2009-10-03 02:24:04 UTC
Hello all, not sure if this is the right place to post about this, but here goes.
Im working through the details of deriving the rotation vector/operator that is then put in matrix form on this website:- https://www.euclideanspace.com/maths/algebra/realNormedAlgebra/quaternions/transforms/derivations/vecExp/index.htm
Once I get to this part:- (-sq(vp•vq) + vq•(-vp x vq + sqvp) , - vq x (-vp x vq + sqvp) + sq(-vp x vq + sqvp) - vq(vp•vq))
I am confused on how to simplify the real part of the quaternion down to - vq•(vp x vq)
I suppose Im confused on how to simplify any part of it :P but one bit at a time.
thanks
from: martinbaker
2009-10-03 08:22:42 UTC
we want to simplify:
-sq*(vp•vq) + vq•(-vp x vq + sq * vp)
expanding out the second part gives:
-sq*(vp•vq) - vq•(vp x vq) + vq•(sq * vp)
the scalar part can be moved outside the brackets: vq•(sq * vp)=sq*(vq•vp)
unlike the cross product the dot product commutes so: vq•vp=vp•vq so we get:
-sq*(vp•vq) - vq•(vp x vq) + sq *(vp•vq)
canceling out the first and last terms leaves the middle term: - vq•(vp x vq)
On the website I have explained things in terms of scalar and vector (s,v) as well as the more usual notation: a + i b + j c + k d this is because different people seem to latch onto different ways of understanding it. So if you find this approach difficult, I would try the other notations, I don't think it is necessary to understand all approaches to use quaternions or understand quaternion, matrix conversions.
Martin
