How about first working on a simple case ... say k=0, x=0 ... then we have to do
Sum((-1)^n*GAMMA(n+1/2)^(1/2)/(GAMMA(1+n)^(1/2)*Pi^(1/2)), n = 0 .. infinity)
Can you do that one? The term goes to 0 very slowly, but we get convergence because it is alternating. I see no reason for a closed-form answer, though.
---
G A Edgar

I am also on Mac Pro, but not Firefox, and doesn't happen with me. Perhaps Firefox has some shortcut method to open a new tab?
---
G A Edgar

On my Mac I use a utility called "MenuMaster" to bind keystrokes to menu choices...
---
G A Edgar

Do you have some reason to think a closed-form solution exists?
---
G A Edgar

Look at the crazy answer to int(g,x) ... Now plug in infinity and -infinity and subtract, to get the answer 0.
---
G A Edgar

So I tried it. JavaViewLib comes from
http://www.javaview.de/maple/
Even though the web site has not been updated for two years, they mention only up to Maple 10, and do not mention Mac at all, still it worked (mostly) on my Mac with Maple 12. As a bonus ... Maple's own VRML output seems to do only VRML version 1 (which is 12 years out of date by now), but using JavaViewLib I can get VRML version 2 (aka VRML97) output files.
---
G A Edgar

I read the iPhone descriptions. "Mathomatic" seems to be the most available so far.
---
G A Edgar

If you have an iPhone, search "algebra" in the app store, and see if one of those does what you want.
Maybe their capabilities are minuscule compared to Maple, but so are their prices!
---
G A Edgar

I can't see Robert's images either ... the "alt" codes show
a[n]=1/Pi*Int(f(x)*cos(n*x),x=0..2*Pi)
b[n]=1/Pi*int(f(x)*sin(n*x),x=0..2*Pi)
---
G A Edgar

/ Library/Frameworks/Maple.framework/Versions/12/lib
so you probably need an admin password to put something there
---
G A Edgar

/ Library/Frameworks/Maple.framework/Versions/12/lib
so you probably need an admin password to put something there
---
G A Edgar

I think he wants his result to be a formula that works for all n ...
Something like this...
> diff(1/(1+x^2),x$n) assuming n::posint;
(n!/2)*(-exp(-(1/2)*n*ln(1+x^2))*sin(n*arctan(1, -x))*x+exp(-(1/2)*n*ln(1+x^2))*cos(n*arctan(1, -x))+exp(-(1/2)*n*ln(1+x^2))*sin(n*arctan(-1, -x))*x+exp(-(1/2)*n*ln(1+x^2))*cos(n*arctan(-1, -x)))/(1+x^2)
[I hope I got that right...]
---
G A Edgar

I think he wants his result to be a formula that works for all n ...
Something like this...
> diff(1/(1+x^2),x$n) assuming n::posint;
(n!/2)*(-exp(-(1/2)*n*ln(1+x^2))*sin(n*arctan(1, -x))*x+exp(-(1/2)*n*ln(1+x^2))*cos(n*arctan(1, -x))+exp(-(1/2)*n*ln(1+x^2))*sin(n*arctan(-1, -x))*x+exp(-(1/2)*n*ln(1+x^2))*cos(n*arctan(-1, -x)))/(1+x^2)
[I hope I got that right...]
---
G A Edgar

* It is already implemented in Mathematica. *
But what about SAGE? (I thought Sage was alec's favorite)
---
G A Edgar

* It is already implemented in Mathematica. *
But what about SAGE? (I thought Sage was alec's favorite)
---
G A Edgar