-
Let
be a linear operator on a vector space . A polynomial is called a minimal polynomial for if is a monic polynomial of least positive degree for which Similarly, for matrices this polynomial is the monic polynomial of least degree such that for
-
(Friedberg 7.11) Let
be a minimal polynomial for a linear operator on a finite dimensional vector space . - If
is any polynomial for which , then divides . In particular divides the characteristic polynomial of . is unique. - Intuition: This immediately follows from applying the division algorithm.
- If
-
(Friedberg 7.12) Let
be a linear operator on a finite-dimensional vector space and let be a basis for . Then the minimal polynomial for is the same as the minimal polynomial for . - Intuition: Immediately follows from Linear Transformation Matrix Isomorphism.
-
(Friedberg 7.12.1) Let
, the minimal polynomial for is the same as the minimal polynomial for . -
(Friedberg 7.13) Let
be a linear operator on a finite dimensional vector space and let be the minimal polynomial for . A scalar is an eigenvalue of if and only if . Hence the characteristic polynomial and the minimal polynomial for
have the same zeros. -
(Friedberg 7.13.1) Let
be a linear operator on a finite dimensional vector space with minimal polynomial and characteristic polynomial . Suppose that factors as Where
are distinct eigenvalues of each with corresponding multiplicities . -
(Friedberg 7.14) Let
be a linear operator on an -dimensional vector space . If is a -cyclic subspace of itself, then the characteristic polynomial and the minimal polynomial for are of the same degree. Hence -
(Friedberg 7.15) Let
be a linear operator on a finite-dimensional vector space . Then is diagonalizable if and only if the minimal polynomial for is of the form Where
are distinct scalars (eigenvalues) -
(Friedberg e7.3.12) Let
be a linear operator on a finite dimensional vector space and suppose that the characteristic polynomial of splits. Let be the distinct eigenvalues of and for each , let be the order of the largest Jordan block corresponding to in a Jordan canonical form. The minimal polynomial of is
Annihilators
- Let
be a linear operator on a finite dimensional vector space and let . The polynomial is called a -annihilator of if is monic, of least degree, and - (Friedberg e7.3.15) Let
be a linear operator on a finite dimensional vector space and let . - The vector
has a unique -annihilator. - The
-annihilator of divides any polynomial for which - If
is the -annihilator of and is the -cyclic subspace generated by , then is the minimal polynomial for and - The degree of the
-annihilator of is if and only if is an eigenvector of .
- The vector