4
$\begingroup$

I think I proved the following proposition. Is this correct and well-known?

Proposition: Let $V$ be a finite dimensional vector space over a field $K$. Let $f: V \to V$ a $K$-linear map. Let $K[X]$ be the polynomial ring. $V$ can be regarded as a $K[X]$-module via $f$. If the characteristic polynomial of $f$ is the minimal polynomial, then $V$ is a cyclic $K[X]$-module, i.e. generated by a single element of $V$ over $K[X]$.

$\endgroup$
6
  • $\begingroup$ Yes, this is a well-known result. $\endgroup$ Commented Apr 14, 2012 at 13:31
  • $\begingroup$ Is there any reference? $\endgroup$ Commented Apr 14, 2012 at 14:11
  • $\begingroup$ This should be covered in virtually any graduate-level algebra book. Check the section on modules, or the material leading up to the proof of Jordan canonical form. $\endgroup$
    – Jim Belk
    Commented Apr 14, 2012 at 16:13
  • $\begingroup$ It will also appear in the material leading up to the Rational Canonical Form; e.g., in Friedberg, Insel, and Spence $\endgroup$ Commented Apr 14, 2012 at 21:39
  • 2
    $\begingroup$ @JimBelk "This should be covered in virtually any graduate-level algebra book." I have Bourbaki and Lang on algebra. I'm afraid I cannot find it covered by them. Would you please tell me the names of some books that do so? $\endgroup$ Commented Dec 10, 2015 at 19:09

1 Answer 1

5
$\begingroup$

This is a consequence of the structure theorem for finitely generated modules over a principal ideal domain $R$, which says that every such module can be written a direct sum of cyclic modules $\bigoplus_{i=1}^lR/(d_i)$ where the generators $d_i$ (invariant factors) divide one another in order: $d_1\mid d_2\mid \cdots \mid d_l$. Take the module to be $V$ as in the question, then the final generator $d_l\in K[X]$ annihilates all the cyclic factors, so (assuming the $d_i$ are taken to be monic) it is the minimal polynomial of $f$, while the characteristic polynomial is the product of all polynomials $d_i$. Clearly the two are equal if and only if $l=1$ (the invariant factors $d_i$ are never $1$) which is the case if and only if $V$ is a cyclic module.

See also this answer to a related question.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .