3
$\begingroup$

According to m.l.boas, one can solve integrations involving a simple pole using residue theorem and using those principles, I can solve integrals like

$$\int_{-\infty}^\infty \frac{\sin{x}}x dx$$

Converting to $$\frac{e^{iz}}z$$ and calculating residue at origin .

But what to do when we encounter something like a fractional pole .

If I am to do

$$\int_{-\infty}^{\infty} \frac{\sin{x^q}}{x^q}dx$$

Then we get a laurent series at origin with fractional power by converting to a complex integral in a similar way . How to proceed with such integrals ?

Apart from way of making use of residue theorem, I am looking for a way of solving this integral, too .

q may be a fraction

$\endgroup$
  • 2
    $\begingroup$ there is no such a thing as a fractional pole - one of the fundamental properties of holomorphic and meromorphic functions is that their special set (zeroes/poles) is discrete with all points of integral order (positive for zeroes, negative for poles); truly fractional powers (e.g square root) have non-discrete singularities in the plane and are not holomorphic/meromorphic on punctured discs but only on cut discs by say a ray (or analytic arc) through the origin $\endgroup$ – Conrad May 29 at 18:47
2
$\begingroup$

Converting to $\dfrac {\sin z^q}{z^q}$ and using the Taylor series for sin, we get: $\dfrac1{z^q}\sum_{n=0}^\infty (-1)^n\dfrac{(z^q)^{2n+1}}{n!}=\sum_{n=0}^\infty (-1)^n\dfrac{z^{2nq}}{n!}$. Thus the function is analytic.

$\endgroup$
  • $\begingroup$ Indeed. @OP - note this is a straightforward generalization of $\lim_{z \to 0} \frac{\sin z}{z} = 1$ by a simple change of variables $z^q \to z$ (since both go to zero simultaneously and the integrand is clearly analytic elsewhere). We have a removable singularity at the origin. $\endgroup$ – Brevan Ellefsen May 29 at 17:27
  • $\begingroup$ This is fine, but how can i proceed with the integral in complex plane ? That wont be possible in the way it was done for sin(x)/x, i mean, converting it to e^(ix)/x $\endgroup$ – user157588 May 30 at 3:42
  • $\begingroup$ Shouldn't you have converted it into $\dfrac{ \sin z}z$? $\endgroup$ – Chris Custer May 30 at 3:54
  • $\begingroup$ Then also it will not have a residue, the dx term, in terms of new variable t=x^q, shall have fractional powers and hence residue theorem wont be applicable . $\endgroup$ – user157588 May 30 at 7:57
  • $\begingroup$ As pointed out, there are no fractional exponents. $\endgroup$ – Chris Custer May 30 at 14:50

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.