Created
November 23, 2022 18:32
-
-
Save whophil/c809692ceb8d5b4a22b70edc103dd87c to your computer and use it in GitHub Desktop.
Extrinsic/intrinsic convention swapped in `scipy.spatial.transform.Rotation.from_euler()` ?
So I guess, following the convention from wikipedia, and accounting for this discrepancy, when using the non-caps intrinsic-rotation(Tait-bryan angles), the right order is R.from_euler('xyz') and when using extrinsic-rotation(Euler-angles) the right order is R.from_euler('ZYX')
Hi @saadehmd - there is an open issue on scipy here where this notebook is discussed. It's been some years since I looked at it. I recall the scipy contributors pointing out some issues in this notebook, but unfortunately I ran out of time to look into the responses. I have workarounds for my particular use cases.
It would be great if somebody else could pick this up!
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
@whophil This was really useful and eye-opening. I've been scratching my head trying to figure this out myself. Didn't get much help googling this discrepancy either. But anyway, glad that somebody actually unit-tested this problem. Has this ever been reported to / fixed by the authors ?