Given 2D view bounds and content bounds, gives you a 3x3 transformation matrix that will scale and translate the content such that it fits to the view.
