Signal8 0.1__tar.gz → 0.4__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Potentially problematic release.
This version of Signal8 might be problematic. Click here for more details.
- {Signal8-0.1 → Signal8-0.4}/PKG-INFO +7 -7
- {Signal8-0.1 → Signal8-0.4}/README.md +6 -6
- Signal8-0.4/Signal8/Signal8.py +153 -0
- Signal8-0.4/Signal8/__init__.py +0 -0
- Signal8-0.4/Signal8/main.py +4 -0
- {Signal8-0.1 → Signal8-0.4}/Signal8.egg-info/PKG-INFO +7 -7
- Signal8-0.4/Signal8.egg-info/SOURCES.txt +14 -0
- Signal8-0.4/Signal8.egg-info/top_level.txt +1 -0
- {Signal8-0.1 → Signal8-0.4}/setup.py +1 -1
- Signal8-0.1/Signal8.egg-info/SOURCES.txt +0 -7
- Signal8-0.1/Signal8.egg-info/top_level.txt +0 -1
- {Signal8-0.1 → Signal8-0.4}/LICENSE +0 -0
- {Signal8-0.1 → Signal8-0.4}/Signal8.egg-info/dependency_links.txt +0 -0
- {Signal8-0.1 → Signal8-0.4}/setup.cfg +0 -0
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.1
|
|
2
2
|
Name: Signal8
|
|
3
|
-
Version: 0.
|
|
3
|
+
Version: 0.4
|
|
4
4
|
Summary: A multi-agent environment inspired by the Lewis Signaling Game, featuring eight unique problem configurations with both static and dynamic adversaries.
|
|
5
5
|
Home-page: https://github.com/ethanmclark1/signal8
|
|
6
6
|
Author: Ethan Clark
|
|
@@ -16,13 +16,13 @@ Sig8 is an adapted version of the Simple environment, originally developed by th
|
|
|
16
16
|
|
|
17
17
|
# Signal8
|
|
18
18
|
|
|
19
|
-
This repository contains a simple multi-agent environment with continuous observations and a discrete action space, inspired by the Lewis Signaling game. The environment incorporates basic simulated physics to create a scenario where
|
|
19
|
+
This repository contains a simple multi-agent environment with continuous observations and a discrete action space, inspired by the Lewis Signaling game. The environment incorporates basic simulated physics to create a scenario where agents must communicate and collaborate effectively to navigate around both static and dynamic adversaries to reach a goal.
|
|
20
20
|
|
|
21
21
|
## Installation
|
|
22
22
|
|
|
23
23
|
```
|
|
24
|
-
git clone https://github.com/ethanmclark1/
|
|
25
|
-
cd
|
|
24
|
+
git clone https://github.com/ethanmclark1/signal8.git
|
|
25
|
+
cd signal8
|
|
26
26
|
pip install -r requirements.txt
|
|
27
27
|
pip install -e .
|
|
28
28
|
```
|
|
@@ -30,10 +30,10 @@ pip install -e .
|
|
|
30
30
|
## Usage
|
|
31
31
|
|
|
32
32
|
```
|
|
33
|
-
import
|
|
33
|
+
import signal8
|
|
34
34
|
|
|
35
|
-
env =
|
|
36
|
-
env.reset(options={
|
|
35
|
+
env = signal8.env(has_dynamic_adversaries=True)
|
|
36
|
+
env.reset(options={"problem_name": "v_cluster"}))
|
|
37
37
|
```
|
|
38
38
|
|
|
39
39
|
## List of Problem Scenarios
|
|
@@ -4,13 +4,13 @@ Sig8 is an adapted version of the Simple environment, originally developed by th
|
|
|
4
4
|
|
|
5
5
|
# Signal8
|
|
6
6
|
|
|
7
|
-
This repository contains a simple multi-agent environment with continuous observations and a discrete action space, inspired by the Lewis Signaling game. The environment incorporates basic simulated physics to create a scenario where
|
|
7
|
+
This repository contains a simple multi-agent environment with continuous observations and a discrete action space, inspired by the Lewis Signaling game. The environment incorporates basic simulated physics to create a scenario where agents must communicate and collaborate effectively to navigate around both static and dynamic adversaries to reach a goal.
|
|
8
8
|
|
|
9
9
|
## Installation
|
|
10
10
|
|
|
11
11
|
```
|
|
12
|
-
git clone https://github.com/ethanmclark1/
|
|
13
|
-
cd
|
|
12
|
+
git clone https://github.com/ethanmclark1/signal8.git
|
|
13
|
+
cd signal8
|
|
14
14
|
pip install -r requirements.txt
|
|
15
15
|
pip install -e .
|
|
16
16
|
```
|
|
@@ -18,10 +18,10 @@ pip install -e .
|
|
|
18
18
|
## Usage
|
|
19
19
|
|
|
20
20
|
```
|
|
21
|
-
import
|
|
21
|
+
import signal8
|
|
22
22
|
|
|
23
|
-
env =
|
|
24
|
-
env.reset(options={
|
|
23
|
+
env = signal8.env(has_dynamic_adversaries=True)
|
|
24
|
+
env.reset(options={"problem_name": "v_cluster"}))
|
|
25
25
|
```
|
|
26
26
|
|
|
27
27
|
## List of Problem Scenarios
|
|
@@ -0,0 +1,153 @@
|
|
|
1
|
+
import math
|
|
2
|
+
import numpy as np
|
|
3
|
+
|
|
4
|
+
from gymnasium.utils import EzPickle
|
|
5
|
+
from utils.scenario import BaseScenario
|
|
6
|
+
from utils.core import Agent, Landmark, World, Action
|
|
7
|
+
from utils.simple_env import SimpleEnv, make_env
|
|
8
|
+
from utils.problems import get_problem, get_problem_list
|
|
9
|
+
|
|
10
|
+
class raw_env(SimpleEnv, EzPickle):
|
|
11
|
+
def __init__(
|
|
12
|
+
self,
|
|
13
|
+
agent_radius=0.1,
|
|
14
|
+
goal_radius=0.1,
|
|
15
|
+
adversary_radius=0.1,
|
|
16
|
+
num_adversaries=4,
|
|
17
|
+
has_dynamic_adversaries=False,
|
|
18
|
+
max_cycles=500,
|
|
19
|
+
continuous_actions=False,
|
|
20
|
+
render_mode=None
|
|
21
|
+
):
|
|
22
|
+
|
|
23
|
+
scenario = Scenario()
|
|
24
|
+
world = scenario.make_world(
|
|
25
|
+
agent_radius,
|
|
26
|
+
goal_radius,
|
|
27
|
+
num_adversaries,
|
|
28
|
+
adversary_radius,
|
|
29
|
+
has_dynamic_adversaries
|
|
30
|
+
)
|
|
31
|
+
|
|
32
|
+
super().__init__(
|
|
33
|
+
scenario=scenario,
|
|
34
|
+
world=world,
|
|
35
|
+
render_mode=render_mode,
|
|
36
|
+
max_cycles=max_cycles,
|
|
37
|
+
continuous_actions=continuous_actions,
|
|
38
|
+
)
|
|
39
|
+
|
|
40
|
+
self.metadata["agent_radius"] = agent_radius
|
|
41
|
+
self.metadata["goal_radius"] = goal_radius
|
|
42
|
+
self.metadata["num_adversaries"] = num_adversaries
|
|
43
|
+
self.metadata["adversary_radius"] = adversary_radius
|
|
44
|
+
|
|
45
|
+
env = make_env(raw_env)
|
|
46
|
+
|
|
47
|
+
class Scenario(BaseScenario):
|
|
48
|
+
def make_world(self, agent_radius, goal_radius, num_adversaries, adversary_radius, has_dynamic_adversaries):
|
|
49
|
+
world = World(has_dynamic_adversaries)
|
|
50
|
+
world.problem_scenarios = get_problem_list()
|
|
51
|
+
|
|
52
|
+
num_dynamic_adversaries = math.floor(num_adversaries / 2) if has_dynamic_adversaries else 0
|
|
53
|
+
num_agents = num_dynamic_adversaries + 1
|
|
54
|
+
|
|
55
|
+
world.agents = [Agent() for _ in range(num_agents)]
|
|
56
|
+
for i, agent in enumerate(world.agents):
|
|
57
|
+
agent.adversary = True if i > 0 else False
|
|
58
|
+
base_name = "adversary" if agent.adversary else "agent"
|
|
59
|
+
base_index = i - 1 if agent.adversary else 0
|
|
60
|
+
agent.name = f"{base_name}_{base_index}"
|
|
61
|
+
agent.collide = False
|
|
62
|
+
agent.silent = True
|
|
63
|
+
agent.blind = agent.adversary
|
|
64
|
+
agent.size = agent_radius
|
|
65
|
+
|
|
66
|
+
world.landmarks = [Landmark() for _ in range(num_adversaries - num_dynamic_adversaries + 1)]
|
|
67
|
+
world.landmarks[0].name = "goal_0"
|
|
68
|
+
world.landmarks[0].collide = False
|
|
69
|
+
world.landmarks[0].movable = False
|
|
70
|
+
world.landmarks[0].size = goal_radius
|
|
71
|
+
|
|
72
|
+
for i, landmark in enumerate(world.landmarks[1:]):
|
|
73
|
+
landmark.name = f"obstacle_{i}"
|
|
74
|
+
landmark.collide = False
|
|
75
|
+
landmark.size = adversary_radius
|
|
76
|
+
|
|
77
|
+
return world
|
|
78
|
+
|
|
79
|
+
def reset_world(self, world, np_random, problem_name="v_cluster"):
|
|
80
|
+
self.get_problem_scenario(world, problem_name)
|
|
81
|
+
world.problem_name = problem_name
|
|
82
|
+
|
|
83
|
+
# set state and color of both agents and landmarks
|
|
84
|
+
for i, agent in enumerate(world.agents):
|
|
85
|
+
if agent.adversary:
|
|
86
|
+
agent.color = np.array([0.5, 0, 0])
|
|
87
|
+
agent.state.p_vel = np.zeros(world.dim_p)
|
|
88
|
+
agent.state.p_pos = np_random.uniform(*zip(*world.dynamic_adversarial_constr[(i+1) % len(world.dynamic_adversarial_constr)]))
|
|
89
|
+
agent.action_callback = self.get_scripted_action
|
|
90
|
+
else:
|
|
91
|
+
agent.color = np.array([0, 0.8, 0])
|
|
92
|
+
agent.state.p_vel = np.zeros(world.dim_p)
|
|
93
|
+
agent.state.p_pos = np_random.uniform(*zip(*world.start_constr))
|
|
94
|
+
agent.goal = world.landmarks[0]
|
|
95
|
+
agent.goal.color = np.array([0, 0, 0.8])
|
|
96
|
+
agent.goal.state.p_vel = np.zeros(world.dim_p)
|
|
97
|
+
agent.goal.state.p_pos = np_random.uniform(*zip(*world.goal_constr))
|
|
98
|
+
|
|
99
|
+
for landmark in world.landmarks[1:]:
|
|
100
|
+
landmark.color = np.array([0.2, 0.2, 0.2])
|
|
101
|
+
landmark.state.p_vel = np.zeros(world.dim_p)
|
|
102
|
+
landmark.state.p_pos = np_random.uniform(*zip(*world.static_adversarial_constr))
|
|
103
|
+
|
|
104
|
+
# Do not need to implement this function
|
|
105
|
+
def reward(self, agent, world):
|
|
106
|
+
return 0
|
|
107
|
+
|
|
108
|
+
def observation(self, agent, world):
|
|
109
|
+
return np.concatenate((agent.state.p_pos, agent.state.p_vel))
|
|
110
|
+
|
|
111
|
+
# Get constraints on entities given the problem name
|
|
112
|
+
def get_problem_scenario(self, world, problem_name):
|
|
113
|
+
problem = get_problem(problem_name, world.has_dynamic_adversaries)
|
|
114
|
+
world.start_constr = problem['start']
|
|
115
|
+
world.goal_constr = problem['goal']
|
|
116
|
+
world.static_adversarial_constr = problem['static_adversary']
|
|
117
|
+
|
|
118
|
+
if world.has_dynamic_adversaries:
|
|
119
|
+
world.dynamic_adversarial_constr = problem['dynamic_adversary']
|
|
120
|
+
|
|
121
|
+
# Get scripted action for adversarial agents s.t. they move in a straight line along their constraint
|
|
122
|
+
def get_scripted_action(self, agent, world):
|
|
123
|
+
action = Action()
|
|
124
|
+
action.u = np.zeros(world.dim_p)
|
|
125
|
+
action.c = np.zeros(world.dim_c)
|
|
126
|
+
|
|
127
|
+
constr = world.dynamic_adversarial_constr[int(agent.name[-1])]
|
|
128
|
+
constr_size = [(abs(point[0]) + abs(point[1])) for point in constr]
|
|
129
|
+
dimension = 'horizontal' if constr_size[0] > constr_size[1] else 'vertical'
|
|
130
|
+
|
|
131
|
+
# Always start off moving in the positive direction
|
|
132
|
+
if dimension == 'horizontal':
|
|
133
|
+
if (agent.state.p_vel == 0).all():
|
|
134
|
+
action.u[0] = +1.0
|
|
135
|
+
else: # If moving, continue in the direction of velocity
|
|
136
|
+
if agent.state.p_pos[0] <= constr[0][0]: # If at left constraint, move right
|
|
137
|
+
action.u[0] = +1.0
|
|
138
|
+
elif agent.state.p_pos[0] >= constr[1][0]: # If at right constraint, move left
|
|
139
|
+
action.u[0] = -1.0
|
|
140
|
+
else: # Otherwise, follow the direction of velocity
|
|
141
|
+
action.u[0] = np.sign(agent.state.p_vel[0])
|
|
142
|
+
else:
|
|
143
|
+
if (agent.state.p_vel == 0).all():
|
|
144
|
+
action.u[1] = +1.0
|
|
145
|
+
else: # If moving, continue in the direction of velocity
|
|
146
|
+
if agent.state.p_pos[1] <= constr[0][1]: # If at bottom constraint, move up
|
|
147
|
+
action.u[1] = +1.0
|
|
148
|
+
elif agent.state.p_pos[1] >= constr[1][1]: # If at top constraint, move down
|
|
149
|
+
action.u[1] = -1.0
|
|
150
|
+
else: # Otherwise, follow the direction of velocity
|
|
151
|
+
action.u[1] = np.sign(agent.state.p_vel[1])
|
|
152
|
+
|
|
153
|
+
return action
|
|
File without changes
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.1
|
|
2
2
|
Name: Signal8
|
|
3
|
-
Version: 0.
|
|
3
|
+
Version: 0.4
|
|
4
4
|
Summary: A multi-agent environment inspired by the Lewis Signaling Game, featuring eight unique problem configurations with both static and dynamic adversaries.
|
|
5
5
|
Home-page: https://github.com/ethanmclark1/signal8
|
|
6
6
|
Author: Ethan Clark
|
|
@@ -16,13 +16,13 @@ Sig8 is an adapted version of the Simple environment, originally developed by th
|
|
|
16
16
|
|
|
17
17
|
# Signal8
|
|
18
18
|
|
|
19
|
-
This repository contains a simple multi-agent environment with continuous observations and a discrete action space, inspired by the Lewis Signaling game. The environment incorporates basic simulated physics to create a scenario where
|
|
19
|
+
This repository contains a simple multi-agent environment with continuous observations and a discrete action space, inspired by the Lewis Signaling game. The environment incorporates basic simulated physics to create a scenario where agents must communicate and collaborate effectively to navigate around both static and dynamic adversaries to reach a goal.
|
|
20
20
|
|
|
21
21
|
## Installation
|
|
22
22
|
|
|
23
23
|
```
|
|
24
|
-
git clone https://github.com/ethanmclark1/
|
|
25
|
-
cd
|
|
24
|
+
git clone https://github.com/ethanmclark1/signal8.git
|
|
25
|
+
cd signal8
|
|
26
26
|
pip install -r requirements.txt
|
|
27
27
|
pip install -e .
|
|
28
28
|
```
|
|
@@ -30,10 +30,10 @@ pip install -e .
|
|
|
30
30
|
## Usage
|
|
31
31
|
|
|
32
32
|
```
|
|
33
|
-
import
|
|
33
|
+
import signal8
|
|
34
34
|
|
|
35
|
-
env =
|
|
36
|
-
env.reset(options={
|
|
35
|
+
env = signal8.env(has_dynamic_adversaries=True)
|
|
36
|
+
env.reset(options={"problem_name": "v_cluster"}))
|
|
37
37
|
```
|
|
38
38
|
|
|
39
39
|
## List of Problem Scenarios
|
|
@@ -0,0 +1,14 @@
|
|
|
1
|
+
LICENSE
|
|
2
|
+
README.md
|
|
3
|
+
setup.py
|
|
4
|
+
Signal8/Signal8.py
|
|
5
|
+
Signal8/__init__.py
|
|
6
|
+
Signal8/main.py
|
|
7
|
+
Signal8.egg-info/PKG-INFO
|
|
8
|
+
Signal8.egg-info/SOURCES.txt
|
|
9
|
+
Signal8.egg-info/dependency_links.txt
|
|
10
|
+
Signal8.egg-info/top_level.txt
|
|
11
|
+
signal8.egg-info/PKG-INFO
|
|
12
|
+
signal8.egg-info/SOURCES.txt
|
|
13
|
+
signal8.egg-info/dependency_links.txt
|
|
14
|
+
signal8.egg-info/top_level.txt
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
Signal8
|
|
@@ -1 +0,0 @@
|
|
|
1
|
-
|
|
File without changes
|
|
File without changes
|
|
File without changes
|