On a Hankel transform integral containing an exponential function and two Laguerre polynomials (Q1919963): Difference between revisions

It seems that the article has a literal value and may be useful in the direction of various endeavour for integral transforms theory. The authors have cited the Hankel (or Bessel) transform integral which appeared in the books of \textit{I. S. Gradshteyn} and \textit{I. M. Ryzhik} [Table of integrals, series, and products (1980; Zbl 0521.33001), \textit{A. Erdélyi} et al. [Tables of integral transforms. Vol. II (1954; Zbl 0058.34103), \textit{F. Oberhettinger} [Tables of Bessel transforms (1972; Zbl 0261.65003), and \textit{A. P. Prudnikov} [Integrals and series, Vols. 1,2 (1986; Zbl 0733.00004)]. Through an elementary derivation, the authors exhibit the incorrectness in the formula No. 7.4222 in Gradshteyn-Ryzhik (loc. cit.) and present a corrected version of the formula \[ \begin{multlined} \int^\infty_0 x^{\nu+1} e^{-\alpha x^2} L_m^{\nu- \sigma} (\alpha x^2) L^\sigma_n (\alpha x^2) J_\nu (xy) dx =\\ {(-1)^{m+n} \over 2 \alpha} \left({y \over 2 \alpha} \right)^\nu \exp \left(-{y^2 \over 4 \alpha} \right) L_n^{\sigma - m + n} \left({y^2 \over 4 \alpha} \right) L_m^{\nu- \sigma +m-n} \left({y^2 \over 4 \alpha} \right),\end{multlined} \] \([y>0,\;\text{Re} \alpha > 0,\;\text{Re} \nu> -1]\) by making it numerically consistent, by interchanging the subscripts \(n\) and \(m\) of the two Laguerre polynomials appearing on the right side of above formula. In fact the authors claim that the formula \[ \int^\infty_0 x^{(\alpha + \beta)/2} e^{-x} L_m^\alpha (x) L_n^\beta (x) J_{\alpha + \beta} \bigl(2 \sqrt {ax} \bigr) dx = (-1)^{m+n} a^{(\alpha + \beta)/2} e^{-a} L_m^{\beta-m +n} (a) L_n^{\alpha + m-n} (a), \] \([\text{Re}(\alpha+\beta)>-1]\), originally given by \textit{A. Erdélyi} [J. Lond. Math. Soc. 13, 146-154 (1938; Zbl 0018.30403)], was correct.