I was going over my notes on classical mechanics and just started to review rotation matrices which is the first topic the book starts with. On page 3

enter image description here

The rotation matrix associated with 1.2a and 1.2b is

\begin{pmatrix} \cos\theta & \sin\theta \\ -\sin\theta & \cos\theta \\ \end{pmatrix}

but when I try to derive the matrix by following the unit vectors $\hat i$ and $\hat j$

I get

\begin{pmatrix} \cos\theta & -\sin\theta \\ \sin\theta & \cos\theta \\ \end{pmatrix}

The one that the book derives would be clockwise rotation, and the one I got would be for counter-clockwise rotation correct?

  • $\begingroup$ There are two points of view : The point $\: \mathrm P\:$ is rotated clockwise and the coordinate system remains fixed or the coordinate system is rotated anticlockwise and the point $\: \mathrm P\:$ remains fixed. $\endgroup$
    – Frobenius
    Commented Apr 21, 2018 at 6:36
  • $\begingroup$ Yea I was just realizing that. I think what the book does is rotate the coordinate system counter-clockwise but then it measures lengths in the new coordinate system. Meaning that they effectively changed basis while the matrix I was computing was in the old basis. Is it correct to think of it in this way? $\endgroup$
    – Elvis
    Commented Apr 21, 2018 at 6:39
  • $\begingroup$ Yes, precisely. $\endgroup$
    – Frobenius
    Commented Apr 21, 2018 at 6:41
  • $\begingroup$ In two dimensions you could solve your problems with complex numbers. Suppose you want to rotate the point $\:\mathrm P : z=x+iy\:$ by an angle $\:\theta\:$ to a new position $\: \mathrm P' : z'=x'+iy'\:$. Don't bother yourself if clock- or anticlock- wise, the sign of $\:\theta\:$ cares for this. \begin{align} z'=e^{i\theta}z\: \Longrightarrow \:x'+iy' & =\left(\cos\theta+i\sin\theta\right)\left(x+iy\right) \tag{01}\\ &=\left(x\cos\theta-y\sin\theta \right)+i\left(x\sin\theta+y\cos\theta \right) \nonumber \end{align} $\endgroup$
    – Frobenius
    Commented Apr 21, 2018 at 7:15
  • $\begingroup$ ...so \begin{equation} \begin{bmatrix} x'\\ y' \end{bmatrix} = \begin{bmatrix} \cos\theta & -\sin\theta\\ \sin\theta & \hphantom{-}\cos\theta \end{bmatrix} \begin{bmatrix} x\\ y \end{bmatrix} \tag{02} \end{equation} $\endgroup$
    – Frobenius
    Commented Apr 21, 2018 at 7:16

1 Answer 1


You can construct the rotation matrix by finding the direction cosines defined as $\lambda_{i,j}=\cos(x'_{i},x_{j})$ so:





The book is right.

  • 1
    $\begingroup$ Sorry but your post doesn't explain anything. My concern is that the book presents a rotation matrix that seems to represent clockwise rotation while what they illustrate is counter-clockwise rotation; is this incorrect? $\endgroup$
    – Elvis
    Commented Apr 21, 2018 at 4:57
  • $\begingroup$ Oh, the problem the book represents a rotation from $(x'_{1},x'_{2})$ system to $(x_{1},x_{2})$ system. $\endgroup$ Commented Apr 21, 2018 at 5:02
  • $\begingroup$ Note that your rotation matrix, which is the inverse of book's matrix, represents the rotation from $(x_{1},x_{2})$ to $(x'_{1},x'_{2})$, maybe that is your confusion. $\endgroup$ Commented Apr 21, 2018 at 5:11
  • $\begingroup$ Correct, I'm going from $(x_{1}, x_{2})$ to $(x'_{1}, x'_{2})$ so shouldn't that transformation be $\begin{pmatrix} x'_{1} \\ x'_{2} \\ \end{pmatrix} = \begin{pmatrix} \cos\theta & -\sin\theta \\ \sin\theta & \cos\theta \\ \end{pmatrix} \begin{pmatrix} x_{1} \\ x_{2} \\ \end{pmatrix} $? $\endgroup$
    – Elvis
    Commented Apr 21, 2018 at 5:44
  • $\begingroup$ To do what the book is doing wouldn't it be $\begin{pmatrix} x_{1} \\ x_{2} \\ \end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta \\ -\sin\theta & \cos\theta \\ \end{pmatrix} \begin{pmatrix} x'_{1} \\ x'_{2} \\ \end{pmatrix} $? $\endgroup$
    – Elvis
    Commented Apr 21, 2018 at 5:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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