1

solving one of the 5 options would be much appreciated as this will give me an idea on how to solve the rest.

Let $\gamma(w,R)$ denote the circular contour $t\mapsto w+Re^{it}$ where $0\lt t\lt2\pi$. Evaluate $$\int_\gamma\dfrac1{1+z^2}dz$$ when $\gamma$ is: $$\gamma(1,1),\quad\gamma(i,1),\quad\gamma(-i,1),\quad\gamma(0,2),\quad\gamma(3i,\pi).$$

Hakim
  • 10,213
mike
  • 35

1 Answers1

2

The hard way

For example, and by definion (without Cauchy's theorems, residues and stuff), taking $\;\gamma:=\gamma(-i,1)\;$

$$z \in\gamma\implies z=-i+e^{it}\;,\;\;t\in [0,2\pi)\;,\;\;dz=ie^{it}dt$$

and the integral thus is

$$\oint_\gamma\frac1{1+z^2}dz=\int_0^{2\pi}\frac{i\;e^{it}\;dt}{1+(-i+e^{it})^2}=\left.\arctan(-i+e^{it})\right|_0^{2\pi}=$$

$$=\left.\frac12i\;\text{Log}\frac{1-(-i+e^{it})}{1+(-1+e^{it})}\right|_0^{2\pi}=\frac12i\left(\log\left|\frac{i}{i}\right|+i\arg 1-\log|1|-i\arg1+2\pi ik\right)=$$

$$=-\pi k\;,\;\;\;k\in\Bbb Z$$

The easy way:

Using Cauchy's Theorems, and since $\;\frac1{z-i}\;$ analytic on and within $\;\gamma\;$:

$$\oint_\gamma\frac{dz}{1+z^2}=\oint_\gamma\frac{\frac1{z-i}}{z+i}dz=\left.2\pi i\frac1{z-i}\right|_{z=-i}=2\pi i\frac1{-2i}=-\pi$$

Timbuc
  • 34,191
  • Hi using, the "hard way" i managed to get to $[arctan(-i + exp(it)]_0^{2\pi}$ however i got that, that equals zero when putting the values in, i dont understand how you went to logs from there in order to get $\pi k$ .... Also for the "easy way" could you explain which part you're solving? as you seem to have a different answer to the "hard way"... Also i seem to be getting zero for every question – mike Nov 20 '14 at 02:13
  • @mike Read the definition of complex arctangent in formula (1) here :http://mathworld.wolfram.com/InverseTangent.html , and remember that the argument is defined only up to multipe of $;2\pi ;$ . As for the easy part: it's exactly the same integral as in the hard part, and the answer is the same...up to an integer multiple of $;2\pi;$ . If we choose the usual principal value (as in the real case), where the angle is in $;(-\pi,,,\pi];$ we get $;-\pi k;$ ( I just realized ther was missing a minus sign in the result of the hard part) – Timbuc Nov 20 '14 at 03:20
  • I have to ask since this makes me cringe. Why do you insist on using limits in displayed math environments for your integrals? – dustin Nov 20 '14 at 03:29
  • @dustin, I think it looks better: comparing $$\int_0^1 =\int\limits_0^1;$$ I always used, and still use, to write the limits directly under/over the integral sign when doing maths with paper and pencil. That's all. – Timbuc Nov 20 '14 at 03:32
  • 1
    Hmmm...looking more closely, perhaps it doesn't look that better. I guess I was "affected" by how the older mathematics formats used to write the integral too small and the limits would look too messy and close to the center part of that "s". I'm seriously considering to change. – Timbuc Nov 20 '14 at 03:37
  • I think int without limits has a cleaner look. – dustin Nov 20 '14 at 03:45
  • @dustin, I think you're probably right, and I'll begin by editing my answer and getting rid of that limits thing. – Timbuc Nov 20 '14 at 03:47
  • hopefully not all of your 495 answers have an integral in it then. – dustin Nov 20 '14 at 03:48
  • Hehe...@dustin, definitely not! Thanks for making me note this. – Timbuc Nov 20 '14 at 03:48