Determination of logarithm (complex)

Click For Summary

Discussion Overview

The discussion revolves around the computation of a complex integral involving a polynomial and its logarithmic derivative. Participants explore the implications of determining the logarithm of a polynomial along a closed path that contains all its zeroes, focusing on the necessity of specifying a branch of the logarithm in this context.

Discussion Character

  • Exploratory, Technical explanation, Debate/contested

Main Points Raised

  • One participant questions the need for a determination of log(p(z)) on the closed path C, suggesting it may pose a problem.
  • Another participant proposes that the necessity of specifying a branch of the logarithm depends on subsequent steps in the computation.
  • A third participant elaborates on differentiating log(p(z)) and presents a series representation involving the zeros of p(z) and their multiplicities.
  • There is a suggestion that specifying a branch of the logarithm may not be necessary since it is used to establish an algebraic identity.
  • One participant expresses agreement with the previous point, indicating a shared understanding.

Areas of Agreement / Disagreement

Participants express differing views on the necessity of specifying a branch of the logarithm for the integral computation, indicating that the discussion remains unresolved.

Contextual Notes

The discussion highlights potential limitations regarding the determination of logarithmic branches and the implications for the integral's evaluation, but these aspects remain unresolved.

quasar987
Science Advisor
Homework Helper
Gold Member
Messages
4,796
Reaction score
32
The question is this:

Consider p(z) a polynomial and C a closed path containing all the zeroes of p in its interior. Compute

\frac{1}{2\pi i}\int_C z\frac{p'(z)}{p(z)}dz

The solution given by the manual starts by saying that

\frac{p'(z)}{p(z)}=(log(p(z)))'.


But there is no determination of log(p(z)) on C. Isn't that a problem?
 
Physics news on Phys.org
I imagine it would depend on what you do next.
 
Next we differentiate log(p(z)), it gives

z\frac{p'(z)}{p(z)} = \sum_{j=1}^k \frac{zn_j}{z-zj}

where zj are the zeros of p(z) and nj their relative order of multiplicity.

Then we use the residue theorem on the integral ansd use the fact that each zj is a simple pole for \sum_{j=1}^k \frac{zn_j}{z-z_j} such that

\sum_{j=1}^k Res(z\frac{p'(z)}{p(z)},z_j) = \sum_{j=1}^k \lim_{z\rightarrow z_j}(z-z_j)\frac{zn_j}{z-z_j} = \sum_{j=1}^k z_jn_j

to get the result.
 
Last edited:
So do you really need to specify a branch of log on the entire curve? You're only using it to prove an algebraic identity.
 
I hear you!
 

Similar threads

  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 4 ·
Replies
4
Views
4K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 3 ·
Replies
3
Views
2K
Replies
7
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K