I'm not sure if this helps, but you can create a polynomial of the type you want a bit simpler:
sage: var("x,y") (x, y) sage: Inds = CartesianProduct(range(5), range(4)) sage: sum([var("a"+str(i)+str(j))*x^i*y^j for i,j in Inds]) a43*x^4*y^3 + a33*x^3*y^3 + a42*x^4*y^2 + a23*x^2*y^3 + a32*x^3*y^2 + a41*x^4*y + a13*x*y^3 + a22*x^2*y^2 + a31*x^3*y + a40*x^4 + a03*y^3 + a12*x*y^2 + a21*x^2*y + a30*x^3 + a02*y^2 + a11*x*y + a20*x^2 + a01*y + a10*x + a00 Now you can reference them on the fly like this: sage: eval("a"+str(3)+str(2)) a32 On Wed, Jun 3, 2009 at 10:51 AM, James Parson <par...@hood.edu> wrote: > > Dear sage-support group, > > I am completely new to computer algebra systems and to computer > programming, and I hope you'll indulge the following beginner's > question. I was wondering if there is a simple way to create a > polynomial of degree d in x and y with symbolic coefficients in Sage. > Here is what I mean: if I were at the board in class, I might write > (in LaTeX transcription) something like > > P(x,y) = \sum_{i+j\leq d} a_{ij} x^i y^j, > > which I would view as an element of \mathbf{Z}[a_{ij},x,y]. I might > then impose some linear conditions on the a_{ij} by insisting that P > (x_t,y_t) = 0 for a list of points > > (x_1,y_1), (x_2,y_2), ... . > > Finally, I might solve the resulting system of linear equations. > > How would you recommend that I set up something like the a_{ij} and P > (x,y) in Sage? In order to make the question more definite, I > illustrate it with an example that I took from a lecture of Doron > Zeilberger on experimental mathematics. He proposed the question of > finding a polynomial of degree d in x and y that vanishes when x and y > are specialized to consecutive Fibonacci numbers. The lines below are > my attempt at a Sage version of his suggested computer search for a > likely solution (originally written in Maple). The program should take > the degree d as an input and then provide a parameterized family of > polynomials of degree d that are likely candidates. > > Here is what I came up with, after an enlightening afternoon of > studying computer manuals: > > d = 4 > e = d+1 > L = [] > M = [] > for i in range(e): > for j in range(e-i): > L.append('a_%s_%s' %(i,j)) > M.append([i,j]) > V = var(' '.join(L)) > P = sum(V[j]*x^(M[j][0])*y^(M[j][1]) for j in range(len(L))) > E = [P(x=fibonacci(n),y=fibonacci(n+1)) for n in range(1,len(V)+6)] > P.substitute(solve(E,V,solution_dict = True)[0]) > > I could not figure out how to create and reference the variables a_ > {ij} conveniently, and so I ended up with the strange lists V and M > above. Even though I got my polynomial P and solved the original > problem to my satisfaction, I still don't think I know how I would > have Sage do something like sum the a_{i.i+1} for 2i+1<=d. Is there a > better way to do this sort of thing? > > > Thanks for your help and indulgence, > > James Parson > > > > --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-support@googlegroups.com To unsubscribe from this group, send email to sage-support-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-support URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---