-
Notifications
You must be signed in to change notification settings - Fork 1
/
Random_class.py
69 lines (58 loc) · 1.99 KB
/
Random_class.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
#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
File description: Implementation of the Random Decision model
"""
# Author: Yoan Russac (yoan.russac@ens.fr)
# License: BSD (3-clause)
# Importations
import numpy as np
class PolicyRandom(object):
def __init__(self, d, name, verbose, omniscient=False):
"""
param:
- d: dimension of the action vectors
- name: additional suffix when comparing several policies (optional)
- verbose: To print informations
- omniscient: Does the policy knows when the breakpoints happen ?
ACTION NORMS ARE SUPPOSED TO BE BOUNDED BE 1
"""
# immediate attributes from the constructor
self.name = name
self.verbose = verbose
self.omniscient = omniscient
# attributes for the re-init
self.t = 0
self.dim = d
self.hat_theta = np.zeros(self.dim)
def select_arm(self, arms):
"""
Selecting a random action
param:
- arms : list of objects Arm with contextualized features
Output:
-------
chosen_arm : index of the pulled arm
"""
assert type(arms) == list, 'List of arms as input required'
kt = len(arms) # available actions at time t, k for the moment
chosen_arm = np.random.randint(kt, dtype='int')
return chosen_arm
def update_state(self, features, reward):
"""
Updating the main parameters for the model, nothing to update in this random policy
:param features: Feature used for updating
:param reward: Reward used for updating
:return: Nothing, but the class instances are updated
"""
self.t += 1
def re_init(self):
"""
Re-init function to reinitialize the statistics while keeping the same hyper-parameters
"""
self.t = 0
def __str__(self):
return 'randomPol' + self.name
@staticmethod
def id():
return 'randomPol'