forked from ElteHupkes/tol-revolve
/
world.py
166 lines (121 loc) · 5.05 KB
/
world.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
# External / system
import sys
import math
import asyncio
import time
import csv
import os
from functools import partial
from .config import str_to_address
from .logging import logger
from .gazebo import WorldManager, RequestHandler
from .build import get_builder, get_simulation_robot
from revolve.spec.msgs.neural_net_pb2 import SendNeuralNetwork
from revolve.spec.msgs.evaluation_result_pb2 import EvaluationResult
class LearningManager(WorldManager):
def __init__(self, conf):
super(LearningManager, self).__init__(
world_address=str_to_address(conf.world_address),
analyzer_address=str_to_address(conf.analyzer_address),
output_directory=conf.output_directory,
pose_update_frequency=conf.pose_update_frequency,
restore=conf.restore_directory)
self.conf = conf
# path to the logging directory
self.path_to_log_dir = None
# create path to logging directory if directory names are given:
if conf.output_directory and conf.log_directory:
# path to the logging directory
self.path_to_log_dir = os.path.join(conf.output_directory, conf.log_directory)
# create logging directory:
try:
os.mkdir(self.path_to_log_dir)
except OSError:
logger.debug("Directory " + self.path_to_log_dir + " already exists.")
self.learners = []
# message passing handlers for each robot
self.robot_handlers = {}
@classmethod
async def create(cls, conf):
"""
Coroutine to instantiate a Revolve.Angle WorldManager
:param conf:
:return:
"""
self = cls(conf=conf)
await self._init()
return self
# def get_world_time(self):
# if self.last_time:
# return self.last_time
# else:
# return 0.0
def get_simulation_sdf(self, robot, robot_name):
"""
:param robot:
:type robot: PbRobot
:param robot_name:
:return:
:rtype: SDF
"""
return get_simulation_robot(robot, robot_name, get_builder(self.conf), self.conf)
def get_snapshot_data(self):
data = super(LearningManager, self).get_snapshot_data()
data['learners'] = self.learners
return data
async def restore_snapshot(self, data):
super(LearningManager, self).restore_snapshot(data)
self.learners = data['learners']
for _, learner in self.learners:
if learner.robot.name not in self.robot_handlers:
await self.create_handler_for_robot(learner.robot.name)
def log_info(self, log_name, log_data):
if self.output_directory and self.path_to_log_dir:
for filename, data in log_data.items():
genotype_log_filename = os.path.join(self.path_to_log_dir, log_name, filename)
with open(genotype_log_filename, "a") as genotype_log_file:
genotype_log_file.write(data)
# async def delete_robot(self, robot):
# await super(LearningManager, self).delete_robot(robot)
# # delete .sdf and .pb files when deleting a robot:
# try:
# os.remove(os.path.join(self.output_directory, 'robot_{0}.sdf'.format(robot.robot.id)))
# except OSError:
# pass
# try:
# os.remove(os.path.join(self.output_directory, 'robot_{0}.pb'.format(robot.robot.id)))
# except OSError:
# pass
async def create_handler_for_robot(self, robot_name):
handler = await RequestHandler.create(
self.connection,
request_class = SendNeuralNetwork,
request_type = "gazebo.msgs.SendNeuralNetwork",
response_class = EvaluationResult,
response_type = "gazebo.msgs.EvaluationResult",
advertise = "/gazebo/default/{0}/modify_neural_network".format(robot_name),
subscribe = "/gazebo/default/{0}/fitness".format(robot_name),
id_attr = "id")
self.robot_handlers[robot_name] = handler
async def add_learner(self, learner, log_name=None, init_brain_list=None):
# create directory for this learner's logs:
if log_name is not None and self.path_to_log_dir is not None:
try:
os.mkdir(os.path.join(self.path_to_log_dir, log_name))
except OSError:
pass
await self.create_handler_for_robot(learner.robot.name)
self.learners.append((log_name, learner))
def run_brain(self, robot_name, brain_msg):
msg = SendNeuralNetwork()
msg.id = self.get_unique_id()
msg.neuralNetwork.CopyFrom(brain_msg)
handler = self.robot_handlers[robot_name]
return handler.do_request(msg)
async def run(self):
futures = []
for log_name, learner in self.learners:
log_callback = None if log_name is None else partial(self.log_info, log_name)
futures.append(learner.run(self, log_callback))
if futures:
await asyncio.wait(futures)