On the possibility of fitting a large set of rare earth nuclei with constant IBM-2 parameters