What is the Correct Integral of Inverse Cosecant?

  • Context: Undergrad 
  • Thread starter Thread starter Euler2718
  • Start date Start date
  • Tags Tags
    Integral Inverse
Click For Summary

Discussion Overview

The discussion revolves around the integral of the inverse cosecant function, specifically the expression ∫ arccsc(x) dx. Participants explore the integration process, particularly focusing on the integration by parts method and the evaluation of the resulting integral involving 1/√(x²-1).

Discussion Character

  • Technical explanation
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • One participant presents their integration by parts approach to evaluate ∫ arccsc(x) dx and arrives at an expression involving ln|√(x²-1)|.
  • Several participants challenge the correctness of the integral ∫ 1/√(x²-1) dx, asserting that it does not yield ln|√(x²-1)|.
  • Another participant inquires about how ∫ 1/√(x²-1) dx leads to ln(x + √(x²-1)), expressing confusion over the integration process.
  • A suggestion is made to use trigonometric substitution to solve the integral, which one participant acknowledges as a method they were unfamiliar with.
  • Another participant proposes an alternative method involving multiplying and dividing by (x + √(x²-1)) to simplify the integral, leading to the logarithmic result.
  • There is a mention of the necessity of including absolute value bars in the final expression.

Areas of Agreement / Disagreement

Participants generally agree on the validity of the integration by parts method but disagree on the evaluation of the integral ∫ 1/√(x²-1) dx. The discussion remains unresolved regarding the most appropriate method for evaluating this integral.

Contextual Notes

Participants express uncertainty about the integration techniques and the implications of their results. The discussion highlights the dependence on specific integration methods and the potential for confusion in evaluating logarithmic expressions.

Euler2718
Messages
90
Reaction score
3
Running into a little trouble when doing this integral by hand:

[tex]\int arccsc(x) dx[/tex]

[tex]u = arccsc(x) \implies du = -\frac{1}{x\sqrt{x^{2}-1}} dx[/tex]

[tex]dv = dx \implies v = x[/tex]

[tex]\int u dv = uv - \int vdu[/tex]

[tex]\int arccsc(x) dx = xarccsc(x) - \int x\cdot -\frac{1}{x\sqrt{x^{2}-1}} dx[/tex]

[tex]\int arccsc(x) dx = xarccsc(x) + \int \frac{1}{\sqrt{x^{2}-1}} dx[/tex]

Now I'm here, and it seems obvious , but this yields the answer to be:

[tex]\int arccsc(x) dx = xarccsc(x) + \ln\left|\sqrt{x^{2}-1}\right| +C[/tex]

When it's suppose to be (from wikipedia):

[tex]\int arccsc(x) dx = xarccsc(x) + \ln\left(x+\sqrt{x^{2}-1}\right) +C[/tex]

Where did I go wrong (skipped ahead in the course, so sorry if it's something really basic I'm missing)?
 
Physics news on Phys.org
That last integral is wrong.
##\frac{d}{dx}ln|\sqrt{x^2-1}|=\frac{x}{x^2-1}\neq\frac{1}{\sqrt{x^2-1}}##.
 
Last edited:
certainly said:
That last integral is wrong.
##\frac{d}{dx}ln|\sqrt{x^2-1}|=\frac{x}{\sqrt{x^2-1}}\neq\frac{1}{\sqrt{x^2-1}}##.

Yes, I'm aware. I'm wondering where I messed up in the integration by parts.
 
Morgan Chafe said:
Yes, I'm aware. I'm wondering where I messed up in the integration by parts.
The integration by parts is right.
However, ##\int \frac{1}{\sqrt{x^2-1}} dx \neq ln|\sqrt{x^2-1}|##.
 
certainly said:
The integration by parts is right.
However, ##\int \frac{1}{\sqrt{x^2-1}} dx \neq ln|\sqrt{x^2-1}|##.

hm..How exactly does ##\int \frac{1}{\sqrt{x^2-1}} dx ## boil down to ##\ln \left( x + \sqrt{x^{2}-1} \right) ## then? I'm a bit lost now.
 
Morgan Chafe said:
hm..How exactly does ##\int \frac{1}{\sqrt{x^2-1}} dx ## boil down to ##\ln \left( x + \sqrt{x^{2}-1} \right) ## then? I'm a bit lost now.
Hint:- Use trigonometric substitution.
 
  • Like
Likes   Reactions: Euler2718
certainly said:
Hint:- Use trigonometric substitution.

Thank you! I never heard of trigonometric substitution lol (like I said, I skipped ahead). Did a secant substitution, and got the real answer. Thank you.
 
if you do not want to use trigonometric subst or doing boring calculation, you can multiply and divide ##\left( 1/ \sqrt{x^{2}-1} \right) ## by

##\left( x + \sqrt{x^{2}-1} \right) ##

in this way the result is immediate! In fact you find

##\left( 1 + x/\sqrt{x^{2}-1} \right)/ \left( x + \sqrt{x^{2}-1} \right) ##

that is f'/f , so the result is ##\log \left( x + \sqrt{x^{2}-1} \right) ##panta rei
 
Last edited:
  • Like
Likes   Reactions: Euler2718
..and obviously with the absolute value bars...
 

Similar threads

  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 6 ·
Replies
6
Views
3K
  • · Replies 12 ·
Replies
12
Views
3K
  • · Replies 11 ·
Replies
11
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 6 ·
Replies
6
Views
3K
  • · Replies 8 ·
Replies
8
Views
4K