Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

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 am trying to show that if $A$ and $B$ are orthonormal bases of a real finite dimensional vector space $V$ and $P$ is the change of basis matrix between $A$ and $B$ then $\det(P) = \pm 1$. Here is my progress:

When working on this problem, I discovered the following fact: If $G_B$ denotes the Gram matrix of the inner product relative to $B$ and $G_A$ denotes the Gram matrix of the inner product relative to $A$ then the Gram matrices and the transition matrix $P$ are related through $$ G_B = P^T G_A P. $$

The original claim is then a corollary of this fact: $$ G_B = P^T G_A P \implies \det(G_B) = \det(P^T) \det(G_A) \det(P) \implies \det(P) = \pm 1 $$ where I have used basic properties of the determinant and the fact that the determinant of the Gram matrix relative to an orthonormal basis is $1$.

So, I think this proves the original claim but I am now stuck with proving the more general claim that $G_B = P^T G_A P$. What would be a good way to approach this?

share|cite|improve this question
Since A and B are both orthonormal bases, which means P preserves inner product, i.e. P is orthogonal transformation, so det(P)=1 or -1. – chaohuang Jul 1 '12 at 21:35
also, since P is orthogonal,P^T = P^(-1) – chaohuang Jul 1 '12 at 21:36
up vote 1 down vote accepted

$$\begin{eqnarray*} (P^T G_A P)_{il} &=& (P^T)_{ij} (G_A)_{jk} P_{kl} \\ &=& P_{ji} \langle x_k,x_j\rangle P_{kl} \\ &=& \langle P_{kl} x_k, P_{ji} x_j\rangle \\ &=& \langle (P^T x)_l, (P^T x)_i\rangle \\ &=& (G_B)_{il} \end{eqnarray*}$$

$\def\det{\mathrm{det}\,}$ Addendum: On second reading I may have misinterpreted the meaning of "this" in the last sentence of your question.

Let $X_{ij} = (x_i)_j$ be the matrix whose rows are the orthonormal basis vectors. $X$ transforms like $X\to P^T X$. We have $X X^T = \mathbb{I}$, since the basis is orthonormal. Then $$(P^T X)(P^T X)^T = P^T X X^T P = P^T P = \mathbb{I},$$ since the new basis is orthonormal. Therefore, $\det(P^T P) = (\det P)^2 = 1$, so $\det P = \pm 1$.

share|cite|improve this answer

Here is a possibly simpler approach:

Let $a_1,...,a_n$, $b_1,...,b_n$ be the elements of bases $A,B$ respectively. By assumption, $b_i = P a_i$. By 'orthonormality', we have $\langle b_i, b_j \rangle = \langle P a_i, P a_j \rangle = \langle a_i, P^TP a_j \rangle = \delta_{ij}$. It follows from this that $P^TP = I$, and that $(\det P)^2 = 1$, from which the result follows (since $\det P$ is real).

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.