Help please! Complex analysis.
Evaluate the contour integral via residue theorem. cos (z)/1-2sin(z)
z0=pi/6
Over what contour?
circle of R=3/2
Centered at the origin?
yes
found the residue at z0=pi/6 r(pi/6)=-sqrt3/2
Alright, so by the residue theorem the integral should evaluate to \(2\pi i\) times the residue, right?
right but the book says the answer is -(pi)(i)
which would mean my residue would have to -1/2
In that case you must have made a mistake with the residue computation. What approach did you use to determine it?
simple pole method
Here's an approach involving the series expansions: \[(z-\pi/6)\frac{\cos z}{1-2\sin z}=\frac{\color{red}1(z-\pi/6)-\frac{(z-\pi/6)^3}{2!}+\frac{(z-\pi/6)^4}{4!}-\cdots}{1\color{red}{-2}\left((z-\pi/6)-\frac{(z-\pi/6)^3}{3!}+\frac{(z-\pi/6)^5}{5!}-\cdots\right)}\to -\frac{1}{2}\]
but can I also do \[\lim_{z \rightarrow \pi/6} (z-z_0)f(z)\]
with out expansion
Perhaps it's a matter of semantics, but around \(z_0=\dfrac{\pi}{6}\), the series expansions IS the same as the function itself.
sorry i'm terrible with that equation maker
No worries, it comes with practice :) Evaluating the limit \[\lim_{z\to\pi/6}\left(z-\dfrac{\pi}{6}\right)\frac{\cos z}{1-2\sin z}\] can be a bit unwieldy without considering the expansions...
(sinz-1/2) (cosz/-2(sinz-1/2)) =cosz at pi/6
wouldn't your (z-z0)= sinz-1/2
No, the definition of residue for a function \(f(z)\) at a (simple) pole \(z_0\) is given as \[\newcommand{res}{\text{Res}} \res\{f(z),z_0\}=\lim_{z\to z_0}(z-z_0)f(z)\]
You seem to be replacing \(z-z_0\) with \(\sin z-\sin(z_0)\).
oh you're right!
so in general then, with sins and cos and e^iz you would have to expand then
right?
I wouldn't say that's the case in general, but it's certainly a good way to think about approaching the computation. For example, if you're given the function \(f(z)=\dfrac{e^z}{z}\), you can tell right away that there's a simple pole at \(z=0\), and so \[\res\left\{\frac{e^z}{z},0\right\}=\lim_{z\to0}z\frac{e^z}{z}=\lim_{z\to0}e^z=1\] with no expansion needed. Of course, if you had elected to use the expansion, yo'd get \[\res\left\{\frac{e^z}{z},0\right\}=\lim_{z\to0}\frac{z+z^2+\frac{z^3}{2!}+\cdots}{z}=1\]
I see, but what about with sin and cos in the denominator
you would at that point have to expand right
I'm only going to answer yes because that's how I would do it :)
I'm sure there are other methods available for that sort of thing.
lol is there any other way to do it?
I know r(z0)=g(z0)/h'(z0)
There is another way that involves Laurent series expansions (a sort of extension of Taylor series). These sorts of series have the form \[\sum_{k=-\infty}^\infty a_k(z-c)^k\] and it turns out that the \(k=-1\) term gives the residue. If I recall correctly, you can also use Cauchy's integral formula. Both these methods are closely related.
oh yeah, now I remember. This was the first way we learned it. Okay, thanks so much for your help!
Think you can answer one more quick question?
Possibly. I have to admit it's been a while since I've taken complex analysis but I'll do my best.
okay, why is it that b^-1 is the residue. More clearly, why is the coefficient in front of 1/(z-z0) the residue? does that mean left over?
Consider the definition given here: http://en.wikipedia.org/wiki/Laurent_series You find that \[a_{-1}=\frac{1}{2\pi i}\oint f(z)\,dz\] I remember learning that evaluating this integral is the same as the residue, but I forget the theory behind it.
the version I know is f(a) =1/2ipi integrate f(z) dz
I know the theory behind that but not sure that translate to a coefficient. Anyways , I really appreciate the help man!
You're welcome!
Take care
Join our real-time social learning platform and learn together with your friends!