9
$\begingroup$

A unimodular matrix $M$ is a square integer matrix having determinant $+1$ or $−1$. A totally unimodular matrix (TU matrix) is a matrix for which every square non-singular submatrix is unimodular. A totally unimodular matrix need not be square itself. Obviously, any totally unimodular matrix has only $0$, $+1$ or $−1$ entries.

Now suppose a $n\times n$ non-singular matrix $A$ is totally unimodular. Can we prove that $A^{-1}$ is also totally unimodular? Or if it is not correct, can we have a counterexample?

$\endgroup$
8
  • $\begingroup$ it seems that invertible TU matrices also form a group like unimodular matrices...but I don't have a proof yet. $\endgroup$
    – Suvrit
    Apr 19, 2013 at 19:13
  • $\begingroup$ Doesn't this follow from looking at the adjoint or cofactor marix? Gerhard "Ask Me About Binary Matrices" Paseman, 2013.04.19 $\endgroup$ Apr 19, 2013 at 22:01
  • 1
    $\begingroup$ S. Sra: Good luck with that proof. You might consider 1 1;0 1 meanwhile. Gerhard "Not Grouplike Under Matrix Multiplication" Paseman, 2013.04.19 $\endgroup$ Apr 19, 2013 at 22:11
  • $\begingroup$ @Gerhard: The inverse of the matrix that you've mentioned is also TU...so I don't understand your comment??? $\endgroup$
    – Suvrit
    Apr 19, 2013 at 22:25
  • $\begingroup$ S. Sra, I am suggesting that I don't know what group structure you are placing on the set of matrices. For the two notions of multiplication I considered, the matrix I gave does not help form a group. Gerhard "How Do You Multiply Them?" Paseman, 2013.04.19 $\endgroup$ Apr 19, 2013 at 22:52

1 Answer 1

14
$\begingroup$

The answer is yes, because if $B=A^{-1}$, then we have an equality between minors: $$B(I,J)=\pm\frac{A(J^c,I^c)}{\det A},$$ for every subsets $I,J\subset[[1,n]]$ of same cardinals. This formula generalizes that giving the entries of $A^{-1}$ in terms of minors of $A$. The $\pm$ sign is not essential to prove the stability of the TU class under inversion.

$\endgroup$
3
  • 1
    $\begingroup$ ah, good old Cramer's rule (essentially!) $\endgroup$
    – Suvrit
    Apr 19, 2013 at 23:03
  • 1
    $\begingroup$ Thank you very much. But I think the formula should be $B(I,J) = \pm \frac{A(J^c, I^c)}{\det A}$. $\endgroup$
    – qianchi
    Apr 20, 2013 at 2:50
  • $\begingroup$ @qianchi. Of course you're right. I'll edit. $\endgroup$ Apr 20, 2013 at 3:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.