It would be convenient to be able to construct a CartesianPose object from a 4x4 transformation object. Similarly, it would be convenient to have a direct method get_transformation_matrix() or similar to return the pose as a 4x4 transformation matrix.

The 3x3 rotation matrix could be handled similarly, though for construction it would take either a 3vector for position or set position to 0.

This feature would make operations with other standard representations even easier.