Algorithm 2 Group detection
Input: A dataset N with skeleton joints in the form $(x,y$) per timestamp
an integer stops_expected with the number of assigned stops
an integer groups_expected representing the number of assigned group locations
Output: A dataset class_group with group membership, and stop locations
	body_identification=spatemp_stop_kmeans(N, stops_expected)
	for$b_i$ in body_indentification:
	trajectory_stops = spatemp_stop$(b_i, t,r)$
	shared_stops=intersection_stops(trajectory_stops)
	class_group=spatemp_stop_kmeans_time(shared_stops, groups_expected)
	return class_group