4
$\begingroup$

I'm looking to calculate $a\int_{-\infty }^{\infty } \frac{e^{\frac{x^2}{a^2+x^2}}}{a^2+x^2} \,\operatorname dx$.

Mathematica10 can't integrate this, but numerical integration gives an answer of 5.50843 for any $a$ I've tried, so I assume that there is an analytic answer to be found.

Is this a standard integral?

Many thanks,

$\endgroup$
2
  • $\begingroup$ Have you tried using complex analysis? $\endgroup$ – UserX Sep 22 '14 at 22:26
  • 3
    $\begingroup$ Perhaps try substituting $x=a \tan{\theta}$. $\endgroup$ – Ron Gordon Sep 22 '14 at 22:39
4
$\begingroup$

Sub $x=a \tan{\theta}$ and the integral becomes

$$\sqrt{e} \int_{-\pi/2}^{\pi/2} d\theta \, e^{-\frac12 \cos{2 \theta}}$$

which is indeed a standard integral, equal to $\pi \sqrt{e} I_0(1/2) \approx 5.50842977...$, where $I_0$ is the modified Bessel function of the first kind of zeroth order.

$\endgroup$
2
  • $\begingroup$ Very nice. Can you point me to a resource that lists that standard integral? I also need the result for the more general integral where the exponent is to the power +1/2.b.cos2θ, where b is a positive constant. $\endgroup$ – Dave Sep 22 '14 at 23:12
  • 1
    $\begingroup$ @Dave See Formula 9.6.16 in Abramowitz and Stegun. $\endgroup$ – Dilip Sarwate Sep 22 '14 at 23:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.