Are complex logarithm rules the same as real number rules?

Click For Summary

Discussion Overview

The discussion centers on the rules of logarithms for complex numbers compared to those for real numbers, particularly focusing on the equation log(z) = -log(1/z) and the implications of branch cuts in complex analysis.

Discussion Character

  • Technical explanation
  • Debate/contested

Main Points Raised

  • One participant asserts that log(z) = -log(1/z) holds for real numbers, questioning if this extends to complex numbers.
  • Another participant agrees but notes exceptions regarding the logarithm of 0 or infinity.
  • A participant introduces the necessity of adding 2πi due to branch choosing issues in complex logarithms.
  • Further elaboration is provided on the branch choice for the logarithm, illustrating with examples that lead to different results depending on the chosen branch.
  • It is mentioned that for positive real numbers, the initial equation holds true, but complexities arise with negative values and branch cuts.
  • One participant emphasizes the importance of caution when applying familiar rules to complex numbers, as unexpected results can occur.

Areas of Agreement / Disagreement

Participants express differing views on the applicability of logarithm rules from real to complex numbers, with no consensus reached on the implications of branch cuts and specific cases.

Contextual Notes

The discussion highlights the limitations of applying real number logarithm rules to complex numbers without considering branch cuts and the implications of multi-valued functions in complex analysis.

daudaudaudau
Messages
297
Reaction score
0
Hi.

I know that for real numbers log(z)=-log(1/z)

is this also true in general for complex numbers?
 
Physics news on Phys.org
yup. no promises about the log of 0 or \infty.
 
But one might need to add 2\pi i somewhere sometimes because of some branch choosing issues.
 
If you choose to use a branch

<br /> \log(z) = \log(|z|) + i\textrm{arg}(z),\quad 0\leq \textrm{arg}(z) &lt; 2\pi<br />

then for example

<br /> \log(-1+i) = \log(\sqrt{2}) + \frac{3\pi i}{4}<br />

and

<br /> \log(\frac{1}{-1+i}) = \log(-\frac{1}{2}(1+i)) = \log(\frac{1}{\sqrt{2}}) + \frac{5\pi i}{4}.<br />

So you've got

<br /> \log(-1+i) + \log(\frac{1}{-1+i}) = 2\pi i,<br />

in contradiction with your equation. But if you choose the branch so that

<br /> -\pi &lt; \textrm{arg}(z) \leq \pi,<br />

then you've got

<br /> \log(-1+i) + \log(\frac{1}{-1+i}) = 0,<br />

as your equation stated. Even with this choice of branch still, for example,

<br /> \log(-1) + \log(\frac{1}{-1}) = 2\pi i,<br />

so actually...

daudaudaudau said:
Hi.

I know that for real numbers log(z)=-log(1/z)

for positive real numbers! :wink:
 
jostpuur said:
But one might need to add 2\pi i somewhere sometimes because of some branch choosing issues.


The essentail point is than ln(1)= 2n\pi i with n being any integer.
 
Thank you for the answers and examples. I understand it much better now.
 
You probably remember the trick where one does something like this:

<br /> 1 = \sqrt{1} = \cdots = -1<br />

with imaginary units. The examples I gave are very similar in nature. Most of the time, a blind use of familiar calculation rules might seem to work, but you never know when something tricky surprises you, if you are not careful.
 

Similar threads

  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 15 ·
Replies
15
Views
4K
  • · Replies 4 ·
Replies
4
Views
4K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 3 ·
Replies
3
Views
1K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 5 ·
Replies
5
Views
2K