forked from CavenaghiEmanuele/Multi-armed-bandit
-
Notifications
You must be signed in to change notification settings - Fork 0
/
bacteria_tests.py
180 lines (152 loc) · 7.04 KB
/
bacteria_tests.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
import pandas as pd
import numpy as np
from numpy import random
from multiprocessing import Pool, cpu_count
from typing import Dict, List
from tqdm import trange
import multi_armed_bandit as mab
class BacteriaDatasetSession():
_test_env: pd.DataFrame
_train_env: pd.DataFrame
_n_test: int
_n_step_train: int
_n_step_test: int
_fixed_step: int
def __init__(self, n_test:int=10, fixed_step:int=1) -> None:
self._n_test = n_test
self._n_step_train = 61 * fixed_step
self._n_step_test = 177 * fixed_step
self._fixed_step = fixed_step
self.build_env(train_test='train')
self.build_env(train_test='test')
def build_env(self, train_test:str):
if train_test == 'train':
df = pd.read_csv('datasets/bacteria/train.csv')
elif train_test == 'test':
df = pd.read_csv('datasets/bacteria/test.csv')
bacteries = list(df.bacteria.unique())
# Build frequencies
frequencies = {}
for registration in df.columns:
if registration != 'bacteria':
frequencies.update(
{registration : [
(bacteries.index(row['bacteria']), row[registration] / df[registration].sum())
for _, row in df.iterrows()
]
})
# Build replay
replay = {}
for i, value in enumerate(frequencies.values()):
if i == 0: replay.update({'probabilities' : [x[1] for x in value]})
else: replay.update({i*self._fixed_step : value})
if train_test == 'train':
self._train_env = mab.BernoulliReplayBandit(replay=replay)
elif train_test == 'test':
self._test_env = mab.BernoulliReplayBandit(replay=replay)
def plot_arms(self, train_test:str, plot_legend:bool = True):
if train_test == 'train':
session = mab.Session(env=self._train_env, agent=[])
session.run(n_step=self._n_step_train, use_replay=True)
self._train_env.plot_arms(render=True, plot_legend=plot_legend)
elif train_test == 'test':
session = mab.Session(env=self._test_env, agent=[])
session.run(n_step=self._n_step_test, use_replay=True)
self._test_env.plot_arms(render=True, plot_legend=plot_legend)
def run(self) -> None:
params = [(_,) for _ in range(self._n_test)]
pool = Pool(cpu_count())
results = pool.starmap(self._run, params)
pool.close()
pool.join()
tmp = {str(agent):[] for agent in results[0]}
for i in range(len(results)):
for key, value in results[i].items():
tmp[str(key)].append(value)
dataset = pd.DataFrame.from_dict(tmp)
dataset.to_csv('results/bacteria/reward_perc.csv', index=False)
return
def _run(self, fake) -> Dict:
agent_list = self._best_agents(n_arms=self._test_env._n_arms) #Build agents
np.random.seed()
session = mab.Session(env=self._test_env, agent=agent_list)
session.run(n_step=self._n_step_test, n_test=1, use_replay=True)
results = {str(agent): session.get_reward_sum(agent) for agent in agent_list}
results.update({"Oracle" : session.get_reward_sum("Oracle")})
return results
def find_params(self) -> None:
path = 'results/bacteria/find_params/'
self._params = {
'f_algo' : [
(0.9, 100), (0.9, 200), (0.9, 400), (0.9, 800),
(0.95, 100), (0.95, 200), (0.95, 400), (0.95, 800),
(0.99, 100), (0.99, 200), (0.99, 400), (0.99, 800),
(0.999, 800), (0.9999, 800), (0.99999, 800)
],
'Sliding Window TS' : [
25, 50, 100, 200,
400, 800, 1600, 3200,
6400, 12800, 25600, 51200,
102400, 204800, 409600
],
'Discounted TS' : [
0.5, 0.6, 0.7, 0.8,
0.9, 0.92, 0.95, 0.97,
0.98, 0.99, 0.999, 0.9999,
0.99999, 0.999999, 0.9999999
]
}
# only to save results
self._params.update({'Max d-sw TS':self._params['f_algo']})
self._params.update({'Min d-sw TS':self._params['f_algo']})
self._params.update({'Mean d-sw TS':self._params['f_algo']})
for i in range(len(self._params['f_algo'])):
params = [
(
self._params['f_algo'][i][0],
self._params['f_algo'][i][1],
self._params['Sliding Window TS'][i],
self._params['Discounted TS'][i]
)
for _ in range(self._n_test)
]
pool = Pool(cpu_count())
results = pool.starmap(self._find_params, params)
pool.close()
pool.join()
for agent in results[0]:
tmp = {str(self._params[str(agent)][i]) : [result[agent] for result in results]}
dataset = pd.concat(
[pd.read_csv(path + agent + '.csv'), pd.DataFrame.from_dict(tmp)],
axis=1, join='inner')
dataset.to_csv(path + agent + '.csv', index=False)
return
def _find_params(self, f_gamma, f_n, sw_n, d_ts_gamma):
n_arms = self._train_env._n_arms
########## BUILD AGENTS ###########
agent_list = [
mab.MaxDSWTS(n_arms=n_arms, gamma=f_gamma, n=f_n, store_estimates=False),
mab.MinDSWTS(n_arms=n_arms, gamma=f_gamma, n=f_n, store_estimates=False),
mab.MeanDSWTS(n_arms=n_arms, gamma=f_gamma, n=f_n, store_estimates=False),
mab.BernoulliSlidingWindowTS(n_arms=n_arms, n=sw_n, store_estimates=False),
mab.DiscountedBernoulliTS(n_arms=n_arms, gamma=d_ts_gamma, store_estimates=False)
]
np.random.seed()
session = mab.Session(env=self._train_env, agent=agent_list)
session.run(n_step=self._n_step_train, n_test=1, use_replay=True)
return {str(agent): session.get_reward_sum(agent) for agent in agent_list}
def _best_agents(self, n_arms) -> List:
return [
mab.MaxDSWTS(n_arms=n_arms, gamma=0.9999, n=800, store_estimates=False),
mab.MinDSWTS(n_arms=n_arms, gamma=0.99, n=800, store_estimates=False),
mab.MeanDSWTS(n_arms=n_arms, gamma=0.9999, n=800, store_estimates=False),
mab.BernoulliThompsonSampling(n_arms=n_arms, store_estimates=False),
mab.BernoulliSlidingWindowTS(n_arms=n_arms, n=12800, store_estimates=False),
mab.DiscountedBernoulliTS(n_arms=n_arms, gamma=0.9999, store_estimates=False),
mab.RandomAlgo(n_arms=n_arms)
]
if __name__ == '__main__':
session = BacteriaDatasetSession(n_test=10, fixed_step=1)
#session.run()
#session.find_params()
session.plot_arms(plot_legend=False, train_test='train')