Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Let $A = \textrm{diag}(a_{1},a_{2},a_{3})$ where $a_{1},a_{2},a_{3}$ are distinct. I am trying to show that every diagonal $3\times3$ matrix cane be made up of linear combinations of $I$, $A$ and $A^{2}$.

I was thinking of writing this matrix $$\begin{pmatrix} 1 & a_{1} & a_{1}^{2}\\ 1 & a_{2} & a_{2}^{2}\\ 1 & a_{3} & a_{3}^{2} \end{pmatrix}$$

and finding its determinant, then using to distinct value property to show that the determinant is non zero, but I didn't seem to get too far with that. I was wondering is there any other way I can approach this?

share|improve this question

3 Answers 3

up vote 1 down vote accepted

I think you have the right idea.

Firstly, the set of diagonal $3\times 3$ matrices constitutes a three-dimensional vector space.

Let $V$ be a three-dimensional vector space and let $\{e_1,e_2,e_3\}$ be a basis for $V$. If $v_1,v_2,v_3$ are elements of $V$, then the assertion that $\{v_1,v_2,v_3\}$ is a basis for $V$ is tantamount to the non-degeneracy of the matrix

$$\begin{pmatrix} a_{11} & a_{12} & a_{13} \\ a_{21} & a_{22} & a_{23} \\ a_{31} & a_{32} & a_{33} \end{pmatrix}$$

if $v_j=\sum_{i=1}^{3} a_{ij} e_i$ is the expansion of $v_j$ in terms of the basis $\{e_1,e_2,e_3\}$.

Now, let us consider the specific case of $V$ being the three-dimensional vector space of diagonal matrices with "standard" basis

$$e_1=\begin{pmatrix} 1 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 0 \\ \end{pmatrix}$$

$$e_2=\begin{pmatrix} 0 & 0 & 0 \\ 0 & 1 & 0 \\ 0 & 0 & 0 \\ \end{pmatrix}$$

$$e_3=\begin{pmatrix} 0 & 0 & 0 \\ 0 & 0 & 0 \\ 0 & 0 & 1 \end{pmatrix}.$$

The matrix you wrote down

$$\begin{pmatrix} 1 & a_1 & a_1^2 \\ 1 & a_2 & a_2^2 \\ 1 & a_3 & a_3^2 \end{pmatrix}$$

is exactly the matrix you need to check for nondegeneracy. In fact, this matrix is the so-called Vandermonde matrix and is nondegenerate if and only if the $a_i$'s are distinct (which is the assumption that you have been given!). (The determinant of the Vandermonde matrix, i.e., the Vandermonde determinant, is $\prod_{1\leq i<j\leq 3} (a_j-a_i)$.)

Hope this helps!

share|improve this answer

You're right: determinant of this matrix is non zero for distinct $a_1,a_2,a_3$, because it's Vandermonde matrix.

share|improve this answer

You can do it as you suggested by using the (well known) evaluation of a Vandermonde determinant, but there is an in my opinion easier approach.

The question is whether $I$, $A$, and $A^2$ span the space of diagonal matrices. Since that space is of dimension$~3$, this is condition equivalent to $I$, $A$, and $A^2$ linearly independent. Suppose to the contrary they are linearly dependent, so $c_0I+c_1A+c_2A^2=0$ for some $(c_0,c_1,c_2)\neq(0,0,0)$. Then $c_0+c_1X+c_2X^2$ is a nonzero polynomial of degree at most$~2$ for which each diagonal entry $a_i$ of $A$ is a root (since that is what $c_0I+c_1A+c_2A^2=0$ means). But such a polynomial can have at most $2$ distinct roots, a contradiction.

share|improve this answer
    
I really like this answer, thanks! –  Johnmgee Jul 5 at 15:21

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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