Rotation Check

[remus]

An interesting feature of a transformation matrix is that its determinant will always be 1 if it is a rotation or combination of rotations so i wrote this little snippet to test whether a given transformation matrix consists solely of a combination of rotations.

Probably not much practical use but someone might find use for it somewhere.

A little word of warning, it'll also return true if you just input the identity matrix (which i suppose is technically a rotation of 0 degrees.)

def rotation_check(trans)

arr = trans.to_a
a1 = arr[0]
a2 = arr[1]
a3 = arr[2]
a4 = arr[4]
a5 = arr[5]
a6 = arr[6]
a7 = arr[8]
a8 = arr[9]
a9 = arr[10]

det_arr = a1*(a5*a9-a6*a8)-a2*(a4*a9-a7*a6)+a3*(a4*a8-a7*a5)

if det_arr == 1 and arr[14] ==0 and arr[13] == 0 and arr[12] == 0
  true
else
  false
end#if

end#def

[Chris Fullmer]

if the transformation is moved off 0,0,0, it will not return 1 yeah? And if the comp has been scaled or skewed it does not return 1, right? Those are the things it checks for more or less, right?

Chris

[remus]

No, it will not return true in any of those cases. Having said that, if i was less lazy it would be fairly trivial to get it to check for translations and 'skews, scales and rotations' (although not each separately.)