# Simulate for i iterations
i = initial_i = 10000
print(f"Iterations remaining {i}")
iterations_list = []
agent_iterations = 0
while True:
agent_a.iterate_rl()
agent_iterations += 1
# Explore until goal is found, then try again
if agent_a.at_goal():
i -= 1
# Train until done
if i == 0:
break
# Reset agent in when retraining
iterations_list.append(agent_iterations)
agent_iterations = 0
agent_a.reset()
agent_a.set_state(init_agent_a)
print(f"Iterations remaining {i}")
print(f"Iterations over time: {iterations_list}")
# Migration step, for a star network
for agent in agents:
# Perform migration(s)
for migration_target in migration_map[agent]:
migration = agent.emigrate(0.01)
migration_target.immigrate(migration)
# Iterate the agents, which now includes the migrated population.
for agent in agents:
agent.iterate_rl()
agent_iterations += 1
# Explore until C finds goal, then try again
if agent_c.at_goal():
i -= 1
# Train until done
if i == 0:
break
# Reset agent in when retraining
iterations_list.append(agent_iterations)
agent_iterations = 0
for agent in agents:
agent.reset()
agent.set_state(init_state_map[agent])
print(f"Iterations remaining {i}")
