[R] Solution to differential equation

[Ben Bolker]

Mike Marchywka <marchywka <at> hotmail.com> writes:


  [snip]

> > did you see my earlier post with link to wolfram integrator? Where i also
> > requested anyone wanting to get rid of a copy of G&R Integral Tables to
> > contact me off list since a dog really did eat mine? I think it came up
> > with "F" or hypergeometrics. The denominator reduces to something like
> > y*(a-y)^n depending on one choice of variables. Analytic of course is
> > in the eye of the beholder but if you can get it in terms of known
> > functions presumably you can find existing code to evaluate or
> > even better known identities for manipulation.
> >
> > http://www.mail-archive.com/r-help <at> r-project.org/msg120242.html
> >
> >

  Yes; as I recall it came up with 
<http://reference.wolfram.com/mathematica/ref/Hypergeometric2F1.html>

  The gsl package has this function, apparently -- it agrees with
Mathematica/Wolfram Alpha's Hypergeometric2F1 for a single set of
inputs (2,3,4,0.5), although apparently the algorithm that GSL has
only converges when |z|=1.