x112)3(1+x14)(1x112)3)+3ln(1+3x112+x1613x112+x16)                23arctan(3x1121x16)2π2arctan(3x1123x1414x16+x13)},

    we derive the following two closed formulae:

    F(1,112,y)=3F2[1,1,121312,2312|y]=11120(1x){Φ(x,1,112)Φ(x,1,1112)}                =11(1x)24012x11{(1x56)ln((1+x14)(1x112)3(1x14)(1+x112)3)+3(1x56)ln(13x112+x161+3x112+x16)                +2π(1+x56)+23(1+x56)arctan(3x1121x16)+2(1+x56)arctan(3x1123x1414x16+x13)},G(1,112,y)=3F2[1,1,321312,2312|y]=11(1x)2144(1+x){Φ(x,1,112)+Φ(x,1,1112)}                =11(1x)228812x11{1+x561+xln((1x14)(1+x112)3(1+x14)(1x112)3)+31+x561+xln(1+3x112+x1613x112+x16)                2π1x561+x231x561+xarctan(3x1121x16)21x561+xarctan(3x1123x1414x16+x13)}.

    F(1,512,y) and G(1,512,y). By employing the two equalities:

    Φ(x,1,512)=1212x5{ln((1x14)(1+x112)3(1+x14)(1x112)3)+3ln(13x112+x161+3x112+x16)                23arctan(3x1121x16)+2π+2arctan(3x1123x1414x16+x13)},Φ(x,1,712)=1212x7{ln((1x14)(1+x112)3(1+x14)(1x112)3)+3ln(13x112+x161+3x112+x16)                +23arctan(3x1121x16)2π2arctan(3x1123x1414x16+x13)},

    we find the following two closed formulae:

    F(1,512,y)=3F2[1,1,121712,1912|y]=3524(1x){Φ(x,1,512)Φ(x,1,712)}                =35(1x)4812x7{(1x16)ln((1+x14)(1x112)3(1x14)(1+x112)3)+3(1x16)ln(1+3x112+x1613x112+x16)                +2π(1+x16)23(1+x16)arctan(3x1121x16)+2(1+x16)arctan(3x1123x1414x16+x13)},G(1,512,y)=3F2[1,1,321712,1912|y]=35(1x)2144(1+x){Φ(x,1,512)+Φ(x,1,712)}                =35(1x)228812x7{1+x161+xln((1x14)(1+x112)3(1+x14)(1x112)3)+31+x161+xln(13x112+x161+3x112+x16)                2π1x161+x+231x161+xarctan(3x1121x16)21x161+xarctan(3x1123x1414x16+x13)}.

    A few explicit expressions for F(m,λ,y) and G(m,λ,y) are recorded as examples, in particular, those for λ=1/2.

    F(2,12,y)=3F2[1,1,3232,52|y]=38(1x)2{Φ(x,1,12)+Φ(x,1,32)},G(2,12,y)=3F2[1,2,3232,52|y]=3(1x)24(1+x)+3(1x)316(1+x){Φ(x,1,12)Φ(x,1,32)};F(3,12,y)=3F2[1,2,3232,72|y]=58(1x)2+1564(1x)3{Φ(x,1,12)Φ(x,1,52)},G(3,12,y)=3F2[1,2,5232,72|y]=5(3x)(1x)224(1+x)+5(1x)432(1+x){Φ(x,1,12)+Φ(x,1,52)};F(2,13,y)=3F2[1,1,3243,83|y]=518(1x)2{Φ(x,1,13)+Φ(x,1,53)},G(2,13,y)=3F2[1,2,3243,83|y]=5(1x)29(1+x)+5(1x)327(1+x){Φ(x,1,13)Φ(x,1,53)};F(2,14,y)=3F2[1,1,3254,114|y]=732(1x)2{Φ(x,1,14)+Φ(x,1,74)},G(2,14,y)=3F2[1,2,3254,114|y]=7(1x)216(1+x)+21(1x)3128(1+x){Φ(x,1,14)Φ(x,1,74)};F(2,25,y)=3F2[1,1,3275,135|y]=825(1x)2{Φ(x,1,25)+Φ(x,1,85)},G(2,15,y)=3F2[1,2,3265,145|y]=9(1x)225(1+x)+18(1x)3125(1+x){Φ(x,1,15)Φ(x,1,95)}.

    By combining the bisection series approach with Whipple's quadtratic transformation formulae, we succeeded in evaluating several remarkable 3F2(y)-series in terms of Lerch's transcendental function, including Yabu's results as very initial examples. However, the remaining problem is how to extend these methods to the generalized hypergeometric series of higher order. The interested reader is encouraged to make further attempts to evaluate the related series explicitly.

    Marta Na Chen: Computation, Writing, and Editing; Wenchang Chu: Original draft, Review, and Supervision. Both authors have read and agreed to the published version of the manuscript.

    The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

    The authors express their sincere gratitude to the two reviewers for their careful reading, positive comments, and valuable suggestions which improved the manuscript during revision.

    Prof. Wenchang Chu is the Guest Editor of special issue “Combinatorial Analysis and Mathematical Constants” for AIMS Mathematics. Prof. Wenchang Chu was not involved in the editorial review and the decision to publish this article. The authors declare no conflicts of interest.



    [1] M. Asakura, T. Yabu, Explicit logarithmic formulas of special values of hypergeometric functions 3F2, Commun. Contemp. Math., 22 (2020), 1950040. https://doi.org/10.1142/S0219199719500408 doi: 10.1142/S0219199719500408
    [2] M. Asakura, N. Otsubo, T. Terasoma, An algebro-geometric study of special values of hypergeometric functions 3F2, Nagoya Math. J., 236 (2019), 47–62. https://doi.org/10.1017/nmj.2018.36 doi: 10.1017/nmj.2018.36
    [3] W. N. Bailey, Generalized hypergeometric series, Cambridge University Press, Cambridge, 1935.
    [4] Y. A. Brychkov, Handbook of special functions: derivatives, integrals, series and other formulas, 1 Ed., New York: Chapman and Hall/CRC, 2008. https://doi.org/10.1201/9781584889571
    [5] M. N. Chen, W. Chu, Evaluation of certain exotic 3F2(1)-series, Nagoya Math. J., 249 (2023), 107–118. https://doi.org/10.1017/nmj.2022.23 doi: 10.1017/nmj.2022.23
    [6] W. Chu, Analytical formulae for extended 3F2-series of Watson–Whipple–Dixon with two extra integer parameters, Math. Comp., 81 (2012), 467–479.
    [7] I. M. Gessel, D. Stanton, Strange evaluations of hypergeometric series, SIAM J. Math. Anal., 13 (1982), 295–308. https://doi.org/10.1137/0513021 doi: 10.1137/0513021
    [8] F. J. W. Whipple, Some transformations of generalized hypergeometric series, Proc. London Math. Soc., 26 (1927), 257–272. https://doi.org/10.1112/plms/s2-26.1.257 doi: 10.1112/plms/s2-26.1.257
    [9] T. Yabu, Explicit logarithmic formulas for hypergeometric function 3F2, P.h.D. Thesiss, Hokkaido University, 2022.
  • This article has been cited by:

    1. Abdelhamid Zaidi, Saleh Almuthaybiri, Explicit evaluations of subfamilies of the hypergeometric function 3F2(1) along with specific fractional integrals, 2025, 10, 2473-6988, 5731, 10.3934/math.2025264
  • Reader Comments
  • © 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1002) PDF downloads(63) Cited by(1)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog