Ramanujan's ``Lost'' notebook. VIII: The entire Rogers-Ramanujan function (Q702705): Difference between revisions
From MaRDI portal
Set profile property. |
Set OpenAlex properties. |
||
Property / full work available at URL | |||
Property / full work available at URL: https://doi.org/10.1016/j.aim.2004.03.012 / rank | |||
Normal rank | |||
Property / OpenAlex ID | |||
Property / OpenAlex ID: W2022765896 / rank | |||
Normal rank |
Revision as of 19:13, 19 March 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Ramanujan's ``Lost'' notebook. VIII: The entire Rogers-Ramanujan function |
scientific article |
Statements
Ramanujan's ``Lost'' notebook. VIII: The entire Rogers-Ramanujan function (English)
0 references
17 January 2005
0 references
In Ramanujan's ``lost'' notebook it is claimed that \[ \sum_{n = 0}^{\infty} \frac{a^nq^{n^2}}{(q;q)_n} = \prod_{n = 1}^{\infty} \left(1 + \frac{aq^{2n-1}}{1-q^ny_1 - q^{2n}y_2 -q^{3n}y_3 - \cdots} \right), \] where \[ \begin{aligned} y_1 &= \frac{(q;q^2)_{\infty}^2}{(1-q)(q^2;q^2)_{\infty}^2}, \\ y_2 &= 0, \\ y_3 &= \frac{(q+q^3)(q;q^2)_{\infty}^2}{(1-q)(1-q^2)(1-q^3)(q^2;q^2)_{\infty}^2} - \frac{(q;q^2)_{\infty}^6 \sum_{n \geq 0} \frac{(2n+1)q^{2n+1}}{1-q^{2n+1}}}{(1-q)^3(q^2;q^2)_{\infty}^6}, \text{and} y_4 &= y_1y_3. \end{aligned} \] Here we make the usual assumption that \(| q| < 1\) and use the standard notation \[ (a;q)_n = \prod_{k=0}^{n-1} (1-aq^k). \] Adjectives like ``bizarre,'' ``disturbing,'' and ``enigmatic'' that are attributed by Andrews to this identity are well-justified. It is completely different from a typical \(q\)-series identity like \[ \sum_{n=0}^{\infty} \frac{a^nq^{n^2}}{(q;q)_n} = \frac{1}{(aq;q)_{\infty}} \left(1 + \sum_{n=1}^{\infty} \frac{(aq;q)_{n-1}(-1)^na^{2n}q^{n(5n-1)/2}(1-aq^{2n})}{(q;q)_n} \right) \] Moreover, it is not at all clear what the ``...'' in the denominator of the right hand side means. The mystery is deftly solved in the present paper using properties of entire functions, theta series, \(q\)-series, and the Stieltjes-Wigert orthogonal polynomials. What Ramanujan is actually presenting is the product representation of the left hand side considered as an entire function in the variable \(a\). It turns out that if \(z_n\) denotes the \(n\)th zero of this function, then \[ z_n = -q^{-2n+1}(1-\sum_{i=1}^{\infty}y_iq^{ni}), \] where the \(y_i\) are uniquely determined by the relation \[ \sum_{j=0}^N \frac{(-1)^{N-j}q^{N-j+1 \choose 2}}{(q;q)_{N-j}} \sum_{k_1 + 2k_2 + \cdots + jk_j = j} \frac{(-1)^{k_1+\cdots+k_j} y_1^{k_1} \cdots y_j^{k_j} \theta_{N-j+1,k_1+\cdots+k_j}}{k_1! \cdots k_j!} = 0. \] Here \(\theta_{m,k}\) is defined by \[ \theta_{m,k} = \sum_{n = -\infty}^{\infty} (-1)^nq^{n^2+mn}n(n-1)\cdots(n-k+1). \] For Part VII, cf. Adv. Math. 89, No. 1, 60--105 (1991; Zbl 0739.11042), Part IX, cf. Adv. Math. 191, No. 2, 408--422 (2005; Zbl 1067.11061)
0 references
Ramanujan's ``lost'' notebook
0 references
Rogers-Ramanujan function
0 references
Stieltjes-Wigert polynomials
0 references
entire functions
0 references