Let $ A$ be a $2\times2$ real orthogonal matrix. Then when does $e^A$ become orthogonal as well? According to my calculations, $A$ must be skew-symmetric also and therefore there are only two possibilities. Am I correct?

Edit~~~ Ok, the two matrices are $\pmatrix{0 &1 \\ -1 &0}$ and $\pmatrix{0 & -1 \\ 1 &0}$. Thank you all for the advice below. Now, are these matrices correct? I think they are, but I would like others to check in case I am missing something.


That does indeed sound correct. (And you can probably generalize fairly easily to higher dimensions.)

As a general principle, it's worth knowing that the tangent space to the orthogonal group at $I$ consists exactly of skew-symmetric matrices; since $A \mapsto \exp A$ sends this tangent space to the orthogonal group (and indeed, ignoring speed for the moment, $t \mapsto \exp(tA)$ is a geodesic through the identity, for any skew-symmetric $A$), there's a pretty strong correspondence between the two.

  • $\begingroup$ Could you pleae show me the two matrices? I actually am not familiar with writing matrices on the latex so...I would like to check my answers are indeed correct. $\endgroup$ – Keith Aug 14 at 14:23
  • 2
    $\begingroup$ To write the identity matrix, you type \pmatrix{1 & 0 \\ 0 & 1} and get $\pmatrix{1 & 0 \\ 0 & 1}$; more generally, \pmatrix{a & b \\ c & d} gives $\pmatrix{a & b \\ c & d}$. To answer your question: no, I won't write out the two matrices for you. That's a good exercise for you. $\endgroup$ – John Hughes Aug 14 at 14:26
  • 1
    $\begingroup$ @Keith See here if you want to write a matrix using MathJax. $\endgroup$ – Thomas Shelby Aug 14 at 14:27
  • $\begingroup$ @JohnHughes OK, thanks to your advice, I wrote those matrices; it seems easier than I thought. Now, are they correct? $\endgroup$ – Keith Aug 14 at 14:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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