A binomial sum
Edwin Clark
eclark at math.usf.edu
Tue Apr 20 01:03:00 CEST 2004
Allowing Maple to think for me, generally
product(a*j+b,j=1..n) =
a^n*GAMMA((a*n+a+b)/a)/GAMMA((a+b)/a)
And also Maple gives:
product(b*j^2+c*j+d,j=1..n);
(n + 1) -c + %1 -c - %1 /
b GAMMA(n + 1 - 1/2 -------) GAMMA(n + 1 - 1/2 -------) /
b b /
/ -c + %1 -c - %1 \
|b GAMMA(1 - 1/2 -------) GAMMA(1 - 1/2 -------)|
\ b b /
2
%1 := sqrt(c - 4 b d)
But it has no answer for product(a*j^3+b*j^2+c*j+d,j=1..n).
--Edwin
On Tue, 20 Apr 2004, Karol PENSON wrote:
> The third product in Henry's mail can be expressed with the gamma
> function. Since the gamma function is as common tool as factorial
> I still believe that the second line below is a closed form of the first.
>
>
> simplify(product(3*j-2,j=1..n));
> (n + 1/2)
> 3 GAMMA(n + 1/3) GAMMA(2/3)
> 1/2 ------------------------------------
> Pi
>
> Cheers,
> Karol A. Penson
>
> On Mon, 19 Apr 2004, Henry Gould wrote:
>
> > Date: Mon, 19 Apr 2004 17:16:38 -0400
> > From: Henry Gould <gould at math.wvu.edu>
> > To: Sequence Fans Mailing List <seqfan at ext.jussieu.fr>
> > Subject: Re: A binomial sum
> >
> > Gentlemen -
> > If we tabulate ANY function, as e.g. a Table of Prime numbers, or admit a
> > Hypergeometric function as a tabulated function, then of course we have a
> > "closed form".
> > We can express the product of the even nunbers 2,4,6,...2n by the "closed
> > form" (2^n)n!.
> > We can express the product of the odd numbers 1,3,5,...2n-1 by the "closed
> > form" (2n)!/(2^n)n!,
> > but we cannot do the same with the product of 1,4,7,10,13,16,...,3n-2. Of
> > course if we admit another this as a tabulated function and assign it a
> > name, e.g. P(n,3), then we ipso facto have a "formula."
> > It is all relative to what we accept as an "elementary" function. My point
> > is that the sum in the query cannot be done with simple factorial products
> > and ratios.
> >
> > Cheers and Salud!
> >
> > Henry
> >
>
>
--
------------------------------------------------------------
W. Edwin Clark, Math Dept, University of South Florida,
http://www.math.usf.edu/~eclark/
------------------------------------------------------------
More information about the SeqFan
mailing list