[ Home ] [ Up ] [ Info ] [ Mail ]

Prove: Taylor’s theorem. Let f(z) be analytic inside a circle C centered at a and let z be a point inside C. Then


ole.gif


Proof. Let z be any point inside C. Construct a circle C1 inside C with center at a and enclosing z. Let w be a point on C1. See Fig. 1.


Our first step is to prove the following lemma. 


Lemma. For any three variables w, a and z the following identity holds:


ole1.gif


                                                 ole2.gif  

ole3.gif

Proof. First we write


ole4.gif                                                                          

Now multiply both numerator and denominator of the right member by the fraction 1/(w - a) to get


ole5.gif

We now wish to employ the following identity:


ole6.gif


Where does this identity come from? It comes from the formula for the sum of a geometric progression

 

5)        a + ar + ar2 + ... + arn-1 = a(1- rn)/(1- r) ,


an identity valid for both real and complex numbers.



Letting a = 1 in 5) we get

 

6)        1 + r + r2 + ... + rn-1 = 1/(1 - r) - rn/(1- r)

 

which, with a transposition, is 4).


Now let r = (z - a)/(w-a) and substitute into 4)



ole7.gif


                                                             ole8.gif


We now substitute 7) into 3) to obtain



ole9.gif


                                                 ole10.gif  


which is 1).


End of Proof.


Using the above lemma we will now proceed to prove our main theorem.


By Cauchy’s integral formula we have


ole11.gif


Substituting 1) into 8) we get


ole12.gif



                                                             ole13.gif


where


ole14.gif


Using Cauchy’s integral formulas


ole15.gif


9) becomes


ole16.gif  


                                                             ole17.gif  


We now wish to show that ole18.gif . Doing so will conclude the proof. 


First note that


             ole19.gif


where ole20.gif is a constant. Second, note that on C1, the function f(w) must acquire some finite maximum value i.e. it is bounded. Why? Because C1 lies in an analytic region and so there can be no singular points on C1 where the function might become infinite. Let M be the maximum of f(w) on C1. The radius r1 of C1 is given by r1 = |w - a|. We will also make use of the following observation:


            |w - z| = |(w - a) - (z - a)| ole21.gif r1 - |z - a|



We now employ Property 5 of line integrals which states that

 

             ole22.gif


where |f(z)| ole23.gif M ( i.e. M is an upper bound of |f(z)| on C) and L is the length of C. The length of C1 is 2πr1 so


ole24.gif


Because ole25.gif < 1 it can be seen that


             ole26.gif


since


             ole27.gif




References

  Spiegel. Complex Variables (Schaum)

  Wylie. Advanced Engineering Mathematics


[ Home ] [ Up ] [ Info ] [ Mail ]