Hi,

I have a Matrix whose entries are polynomials in several formal parameters (the matrix is sparse and the polynomials are rather simple, though inverses of the parameters may also arise).
Then, when I compute the kernel with LinearAlgebra-NullSpace, maple naturally gives a basis of solutions over the same ring of polynomials.

Now for some reason there are some parameters that I don't want to see in the solutions (all but two of them, actually).

How can I compute the part of the kernel that lives in $\mathbb{Z}[a,a^{-1}, b,b^{-1}]$, i.e. that involves only the first two parameters?

Thank you,

NoThik

Edit : the coefficients of the polynomials are integers, and I expect the kernel elements to have integer coefficients as well.

Hi,

I have a linear problem A*X = B, with dimensions of A approximately 500*1300, and B is a vector with only one nonzero coordinate.

I feed it to LinearSolve, and there are a *lot* of solutions - presented as one vector with linear polynomial coordinates.

How can I get only one (hopefully with as many zeros as possible) ?

Thank you in advance.

NoThik

Hi,

I wrote a procedure and saved it into a .mla library, then accidentally lost the file containing the initial code. Can I get it back from the library ? I didn't find the answer in the help on LibraryTools.

Thank you.

Hi,
I need to solve a big linear system (sparse, if it helps). I first used the "Matrices" and it worked perfectly, but then the matrix got bigger and I got the message "assigning to a long list, please use Arrays". So I learned how to manipulate Arrays and basically wrote the whole program again, but I couldn´t find how to solve A.X=B when A and B are Arrays. And it seems like nobody talks about that on the internet :( please don´t tell me that in 2012, with...