What is the inverse cosine of 2?

2 Answers
Nov 3, 2015

It does not exist.

Explanation:

The range of the cosine function is only from 1 to -1.
The curve doesn't go past these values in the y-axis (as you've mentioned cosine inverse).

Take a look at the cosine curve.
graph{cosx [-15.8, 15.79, -7.9, 7.9]}

Nov 3, 2015

For Real cosine this does not exist.
For Complex cosine: #cos^-1 (2) = i ln(2+sqrt(3))#

Explanation:

As a Real valued function of Real angles #cos(x): RR->[-1, 1]# does not take the value #2#, or indeed any value in #(-oo, -1) uu (1, oo)#, so #cos^(-1)(2)# is undefined.

The definition of #cos# can be extended to Complex numbers as follows:

#e^(ix) = cos(x) + i sin(x)#
#cos(-x) = cos(x)#
#sin(-x) = -sin(x)#

Hence:

#cos(x) = (e^(ix)+e^(-ix))/2#

Then we can define #cos(z):CC->CC# as follows:

#cos(z) = (e^(iz)+e^(-iz))/2#

Then #cos^-1 (2)# is a solution of #cos(z) = 2#, that is:

#(e^(iz)+e^(-iz))/2 = 2#

Let #t = e^(iz)#, then we have:

#(t+1/t)/2 = 2#

Hence:

#t^2-4t+1 = 0#

Hence:

#t = 2+-sqrt(3)#

That is:

#e^(iz) = 2+-sqrt(3)#

So:

#iz = ln(e^(iz)) = ln(2+-sqrt(3))#

So:

#z = ln(2+-sqrt(3))/i = +-i ln(2+sqrt(3))#

By convention, the principal value is the solution with positive coefficient of #i#, that is #i ln(2+sqrt(3))#

In fact, #cos(z) = 2# for #z = 2npi +-i ln(2+sqrt(3))# for any #n in ZZ#.