First, going through what it means to be positive definite and non-singular:
Positive definite implies
- $\det(A) > 0$
- All eigenvalues of $A$ are positive, and so $0$ is not an eigenvalue of $A$
Nonsingular implies
- $\det(A) \neq 0$
- All eigenvalues of A are nonzero
- The product of eigenvalues of $A$ $= \det(A)$
It seems as though these two characterizations go hand in hand, though I assume negative eigenvalues could form a non-singular matrix but not a positive definite matrix. Can this be proven directly, or do I need to figure out how to prove by contradiction?
Thanks!
Best Answer
Here's a one-line proof by contraposition:
Suppose there is nonzero $x$ such that $Ax = 0$. Then $\langle x, Ax \rangle = 0$, and so $A$ is not positive definite.