The integral I am trying to evaluate is:

$$I = \int_{-\infty}^\infty \frac{x}{1+x^2}\sin x\ dx = \int_{-\infty}^\infty f(x)\ dx$$

The standard approach to this is to realise $\sin x$ as the complex part of $e^{ix}$ and take a contour integral over a semicircle on the upper half-plane. I have no trouble doing this and getting the (correct) answer of $I = \frac\pi e$.

However, I see no reason why one shouldn't be able to do the aforementioned contour integral directly, without switching $\sin x$ for the exponential. When, I do this, I get a different answer, and I'm not sure where I'm going wrong.

The integrand has a single simple pole on the upper half-plane at $z=i$. We can compute the residue:

$$\text{Res}(f, i) = \lim_{z\to i}(z-i)f(z) = \lim_{z\to i}\frac{z}{z+i}\sin z = \frac12i\sinh(1) = \frac i4(e-\frac1e)$$

By the residue theorem, we get an unexpected answer:

$$I = 2\pi i\ \text{Res}(f, i) = \frac\pi 2(\frac1e - e)$$

What's the issue here?

  • 3
    $\begingroup$ Have you checked that the contribution to the integral from the semicircle goes to 0 as $r \rightarrow \infty$? $\endgroup$ – user_of_math May 27 '16 at 16:33
  • $\begingroup$ Ah. I thought I had, but I see now that this is not in fact the case. Thank you! $\endgroup$ – Guy Paterson-Jones May 27 '16 at 18:23

This interesting question reminds us that Residue theorem is correct only when $$ \lim_{z\to a+i\infty}f(z) = 0. $$


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.