def policy_evaluation(self, policy, context, desired_action, delta, r,
epsilon):
if policy != 'LinThompSamp':
print("We don't support other bandit algorithms now!")
else:
historystorage = history.MemoryHistoryStorage()
modelstorage = model.MemoryModelStorage()
policy = linthompsamp.LinThompSamp(self.actions, historystorage,
modelstorage, self.d, delta, r,
epsilon)
seq_error = np.zeros(shape=(self.t, 1))
for t in range(self.t):
history_id, action = policy.get_action(context[t])
if desired_action[t][0] != action:
policy.reward(history_id, 0)
# sum_error += 1
if t == 0:
seq_error[t] = 1.0
else:
seq_error[t] = seq_error[t - 1] + 1.0
else:
policy.reward(history_id, 1)
if t > 0:
seq_error[t] = seq_error[t - 1]
return seq_error
def policy_generation(bandit, actions):
historystorage = history.MemoryHistoryStorage()
modelstorage = model.MemoryModelStorage()
if bandit == 'Exp4P':
policy = exp4p.Exp4P(actions,
historystorage,
modelstorage,
delta=0.5,
pmin=None)
elif bandit == 'LinUCB':
policy = linucb.LinUCB(actions, historystorage, modelstorage, 0.3, 20)
elif bandit == 'LinThompSamp':
policy = linthompsamp.LinThompSamp(actions,
historystorage,
modelstorage,
d=20,
delta=0.61,
r=0.01,
epsilon=0.71)
elif bandit == 'UCB1':
policy = ucb1.UCB1(actions, historystorage, modelstorage)
elif bandit == 'Exp3':
policy = exp3.Exp3(actions, historystorage, modelstorage, gamma=0.2)
elif bandit == 'random':
policy = 0
return policy
def test_update_reward(self):
policy = linthompsamp.LinThompSamp(self.actions, self.historystorage,
self.modelstorage, self.d,
self.delta, self.R, self.epsilon)
history_id, action = policy.get_action([[1, 1], [2, 2], [3, 3]])
policy.reward(history_id, 1.0)
self.assertEqual(
policy._historystorage.get_history(history_id).reward, 1)
def test_initialization(self):
policy = linthompsamp.LinThompSamp(self.actions, self.historystorage,
self.modelstorage, self.d,
self.delta, self.R, self.epsilon)
self.assertEqual(self.actions, policy._actions)
self.assertEqual(self.d, policy.d)
self.assertEqual(self.R, policy.R)
self.assertEqual(self.epsilon, policy.epsilon)
def test_model_storage(self):
policy = linthompsamp.LinThompSamp(self.actions, self.historystorage,
self.modelstorage, self.d,
self.delta, self.R, self.epsilon)
history_id, action = policy.get_action([[1, 1], [2, 2], [3, 3]])
policy.reward(history_id, 1.0)
self.assertTrue(
(policy._modelstorage._model['B'].shape == (2, 2)) == True)
self.assertEqual(len(policy._modelstorage._model['muhat']), 2)
self.assertEqual(len(policy._modelstorage._model['f']), 2)
def test_get_first_action(self):
policy = linthompsamp.LinThompSamp(self.actions, self.historystorage,
self.modelstorage, self.d,
self.delta, self.R, self.epsilon)
history_id, action = policy.get_action([[1, 1], [2, 2], [3, 3]])
self.assertEqual(history_id, 0)
self.assertIn(action, self.actions)
self.assertTrue(
(policy._historystorage.get_history(history_id).context == [[1, 1],
[2, 2],
[3,
3]]))
def policy_generation(bandit, actions):
"""
Parameters
----------
bandit： 赌博机算法
actions：动作即推荐的电影
Returns
-------
policy： 生成的策略
"""
historystorage = history.MemoryHistoryStorage() # 内存中历史存储记录
modelstorage = model.MemoryModelStorage() # 内存中模型存储，为了统一
if bandit == 'Exp4P':
policy = exp4p.Exp4P(historystorage,
modelstorage,
actions,
delta=0.5,
p_min=None)
elif bandit == 'LinUCB':
#policy = linucb.LinUCB(historystorage, modelstorage, actions, 0.3, 20)
policy = linucb.LinUCB(history_storage=historystorage,
model_storage=modelstorage,
action_storage=actions,
alpha=0.3,
context_dimension=18)
elif bandit == 'LinThompSamp':
policy = linthompsamp.LinThompSamp(
historystorage,
modelstorage,
actions,
#d=20, Supposed to be context dimension
context_dimension=18,
delta=0.61,
R=0.01,
epsilon=0.71)
elif bandit == 'UCB1':
policy = ucb1.UCB1(historystorage, modelstorage, actions)
elif bandit == 'Exp3':
policy = exp3.Exp3(historystorage, modelstorage, actions, gamma=0.2)
elif bandit == 'random':
policy = 0
return policy
def test_reward_order_descending(self):
policy = linthompsamp.LinThompSamp(self.actions, self.historystorage,
self.modelstorage, self.d,
self.delta, self.R, self.epsilon)
history_id, action = policy.get_action([[1, 1], [2, 2], [3, 3]])
history_id_2, action_2 = policy.get_action([[0, 1], [2, 3], [7, 5]])
policy.reward(history_id_2, 1)
self.assertEqual(
policy._historystorage.get_history(history_id).context,
[[1, 1], [2, 2], [3, 3]])
self.assertEqual(
policy._historystorage.get_history(history_id_2).context,
[[0, 1], [2, 3], [7, 5]])
self.assertEqual(
policy._historystorage.get_history(history_id).reward, None)
self.assertEqual(
policy._historystorage.get_history(history_id_2).reward, 1)
def policy_generation(bandit, actions):
historystorage = history.MemoryHistoryStorage()
modelstorage = model.MemoryModelStorage()
if bandit == 'Exp4P':
policy = exp4p.Exp4P(historystorage,
modelstorage,
actions,
delta=0.5,
p_min=None)
elif bandit == 'LinUCB':
#policy = linucb.LinUCB(historystorage, modelstorage, actions, 0.3, 20)
policy = linucb.LinUCB(history_storage=historystorage,
model_storage=modelstorage,
action_storage=actions,
alpha=0.3,
context_dimension=18)
elif bandit == 'LinThompSamp':
policy = linthompsamp.LinThompSamp(
historystorage,
modelstorage,
actions,
#d=20, Supposed to be context dimension
context_dimension=18,
delta=0.61,
R=0.01,
epsilon=0.71)
elif bandit == 'UCB1':
policy = ucb1.UCB1(historystorage, modelstorage, actions)
elif bandit == 'Exp3':
policy = exp3.Exp3(historystorage, modelstorage, actions, gamma=0.2)
elif bandit == 'random':
policy = 0
return policy