2

We know that $$e_q(z)=\sum_{j\geq 0} \frac{z^j}{(q;q)_j}=\frac{1}{(z;q)_{\infty}}$$

where $(a;q)_{\infty}=\prod_{i=0}^{\infty}(1−aq^i)$ denotes the q-shifted factorial.

The limit between the $q$-exponential and the ordinary exponential is

$$\lim_{q\rightarrow1}e_q((1-q)z)=\lim_{q\rightarrow1}\frac{1}{((1-q)z;q)_{\infty}}=e^z.$$

My question is how to prove

$$\lim_{q\rightarrow1}\frac{1}{((1-q)\sqrt{1-q^2}x;q)_{\infty}}=e^{\sqrt{2}x}?$$

I'm glad for your help.

othmane
  • 95

2 Answers2

4

We know that $(q;q)_n=(1-q)^n[n]_q!\;$ and $\lim_{q\to1} [n]_q!=n!.\quad$ Therefore $$e_q((1-q)(1+q)x) = \sum_{n=0}^\infty \frac{(1-q)^n(1+q)^nx^n}{(q;q)_n}=\sum_{n=0}^\infty \frac{(1+q)^n x^n}{[n]_q!}.$$ The limit as $q\to1$ gives $e^{2x}$ which is justified because of absolute convergence.

Somos
  • 37,457
  • 3
  • 35
  • 85
0

Mathematica can not find the limit lim

QExponential[z_, q_] := 1/QPochhammer[z, q];
lim=Limit[1/(QPochhammer[(1 - q) Sqrt[1 - q^2]*x, q]), q -> 1]

By the way, I don't think the proposed limit is correct.

QExponential[z_, q_] := 1/QPochhammer[z, q];
Limit[1/(QPochhammer[(1 - q) Sqrt[1 - q^2]*x, q]), 
   q -> 0.9999] /. {x -> 3} // N  (* 1.04334 *)
Exp[Sqrt[2]*x] /. {x -> 3} // N   (* 69.5914 *)