Subject: renaming undocumented Bessel functions g0, g1, gn
From: Raymond Toy
Date: Tue, 08 Aug 2006 13:04:22 -0400
>>>>> "Robert" == Robert Dodier <robert.dodier at gmail.com> writes:
Robert> On 8/7/06, Raymond Toy <raymond.toy at ericsson.com> wrote:
>> Why not call them scaled_bessel_i?
Robert> OK by me. In renaming g0 --> bessel_g0, etc, I was just
Robert> preserving the existing function names.
But since you're the first one I know of to complain about it, it
seems not particularly important to preserve the names.
However, maybe we should just get rid of them completely? Maxima
doesn't know any mathematical properties of these functions so it
seems their only utility is to evaluate Bessel I for large arguments
without overflowing.
If we told Maxima the properties of such functions, then it would make
sense to keep them.
Also, there seems to be a bug:
bessel_g0(10.0) -> 0.1278333...
bessel_gn(10.0,0) -> 0.1278333...
bessel_gn(10.0,0.0) -> exp(-10.0)*bessel_i(n,x)
The order of the arguments for bessel_gn is backward from bessel_i and
friends. And bessel_gn(10.0,0.0) is totally wrong. (A simple typo in
the code.) And the code for bessel_gn can evaluate it for real values
of the index, not just integer values.
Ray