>>>>> "Barton" == Barton Willis <willisb at unk.edu> writes:
Barton> -----Raymond Toy wrote: -----
>>
Barton> assoc_legendre_p(n,0,x) - legendre_p(n,x),
Barton> assoc_legendre_q(n,0,x) - legendre_q(n,x).
>>
>> Is that supposed to be an equal sign? Is that
>> conversion supposed to happen automatically? If so, it doesn't for me.
Barton> Oh--make that assoc_legendre_p(n,0,x) - legendre_p(n,x) = 0. And
Barton> no, orthopoly doesn't simplify assoc_legendre_p(n,0,x) to legendre_p(n,x)
Barton> even when it has been told that n is a positive integer. Actually,
Barton> orhthopoly knows very few simplifications.
Oh, ok.
Barton> As hyp and friends get better, maybe all of orthopoly, except maybe
Barton> the numerical stuff, can be expunged or blended in with hyp.lisp.
Barton> Or maybe hyp.lisp and friends are already better than orthopoly.
I think think orthopoly and hyp serve two different purposes. hyp is
(currently) meant to simplify hypergeometric functions into "simpler"
functions like Legendre and other orthogonal polynomials. It knows
very little else.
Ray