1
$\begingroup$

I recently learnt about inverse trigonometric functions. While studying, I came across a formula

$$\arcsin\left(2x\sqrt{1-x^2}\right)=2\arcsin x$$

The proof of this was given by substituting $x$ by $\sin\phi$. However if we substitute $x$ by $\cos\phi$, we can prove that the above expression equals $2\arccos x$. And indeed while going through another book I found that the above expression equals $2\arcsin x$ for $x$ between $-1/\sqrt{2}$ and $1/\sqrt{2}$ and $2\arccos x$ for $x$ between $1/\sqrt{2}$ and $1$. I graphed the function in GeoGebra and wanted to post the images but I can't. Links: arcsin(2x√1-x²) 2arcsin(x) 2arccos(x)

But it was clear that $\arcsin\left(2x\sqrt{1-x^2}\right)$ is not equal to $\arcsin x$ throughout its domain.

So the question is:why isn't $\arcsin\left(2x\sqrt{1-x^2}\right)=2\arcsin x$ throughout its domain? Why is there another definition for it?

Also as a bonus question, what can be a definition for the expression for $x$ between $-1/\sqrt{2}$ and $-1$?

Please explain in a simple way.

I'm new to high level mathematics that is common here in mathstackexchange :)

$\endgroup$
  • $\begingroup$ What is $\phi$? You could post a link to the images you mentioned. Somebody with sufficient reputation could then edit your question to post the images. $\endgroup$ – N. F. Taussig Sep 16 '18 at 17:25
  • $\begingroup$ I believe $\phi$ is a random name for a variable that satisfies $x=\sin\phi$ $\endgroup$ – AnotherJohnDoe Sep 16 '18 at 17:26
  • $\begingroup$ Hang on. I'll edit the question. $\endgroup$ – Garry Host Sep 16 '18 at 17:26
  • $\begingroup$ ø is an arbitrary variable $\endgroup$ – Garry Host Sep 16 '18 at 17:30
  • $\begingroup$ @GarryHost, See math.stackexchange.com/questions/1764431/… $\endgroup$ – lab bhattacharjee Sep 17 '18 at 10:03
1
$\begingroup$

Substituting $x = \sin \phi$ gives on the LHS, by trig addition theorem:

$\arcsin(2x\sqrt{1-x^2}) = \arcsin(2\sin \phi \cos \phi) = \arcsin(\sin ( 2\phi))$

Note that $\sin ( 2\phi)$ is increasing with $\phi$ only for $0 \le \phi \le \pi/4$ so

$\arcsin(\sin ( 2\phi)) = 2 \phi = 2 \arcsin(x)$ is only valid for the corresponding $0\le x \le \sin \pi/4 = \frac{1}{\sqrt 2}$ .

However, for $\pi/4 \le \phi \le \pi/2$, $\sin ( 2\phi)$ is decreasing, so substituting, $x = \sin \phi$ gives in this range $\frac{1}{\sqrt 2} \le x \le 1$ and

$\arcsin(2x\sqrt{1-x^2}) = \arcsin(2\sin \phi \cos \phi) = \arcsin(\sin ( 2\phi)) = \pi - 2 \phi = \pi - 2 \arcsin x = 2 \arccos x$

So the answer in your "other book" is accurate.

$\endgroup$
0
$\begingroup$

There are two issues. One is the $\pm$ sign in $x=\sin \phi\implies \sqrt{1-x^2}=\pm\cos\phi$. But even if the $\pm$ sign is a $+$ so that $\arcsin (2x\sqrt{1-x^2})=\arcsin\sin 2\phi$, that doesn't necessarily simplify to $2\phi$.

$\endgroup$
  • $\begingroup$ for arcsin(sin2ø) to simplify to 2ø, 2ø has to lie between -π/2 to π/2 right? $\endgroup$ – Garry Host Sep 16 '18 at 17:42
  • $\begingroup$ @GarryHost Right. $\endgroup$ – J.G. Sep 16 '18 at 18:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.