rv = np.random.randint(0, nstrats)
node_strat = {node: robot_strategies[rv]}
edge_strats = {e: robot_strategies[rv] for e in edges}
self.internal_policy = {'nodes': nodes,
'edges': edges,
'quantity': initial_balls,
'node_strats': node_strat,
'edge_strats': edge_strats,
'delta': np.zeros(node + 1)
}
return self
robot_udo = UDO(udo=robot(G, balls), masked_members=['obj'])
initial_conditions = {'balls': balls, 'network': G, 'robot': robot_udo}
def update_balls(params, step, sL, s, _input):
delta_balls = _input['robot'].internal_policy['delta']
new_balls = s['balls']
for e in G.edges:
move_ball = delta_balls[e]
src = e[0]
dst = e[1]
if (new_balls[src] >= move_ball) and (new_balls[dst] >= -move_ball):
new_balls[src] = new_balls[src] - move_ball
new_balls[dst] = new_balls[dst] + move_ball
key = 'balls'
def read(self, ds_uri):
self.ds = SilentDF(pd.read_csv(ds_uri))
return self
def write(self, ds_uri):
pd.to_csv(ds_uri)
# Generic update function
pass
# can be accessed after an update within the same substep and timestep
state_udo = UDO(udo=udoExample(0, DF), masked_members=['obj', 'perception'])
policy_udoA = UDO(udo=udoExample(0, DF), masked_members=['obj', 'perception'])
policy_udoB = UDO(udo=udoExample(0, DF), masked_members=['obj', 'perception'])
def udo_policyA(_g, step, sL, s, **kwargs):
s['udo_policies']['udo_A'].updateX()
return {'udo_A': udoPipe(s['udo_policies']['udo_A'])}
def udo_policyB(_g, step, sL, s, **kwargs):
s['udo_policies']['udo_B'].updateX()
return {'udo_B': udoPipe(s['udo_policies']['udo_B'])}
policies = {"p1": udo_policyA, "p2": udo_policyB}
T = 50 #iterations in our simulation
n = 3 #number of boxes in our network
m = 2 #for barabasi graph type number of edges is (n-2)*m
G = nx.barabasi_albert_graph(n, m)
k = len(G.edges)
# class udoExample(object):
# def __init__(self, G):
# self.G = G
# self.mem_id = str(hex(id(self)))
g = UDO(udo=G)
print()
# print(g.edges)
# print(G.edges)
# pp.pprint(f"{type(g)}: {g}")
# next
balls = np.zeros(n,)
for node in g.nodes:
rv = np.random.randint(1,25)
g.nodes[node]['initial_balls'] = rv
balls[node] = rv
# pp.pprint(balls)
# next