Taylor Series expansion

**zizou1089** · May 4th 2010, 09:19 AM

After working out the poles to:
$<br /> f(z)=\frac{1}{2z^{2}+(6-4i)z-12i}$

I have the answer z1=2i and z2=3

Given this how would we find the taylor series if:

$\left | z \right |< \left | z1 \right |$

**Failure** · May 4th 2010, 09:50 AM

Originally Posted by zizou1089

After working out the poles to:
$<br /> f(z)=\frac{1}{2z^{2}+(6-4i)z-12i}$

I have the answer z1=2i and z2=3

Shouldn't we have $z_2=-3$ , or have you given us the wrong $f(z)$ ? For the rest of this post I am assuming that your $f(z)$ is right, and that $z_2=-3$ .

Given this how would we find the taylor series if:

$\left | z \right |< \left | z1 \right |$

First you do a partial fraction decomposition by requiring that
$f(z)=\frac{a}{2i-z}+\frac{b}{3+z}=-\frac{1}{2(2i-z)(3+z)}$
which gives, if I am not misaken, that $a=b=-\frac{1}{6+4i}$ . (I might be mistaken, however, so please to the math yourself.)

Now you can develop each of these two fractions with the help of a geometric series like this:
$\frac{a}{2i-z}+\frac{b}{3+z} = \frac{a}{2i}\cdot \frac{1}{1-\frac{z}{2i}}+\frac{b}{3}\cdot \frac{1}{1-\left(-\frac{z}{3}\right)}$

$=\frac{a}{2i}\sum_{n=0}^\infty \left(\frac{z}{2i}\right)^n+\frac{b}{3}\sum_{n=0}^ \infty \left(-\frac{z}{3}\right)^n=\ldots$

**zizou1089** · May 4th 2010, 12:46 PM

Ok, wouldnt $\left |2i \right |$ =2?

**zizou1089** · May 4th 2010, 01:14 PM

**Failure** · May 4th 2010, 08:06 PM

Originally Posted by zizou1089

In that case, you need a Laurent-Series development, because the geometric series that we used for the first term of the partial fraction decomposition no longer converges for such values of z.

You can get it, starting from the same partial fraction decomposition, like this

$\frac{a}{2i-z}+\frac{b}{3+z} = -\frac{a}{z}\cdot \frac{1}{1-\frac{2i}{z}}+\frac{b}{3}\cdot \frac{1}{1-\left(-\frac{z}{3}\right)}$

$=-\frac{a}{z}\sum_{n=0}^\infty \left(\frac{2i}{z}\right)^n+\frac{b}{3}\sum_{n=0}^ \infty \left(-\frac{z}{3}\right)^n=\ldots$

**zizou1089** · May 5th 2010, 03:08 AM

Originally Posted by Failure

In that case, you need a Laurent-Series development, because the geometric series that we used for the first term of the partial fraction decomposition no longer converges for such values of z.

You can get it, starting from the same partial fraction decomposition, like this

$\frac{a}{2i-z}+\frac{b}{3+z} = -\frac{a}{z}\cdot \frac{1}{1-\frac{2i}{z}}+\frac{b}{3}\cdot \frac{1}{1-\left(-\frac{z}{3}\right)}$

$=-\frac{a}{z}\sum_{n=0}^\infty \left(\frac{2i}{z}\right)^n+\frac{b}{3}\sum_{n=0}^ \infty \left(-\frac{z}{3}\right)^n=\ldots$

Okay Im starting to get the hang of this, its just aint $\left | 2i \right |=2$ ? For the first part when we divide $\left | z \right |< \left | z1 \right |$ by $\left | z1 \right |$ we have

$\frac{\left | z \right |}{\left | 2i \right |}< 1$

which is equal to $\frac{\left | z \right |}{2}< 1$
therefore:

$\frac{a}{2}\cdot \frac{1}{1-\frac{z}{2}}+\frac{b}{3}\cdot \frac{1}{1-\left(-\frac{z}{3}\right)}$ ??

**Failure** · May 5th 2010, 05:58 AM

Originally Posted by zizou1089

Okay Im starting to get the hang of this,

Great!

its just aint $\left | 2i \right |=2$ ? For the first part when we divide $\left | z \right |< \left | z1 \right |$ by $\left | z1 \right |$ we have

$\frac{\left | z \right |}{\left | 2i \right |}< 1$

which is equal to $\frac{\left | z \right |}{2}< 1$
therefore:

$\frac{a}{2}\cdot \frac{1}{1-\frac{z}{2}}+\frac{b}{3}\cdot \frac{1}{1-\left(-\frac{z}{3}\right)}$ ??

Well, yes, the condition for the geometric series expansion $\frac{1}{1-z}=\sum_{n=0}^\infty z^n$ to converge is that $|z|<1$ holds.
Therefore, if $|z| < |2i|$ , we have that $\left|\frac{z}{2i}\right| < 1$ , and thus the series $\sum_{n=0}^\infty \left(\frac{z}{2i}\right)^n$ converges. But if $|2i|< |z|$ , we have $1<\left|\frac{z}{2i}\right|$ , thus $\sum_{n=0}^\infty \left(\frac{z}{2i}\right)^n$ diverges.

By a purely algebraic transformation of that first term of the partial fraction decomposition we can use the (for $|2i| < |z|$ ) convergent geometric series $\sum_{n=0}^\infty \left(\frac{2i}{z}\right)^n$ , instead.

Simililarly, if we were told to develop the given function for $|-3|<|z|$ , we would have to transform the second term of the partial fraction decomposition in such a way that we can develop it in a, for $\left|\frac{-3}{z}\right|<1$ , convergent series of the form $\sum_{n=0}^\infty\left(-\frac{3}{z}\right)^n$ .

**zizou1089** · May 5th 2010, 08:37 AM

Originally Posted by Failure

Great!

Well, yes, the condition for the geometric series expansion $\frac{1}{1-z}=\sum_{n=0}^\infty z^n$ to converge is that $|z|<1$ holds.
Therefore, if $|z| < |2i|$ , we have that $\left|\frac{z}{2i}\right| < 1$ , and thus the series $\sum_{n=0}^\infty \left(\frac{z}{2i}\right)^n$ converges. But if $|2i|< |z|$ , we have $1<\left|\frac{z}{2i}\right|$ , thus $\sum_{n=0}^\infty \left(\frac{z}{2i}\right)^n$ diverges.

By a purely algebraic transformation of that first term of the partial fraction decomposition we can use the (for $|2i| < |z|$ ) convergent geometric series $\sum_{n=0}^\infty \left(\frac{2i}{z}\right)^n$ , instead.

Simililarly, if we were told to develop the given function for $|-3|<|z|$ , we would have to transform the second term of the partial fraction decomposition in such a way that we can develop it in a, for $\left|\frac{-3}{z}\right|<1$ , convergent series of the form $\sum_{n=0}^\infty\left(-\frac{3}{z}\right)^n$ .

Ok I seem to be getting more questions correct, one last example Im having difficulty with is what if,

$|z-z1| < |z2-z1|$ ??

Thanks!

**Failure** · May 5th 2010, 09:18 AM

Originally Posted by zizou1089

Ok I seem to be getting more questions correct, one last example Im having difficulty with is what if,

$|z-z1| < |z2-z1|$ ??

Thanks!

In this case you cannot develop the series around $0$ anymore, but have instead to develop it around $z_1$ : which means, that you want to develop your function into Laurent-series that do not have terms of the form $a_n z^n$ but terms of the form $a_n (z-z_1)^n$ instead.

The first term of your partial fraction decomposition is already (almost) in that form, namely
$\frac{a}{2i-z}=-a(z-2i)^{-1}$ .

For the second term of your partial fraction decomposition you can substitute $(z-2i)+2i$ for $z$ and try to transform it into a power series in terms of $(z-2i)^n$ .

**zizou1089** · May 6th 2010, 05:34 PM

Originally Posted by Failure

In this case you cannot develop the series around $0$ anymore, but have instead to develop it around $z_1$ : which means, that you want to develop your function into Laurent-series that do not have terms of the form $a_n z^n$ but terms of the form $a_n (z-z_1)^n$ instead.

The first term of your partial fraction decomposition is already (almost) in that form, namely
$\frac{a}{2i-z}=-a(z-2i)^{-1}$ .

For the second term of your partial fraction decomposition you can substitute $(z-2i)+2i$ for $z$ and try to transform it into a power series in terms of $(z-2i)^n$ .

I dont seem to be getting the right answer for this last question, help needed please! Thanks in advance!

**Failure** · May 6th 2010, 07:58 PM

Originally Posted by zizou1089

I dont seem to be getting the right answer for this last question, help needed please! Thanks in advance!

I have only very little time for this (it's early in the morning here in Switzerland). So let me start with the partial fraction decomposition that we've already used successfully for the other cases and try to develop it as series of integral powers of $(z-z_1)$ , that is $(z-2i)$ :
$\frac{a}{2i-z}+\frac{b}{3+z} = -a(z-2i)^{-1}+\frac{b}{3+2i+(z-2i)} =-a(z-2i)^{-1}+\frac{b}{3+2i}\cdot\frac{1}{1-\left(-\frac{z-2i}{3+2i}\right)}$

$=-a(z-2i)^{-1}+\frac{b}{3+2i}\cdot\sum_{n=0}^\infty \left(-\frac{z-2i}{3+2i}\right)^n=\ldots$

Note that this series converges, as required, if $|z-2i| < |-3-2i|=|3+2i|$ , because in that case we have that $\left|-\frac{z-2i}{3+2i}\right|<1$

**zizou1089** · May 7th 2010, 05:18 AM

Originally Posted by Failure

I have only very little time for this (it's early in the morning here in Switzerland). So let me start with the partial fraction decomposition that we've already used successfully for the other cases and try to develop it as series of integral powers of $(z-z_1)$ , that is $(z-2i)$ :
$\frac{a}{2i-z}+\frac{b}{3+z} = -a(z-2i)^{-1}+\frac{b}{3+2i+(z-2i)} =-a(z-2i)^{-1}-\frac{b}{3+2i}\cdot\frac{1}{1-\left(-\frac{z-2i}{3+2i}\right)}$

$=-a(z-2i)^{-1}+\frac{b}{3+2i}\cdot\sum_{n=0}^\infty \left(-\frac{z-2i}{3+2i}\right)^n=\ldots$

Note that this series converges, as required, if $|z-2i| < |-3-2i|=|3+2i|$ , because in that case we have that $\left|-\frac{z-2i}{3+2i}\right|<1$

Thanks that helped alot.

Thread: Taylor Series expansion

LinkBack

Thread Tools

Display

Taylor Series expansion

Similar Math Help Forum Discussions

expansion of e^x in Taylor series

Taylor Series Expansion Around 0.25

Taylor series expansion

taylor series expansion for log(1-x/1+x)

Taylor Series Expansion of an eqn

Search Tags