tabulate_behavior_data and random target chains

random target trials don't always start/end with a fixed event code, rather they have a fixed base event (16 = target on, e.g.) but the data is different on each trial (16, 17, 18, etc.)

need a way for tabulate_behavior_data to capture these trials. we've been using exp_data['events']['event'] for this purpose in other trial alignment functions.