Integration result ln || - confusing and apparent contradictions

Click For Summary

Discussion Overview

The discussion revolves around the integral of sec x, specifically the expression ln|sec x + tan x| + c, and the implications of the modulus sign in this context. Participants explore the nature of real integrals, the behavior of the logarithm function, and the conditions under which the integral is defined, touching on both theoretical and practical aspects of integration.

Discussion Character

  • Technical explanation
  • Debate/contested
  • Mathematical reasoning

Main Points Raised

  • One participant questions the necessity of the modulus sign in the integral of sec x, noting that the ln operator can yield complex values for negative arguments.
  • Another participant emphasizes that real integrals cannot take complex values, discussing the derivative of ln|sec x + tan x| and the implications of the signum function in this context.
  • A participant points out that the expression sec x + tan x can be negative or zero when cos x is negative, raising concerns about the definition of the logarithm in these cases.
  • One participant references the integral of 1/x and its even nature, suggesting that the use of absolute values in the antiderivative is a way to ensure the function remains defined across its domain.

Areas of Agreement / Disagreement

Participants express differing views on the role of the modulus sign and the conditions under which the integral of sec x is defined. There is no consensus on the necessity of the modulus or the implications of negative values for sec x and tan x.

Contextual Notes

Participants note that the relationship between sec x and tan x may lead to complications when cos x is negative, which affects the validity of the logarithm in certain intervals. The discussion highlights the importance of understanding the conditions under which these mathematical expressions are valid.

spaghetti3451
Messages
1,311
Reaction score
31
The integral of sec x is ln|sec x + tan x| + c. That's what the internet is telling me.

Now, I am having trouble understanding the modulus sign. I understand that the ln operator gives real numbers for positive values of the argument and complex numbers for negative values of the argument. So, if the ln operator is defined for negative values, why incorporate the modulus sign?

Also, I differentiated ln |sec x + tan x| + c and I get |sec x|. Doesn't that mean that the integral of sec x is only defined for positive values of sec x? But that contradicts common sense when we have a look at the graph of sec x (google 'graph sec x' for a display of the graph). It's clear the function does have an integral for negative values of sec x, right?
 
Physics news on Phys.org
Real integrals can't take complex values, to be precise.
Differentiating the expression you gave, we get this: We apply the chain rule once to get the derivative is
[tex]\frac{1}{\sec x + \tan x}\frac{d}{dx}|\sec x + \tan x|[/tex]
Now, we apply the chain rule once more. The derivative of the absolute value function is the signum function, defined everywhere but at 0. Hence this yields [itex]\mathrm{sgn}(\sec x + \tan x)\sec(x)[/itex] as the derivative. Now, the expression inside the signum function is positive or zero because of this trigonometric simplification:

[tex]\sec x + \tan x = \frac{1+\sin x}{\cos x} = \frac{1+\sin x}{\sqrt{1-\sin^2 x}} \geq 1+\sin x \geq 0[/tex]

Hence we are allowed to discard the signum function (for it will be 1) and rewrite the derivative as [itex]\sec(x)[/itex].

I am curious how you obtained the result you did. To obtain a more general answer, try differentiating and integrating a general reciprocal.
 
Millennial said:
Real integrals can't take complex values, to be precise.
Differentiating the expression you gave, we get this: We apply the chain rule once to get the derivative is
[tex]\frac{1}{\sec x + \tan x}\frac{d}{dx}|\sec x + \tan x|[/tex]
Now, we apply the chain rule once more. The derivative of the absolute value function is the signum function, defined everywhere but at 0. Hence this yields [itex]\mathrm{sgn}(\sec x + \tan x)\sec(x)[/itex] as the derivative. Now, the expression inside the signum function is positive or zero because of this trigonometric simplification:

[tex]\sec x + \tan x = \frac{1+\sin x}{\cos x} = \frac{1+\sin x}{\sqrt{1-\sin^2 x}} \geq 1+\sin x \geq 0[/tex]

Hence we are allowed to discard the signum function (for it will be 1) and rewrite the derivative as [itex]\sec(x)[/itex].

I am curious how you obtained the result you did. To obtain a more general answer, try differentiating and integrating a general reciprocal.

One problem. [itex]\cos x = \sqrt{1-\sin^2 x}[/itex] is only true when [itex]\cos x[/itex] is positive.
 
I'd like to remind you that if cos x is negative, due to the identity [itex]|\sec x|\geq |\tan x|[/itex], the expression [itex]\sec x + \tan x[/itex] is negative or zero: values where the real logarithm is not defined.
 
How about going back to ∫1/x dx = ln|x| + C?
My calc textbook said that since 1/x is odd, ∫1/x dx is even, so you can take the antiderivative to be ln|x| + C with absolute values to make it even. Then you can apply it to the problem with ∫secx dx.
 

Similar threads

  • · Replies 8 ·
Replies
8
Views
4K
  • · Replies 14 ·
Replies
14
Views
4K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 7 ·
Replies
7
Views
2K
Replies
5
Views
3K
  • · Replies 19 ·
Replies
19
Views
5K