We take out the results and test the output

> l:=%: The output is a sequence with the Frobenius form and the change matrix.

> F:=l[1];

> P:=l[2];

> evalm(inverse(P)&*A&*P-F);