# General advice on unsolvable equations

• Nov 4th 2007, 09:09 AM
Optiminimal
http://img232.imageshack.us/img232/9...mulaud2.th.jpg
t and horax are angles.
r, MC and VM are distances.

I'd like to solve for t, but math softwares tell me they find no solution.

While I'd very much appreciate a solution to the specific problem at hand, I am also very interested in how one should proceed on occasions like this. What I know to do is to make guesses and close in on a solution with the bisection method (or Newton Raphson if tehre is a derivative). But are there other approaches?

Could one use some statistics approach where one samples a lot of guesses on different parameter values and then fit new and simple functions t = f(horax,r,MC,VM) to that data?

• Nov 4th 2007, 06:58 PM
topsquark
Quote:

Originally Posted by Optiminimal
http://img232.imageshack.us/img232/9...mulaud2.th.jpg
t and horax are angles.
r, MC and VM are distances.

I'd like to solve for t, but math softwares tell me they find no solution.

While I'd very much appreciate a solution to the specific problem at hand, I am also very interested in how one should proceed on occasions like this. What I know to do is to make guesses and close in on a solution with the bisection method (or Newton Raphson if tehre is a derivative). But are there other approaches?

Could one use some statistics approach where one samples a lot of guesses on different parameter values and then fit new and simple functions t = f(horax,r,MC,VM) to that data?

Is it possible that the software is being literal and telling you that there are no real solutions? (Such as in the case of $x^2 + 1 = 0$. This one "has no solution" as well.)