In order to keep it simple, I would try to model a frame element of an assumed effective width (say 8') and get the forces on that discrete element. In order to get the moment due to eccentricity, you would have to model a "rigid" link supporting the walking column. I have done (2) as KootK suggests and designed the transition reinforcement. A simple check would a T/C check based on the floor-to-floor height, but if that is too conservative, you might consider additional diaphragms.