Register to reply

Sum to Product Trigonometric identity does not work

Share this thread:
CraigH
#1
Apr27-13, 02:34 PM
P: 198
Hi,

The identity

[itex]sin(u) + sin(v) = 2 * sin (\frac{u+v}{2}) * cos(\frac{u-v}{2})[/itex]
http://en.wikipedia.org/wiki/List_of...uct_identities

Does not always work. I put the equation :

[itex](sin(u) + sin(v)) - (2 * sin (\frac{u+v}{2}) * cos(\frac{u-v}{2}))[/itex]

With u equal to -4.1 and v equal to 99 into wolfram alpha and it gave me the answer -1.11022x10^-16
http://www.wolframalpha.com/input/?i...9%2F2%29%29%29

If the identity is true, shouldn't the answer always be 0?

What's going on here?

Thanks
Phys.Org News Partner Mathematics news on Phys.org
'Moral victories' might spare you from losing again
Fair cake cutting gets its own algorithm
Effort to model Facebook yields key to famous math problem (and a prize)
HS-Scientist
#2
Apr27-13, 02:44 PM
P: 181
It seems like a rounding error to me. You can verify that the identity is in fact always true by using the half angle and angle addition formulae.
micromass
#3
Apr27-13, 02:49 PM
Mentor
micromass's Avatar
P: 18,036
The problem is not that the formula doesn't work, but with the fact that your calculator is incapable of precisely calculating the sine or cosine of an angle.

CraigH
#4
Apr27-13, 02:56 PM
P: 198
Sum to Product Trigonometric identity does not work

Ah okay, thank you for answering.
One thing though... If wolfram alpha knows that it can only calculate the sine or cosine of an angle to a certain precision, shouldn't it give the final answer to that precision, or less, so that it avoids giving misleading answers like the one it gave me.
micromass
#5
Apr27-13, 02:58 PM
Mentor
micromass's Avatar
P: 18,036
Quote Quote by CraigH View Post
Ah okay, thank you for answering.
One thing though... If wolfram alpha knows that it can only calculate the sine or cosine of an angle to a certain precision, shouldn't it give the final answer to that precision, or less, so that it avoids giving misleading answers like the one it gave me.
It would be better if they did that. But I've never seen a calculator doing it. They rather count on the users to know about the fallibility of the program.


Register to reply

Related Discussions
In binary can we have a value with deci centi mili or more lower valued prefix? Computers 14
Proving the contracted epsilon identity Differential Geometry 2
Proving the Jacobi identity from invariance Differential Geometry 2
How to work out frictional torque and resisting motion Introductory Physics Homework 1