I'm not too sure what your question is.
A basis of a matrix is the rows or columns that are linearly independent.
That is,
\(\displaystyle c1[\vec{v1}]+c2[\vec{v2}]+...+cn[\vec{vn}] = \vec{0}\) only has the trivial solution.
If you want to determine two different basis for the matrix A, you could find the basis for the sub-spaces row(A) and col(A).
Take your matrix and put it in RREF is the first step.
You can also do a check, rank(A) = dim(row(A)) = dim(col(A)).
And for a m x n matrix, nullity(A) + rank(A) = n