# Variable Matrices on Sage

I am using Sagemath 10. I wish to work with matrices whose entries are multivariate polynomial ring variables (nothing funky with the base ring, `R = PolynomialRing(QQ, 100, 'x'); R.gens()`

is what I am trying). I create a `matrix.zero(n,n)`

for reasonable `n`

(currently just 3) and then try to alter each entry manually. However, Sage complains that the entries can't be non-constant polynomials.

So I tried to provide the ring above as an argument by writing `matrix.zero(R, n,n)`

, but this generated the error

`TypeError: unbound method Parent.category() needs an argument`

I have also tried using `M=MatrixSpace(R,n,n); M(L)`

where `L`

is a list of appropriate design. It still complains about getting non-constant polynomial entries.

How do I circumvent this issue? I only plan to analyse the determinant of a specific kind of matrix. Based on this information, feel free to offer alternatives with more restricted functionality than just a matrix. Thank you!

like this

@FrédéricC : could you resend your comment as an answer (possibly with amplification/explanations ? This would be useful for the edification of future generations of this site users ;-)...

Before that, I must remark that it didn't resolve the issue. It complains about getting too many arguments. Which makes sense, because as far as I recall,

`matrix()`

takes in a list of lists and returns the corresponding matrix.You say "it complains". What command yields the complaint? The comment from @FrédéricC was intended (I believe) to address your first question: how to create a zero matrix that can be modified one entry at a time.

Okay, I am not sure what happened. The command

`matrix(R,n,n)`

seemed to be yielding this error. But when I ran the notebook today, stuff seems to be fine. As of this moment, I am unable to recover the original code I wrote which was causing issues. There is a small chance I just did not run the appropriate cells. Apologies.