S.unit does not equal to S / S.theta?

[x35yao]

Hi guys,

I am reading the book "Robotics, Vision, and Control". On page 76, it says S.unit() is the same as S / S.theta. I tried it on my end and got different results.

Any information will be appreciated.
Best,
Luke

[petercorke]

Hi @x35yao, thank you for your interest in RVC3-python, and apologies for the very long delay in replying. Various life events got in the way, and since May this year I've been doing a lot of work across RTB, Swift and SpatialGeometry.

You are right, and thank you for reporting it: this is a bug in the Spatial Math Toolbox (SMTB), not the book. For a twist S, S.unit() should equal S / S.theta for a general twist, as on p.76. In released versions of SMTB the rotational and prismatic cases in Twist3.unit() and Twist2.unit() are swapped, so for an ordinary screw twist it silently drops the rotation, and for a pure rotation or pure translation it raises an error.

The fix is in SMTB pull request #225, which also adds tests. SMTB is now maintained by RAI and I do not control PR merging and releases. Until the fix is released, S / S.theta gives the result described in the book. The exception is a pure translation, where S.theta is zero so the division fails, and that case needs the fix.

Thanks again for finding this. If you find anything else, please post an Issue rather than a discussion thread.