I am curious if anyone can point me to a reference where the Euler angle coordinates are visualized as a parametrization of $\mathbb{R}P^3$. Bonus points if there are visualization aids for the quaternion and axis-angle parametrization as well.


  • $\begingroup$ Presumably you're asking about projective space because you're aware that $\mathbf{RP}^{3}$ is diffeomorphic to the rotation group $SO(3)$...? $\endgroup$ Commented Jun 26, 2017 at 20:44
  • $\begingroup$ Yes, I am speaking of Real Projective 3-Space. As I understand it, unit quaternions provide a Cartesian parametrization of $S^3$, which we chop in half and identify antipodal points at the equator to get $\mathbb{R}P^3$. $\endgroup$
    – Evan
    Commented Jun 26, 2017 at 20:45
  • $\begingroup$ Not sure what you mean by "at the equator." We regard a vector $(x,y,z)$ in $\mathbb R^3$ as a "pure" quaternion, zero real part, $ v =xi + yj + zk.$ Given a unit quaternion $q,$ with conjugate $\bar{q},$ our rotated vector is $q v \bar{q}.$ Short proof that this is also a pure quaternion. Meanwhile, $q$ and $-q$ give precisely the same rotation. $\endgroup$
    – Will Jagy
    Commented Jun 26, 2017 at 21:31
  • $\begingroup$ I'm talking about the equator of $S^3$, which is a 2-sphere, $S^2$. Unit quaternions (with non zero real part) provide a Cartesian cover of $S^3$. $\endgroup$
    – Evan
    Commented Jun 26, 2017 at 21:32
  • $\begingroup$ My earlier comment was unclear. The questions were: 1. Are you happy with Euler angles and axis-angle descriptions of $SO(3)$? 2. Are you happy with the axis-angle description of the diffeomorphism between $SO(3)$ and $\mathbf{RP}^{3}$? and 3. Are you (therefore) looking for a description in the language of projective space rather than in terms of rotation matrices/unit quaternions? $\endgroup$ Commented Jun 29, 2017 at 0:33


You must log in to answer this question.

Browse other questions tagged .