Log Message: 
Fixed bad bug (but with little consequences): the matrix for calculating eigenvectors (for calculation of rotation axes) was transposed, resulting in a few operators in a few space groups to have wrong axes (for instance in H 3 2 or P 32 2 1). Fortunately we weren't using the axes calculation for anything other than output in test programs (that's why it didn't have consequences). Added testing for comparison of axis vectors versus screw translation components (both must be in same direction).
