Is there an algebraic explanation for the identity atan(a/b) + atan(b/a) = pi/2?

  • Context: Undergrad 
  • Thread starter Thread starter Nick89
  • Start date Start date
Click For Summary
SUMMARY

The identity atan(a/b) + atan(b/a) = pi/2 is proven using the properties of right triangles and the tangent function. By defining tan(α) = a/b and tan(β) = b/a, it follows that tan(α + β) = (tan(α) + tan(β)) / (1 - tan(α)tan(β)). This results in a denominator of zero, indicating that α + β = pi/2. The algebraic proof confirms the identity holds true when (a, b) > 0.

PREREQUISITES
  • Understanding of trigonometric functions, specifically tangent and arctangent.
  • Familiarity with the properties of right triangles.
  • Basic knowledge of limits and handling indeterminate forms.
  • Experience with algebraic manipulation of equations.
NEXT STEPS
  • Study the properties of the tangent function and its relationship with angles in right triangles.
  • Learn about limits and how to handle indeterminate forms in calculus.
  • Explore the use of Maple for symbolic computation and algebraic proofs.
  • Investigate other trigonometric identities and their geometric interpretations.
USEFUL FOR

Mathematicians, students studying trigonometry, educators teaching geometry, and anyone interested in algebraic proofs of trigonometric identities.

Nick89
Messages
553
Reaction score
0
Hey,

I came across this 'identity' today and was wondering if there was any algebraical explanation to this...

Basically I had to show:
\arctan( \frac{a}{b}) + \arctan( \frac{b}{a}) = \frac{\pi}{2}
(And if necessary, (a,b) > 0 )

It is pretty easy to show when you draw a right triangle:
Code:
     /|
    /c|
   /  |
  /   |a
 /d   |
/_____|
   b
(a and b are the sides while c and d are the angles)

Now, \tan d = \frac{a}{b} and \tan c = \frac{b}{a} and because it is a right triangle, c + d = \frac{\pi}{2} = \arctan( \frac{a}{b}) + \arctan( \frac{b}{a}).But I was wondering if you can also proof this algebraically?
I typed it into Maple and got \frac{1}{2} \text{signum}(\frac{a}{b}) \pi. If my memory serves me well, signum is always 1 if a/b > 0 and -1 if a/b < 0, so if a/b > 0 this holds...

So yeah, just wondering... I can't see any way to do it algebraically...
 
Mathematics news on Phys.org
Let tan\alpha =\frac{a}{b} and tan\beta =\frac{b}{a}

Then consider

tan(\alpha + \beta)=\frac{tan\alpha +tan\beta}{1-tan\alpha tan\beta}

and what that gives.
 
Thanks, I understand now!

\tan \alpha = \frac{a}{b} \text{ , } \tan \beta = \frac{b}{a}
\arctan \frac{a}{b} + \arctan \frac{b}{a} = \alpha + \beta

\tan(\alpha + \beta) = \frac{ \frac{a}{b} + \frac{b}{a} }{ 1 - \frac{ab}{ab}} = \frac{ \frac{a}{b} + \frac{b}{a} }{0} = \infty*

If \tan(\alpha + \beta) = \infty then \alpha + \beta = \frac{\pi}{2}.

*Can you say this so easily? Shouldn't you handle the divide by zero better? I know if a denominator tends to zero, the fraction tends to infinity, but you can't divide by 0 exactly... I also can't see any way I could take a limit here?
 

Similar threads

  • · Replies 11 ·
Replies
11
Views
2K
Replies
2
Views
1K
  • · Replies 13 ·
Replies
13
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 14 ·
Replies
14
Views
2K
  • · Replies 13 ·
Replies
13
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K