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.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I think (hope) I'm on the right track with this problem, but there are details that I can't seem to work out. I've also struggled to find examples of this sort of problem to assist me.

$\rho (\mathbf{x}, \mathbf{y}) = \sqrt{(\mathbf{x} - \mathbf{y})^T \mathbf{A} (\mathbf{x} - \mathbf{y})}$,
where $\mathbf{x}=(x_1,\ldots,x_n)$, $\mathbf{y}=(y_1,\ldots,y_n)$, and $\mathbf{A}$ is an $n \times n$ symmetric positive-definite real matrix. Taking $\varepsilon$ to be the ordinary Euclidean metric, show that $(\mathbb{R}^n,\varepsilon)$ and $(\mathbb{R}^n,\rho)$ are isometric.

My thought process so far: I need to show that $\exists f:\mathbb{R}^n \rightarrow \mathbb{R}^n$, which is surjective, such that
$\rho(f(\mathbf{x}),f(\mathbf{y})) = \varepsilon(\mathbf{x},\mathbf{y})$.
What I'm thinking is to use Sylvester's Law of Inertia, so that
$(f(\mathbf{x})-f(\mathbf{y}))^T \mathbf{A} (f(\mathbf{x})-f(\mathbf{y})) = \sum_{i=1}^n w_i^2$.
This would reduce the equation to $\sum_{i=1}^n w_i^2 = \sum_{i=1}^n (x_i - y_i)^2$, but I'm too tentative to assert that $w_i = x_i-y_i$ for some $f$.

How should I proceed? Or am I completely off-track here? Please be brutal.

share|cite|improve this question
up vote 2 down vote accepted

Positive definite real symmetric matrix must be of the form $\mathbf{A}=\mathbf{P}\mathbf{P}^T$ for some regular matrix $\mathbf P$. (Since it's diagonal form is the identity matrix $\mathbf{I}$. The matrix $\mathbf{P}$ can be even chosen to be orthogonal.)

Now define $f(\mathbf{x})=(\mathbf{P}^T)^{-1}\mathbf{x}$

$\rho (f(\mathbf{x}), f(\mathbf{y})) = \sqrt{(\mathbf{x} - \mathbf{y})^T \mathbf{P}^{-1} \mathbf{P} \mathbf{P}^T (\mathbf{P}^T)^{-1} (\mathbf{x} - \mathbf{y})} = \sqrt{(\mathbf{x} - \mathbf{y})^T (\mathbf{x} - \mathbf{y})} = \varepsilon(\mathbf x, \mathbf y)$.

This is probably very similar to what you wanted to do, but I did not quite exactly follow your last conclusion.

share|cite|improve this answer

Your Answer


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.