Elliptic Integrals * request help *

[astropi]

Homework Statement

I need to write a program (or use one that's already written) to compute the elliptic integral of the 1st and 2nd kind - ie F and E respectively. Please note that this is NOT a homework assignment. Thus, I do not actually need to write the code, and will be happy to use something that's already written. (In fact, I would be surprised if this hasn't been done quite a lot already, I just can't seem to find any code/library)!

Homework Equations

No equations. I looked to see if I could find this code already implemented in C or Fortran, and no luck thus far. So, I decided to try and see if I could get my own code to work via Numerical Recipes in C. I wrote the code, and the code is giving me a value of 2, always. Clearly it's not working.

The Attempt at a Solution

I will be happy to e-mail someone my code if he/she is willing to look at it (it's written in C++). Consequently, if anyone knows if this is already implemented in some library, please let me know. Also, please note that I am a beginner programmer, so I apologize ahead of time if my questions are "simple"! Thank you.

 

[phyzguy]

Probably nobody is going to want to wade through your code and figure out what you did wrong. If it's always giving 2, you probably have something simple wrong. I suggest you simplify the problem to trying to calculate something easy (like x^2 or cos(x)), and get this working, then move on to more complicated functions. The Wolfram functions site is a great resource for formulas on functions like this:

http://functions.wolfram.com/

 

[awkward]

You might be able to use the Gnu Scientific Library:

http://www.gnu.org/software/gsl/

 

[RSW00001]

Not sure if you're looking for "complete" elliptic integrals or not. If you want the complete elliptic integrals of first and second kind, there's a very efficient algorithm using arithmetic-geometric means. I discuss it on my site, and have example code in Open Office Basic and Javascript.
http://electronbunker.sasktelwebsite.net/CalcMethods1a.html

 

[astropi]

Thanks everyone for your helpful replies. Shortly after I posted I figured out what I did wrong (isn't that always the case :)

I checked my results with Abramowitz and Stegun and they agree perfectly. I would eventually like to get the GNU library to work, I haven't done that yet.

cheers,

-astropi