From e1ce40085b77bf0c85658b94892b504cc3562932 Mon Sep 17 00:00:00 2001
From: MLErik <baerenjesus@gmail.com>
Date: Thu, 7 Nov 2019 15:38:29 -0500
Subject: [PATCH] updated files to new malfunction behavior
---
torch_training/multi_agent_inference.py | 11 ++++++-----
torch_training/multi_agent_training.py | 13 +++++++------
torch_training/render_agent_behavior.py | 12 ++++++------
torch_training/training_navigation.py | 14 ++++++++------
4 files changed, 27 insertions(+), 23 deletions(-)
diff --git a/torch_training/multi_agent_inference.py b/torch_training/multi_agent_inference.py
index 1866443..0f9d1e5 100644
--- a/torch_training/multi_agent_inference.py
+++ b/torch_training/multi_agent_inference.py
@@ -3,6 +3,7 @@ from collections import deque
import numpy as np
import torch
+from flatland.envs.malfunction_generators import malfunction_from_params
from flatland.envs.observations import TreeObsForRailEnv
from flatland.envs.predictions import ShortestPathPredictorForRailEnv
from flatland.envs.rail_env import RailEnv
@@ -36,12 +37,12 @@ n_agents = 10
observation_builder = TreeObsForRailEnv(max_depth=2)
# Use a the malfunction generator to break agents from time to time
-stochastic_data = {'prop_malfunction': 0.0, # Percentage of defective agents
- 'malfunction_rate': 30, # Rate of malfunction occurence
- 'min_duration': 3, # Minimal duration of malfunction
- 'max_duration': 20 # Max duration of malfunction
+stochastic_data = {'malfunction_rate': 8000, # Rate of malfunction occurence of single agent
+ 'min_duration': 15, # Minimal duration of malfunction
+ 'max_duration': 50 # Max duration of malfunction
}
+
# Custom observation builder
TreeObservation = TreeObsForRailEnv(max_depth=2, predictor=ShortestPathPredictorForRailEnv())
@@ -61,7 +62,7 @@ env = RailEnv(width=x_dim,
max_rails_in_city=2),
schedule_generator=sparse_schedule_generator(speed_ration_map),
number_of_agents=n_agents,
- stochastic_data=stochastic_data, # Malfunction data generator
+ malfunction_generator_and_process_data=malfunction_from_params(stochastic_data),
obs_builder_object=TreeObservation)
env.reset(True, True)
diff --git a/torch_training/multi_agent_training.py b/torch_training/multi_agent_training.py
index 2e20c63..93c7eff 100644
--- a/torch_training/multi_agent_training.py
+++ b/torch_training/multi_agent_training.py
@@ -2,9 +2,11 @@ import getopt
import random
import sys
from collections import deque
-
# make sure the root path is in system path
from pathlib import Path
+
+from flatland.envs.malfunction_generators import malfunction_from_params
+
base_dir = Path(__file__).resolve().parent.parent
sys.path.append(str(base_dir))
@@ -41,10 +43,9 @@ def main(argv):
# Use a the malfunction generator to break agents from time to time
- stochastic_data = {'prop_malfunction': 0.0, # Percentage of defective agents
- 'malfunction_rate': 30, # Rate of malfunction occurence
- 'min_duration': 3, # Minimal duration of malfunction
- 'max_duration': 20 # Max duration of malfunction
+ stochastic_data = {'malfunction_rate': 8000, # Rate of malfunction occurence of single agent
+ 'min_duration': 15, # Minimal duration of malfunction
+ 'max_duration': 50 # Max duration of malfunction
}
# Custom observation builder
@@ -66,7 +67,7 @@ def main(argv):
max_rails_in_city=3),
schedule_generator=sparse_schedule_generator(speed_ration_map),
number_of_agents=n_agents,
- stochastic_data=stochastic_data, # Malfunction data generator
+ malfunction_generator_and_process_data=malfunction_from_params(stochastic_data),
obs_builder_object=TreeObservation)
# After training we want to render the results so we also load a renderer
diff --git a/torch_training/render_agent_behavior.py b/torch_training/render_agent_behavior.py
index dd81bea..b32e4df 100644
--- a/torch_training/render_agent_behavior.py
+++ b/torch_training/render_agent_behavior.py
@@ -3,6 +3,7 @@ from collections import deque
import numpy as np
import torch
+from flatland.envs.malfunction_generators import malfunction_from_params
from flatland.envs.observations import TreeObsForRailEnv
from flatland.envs.rail_env import RailEnv
from flatland.envs.rail_generators import sparse_rail_generator
@@ -37,10 +38,9 @@ min_dist = 5
observation_builder = TreeObsForRailEnv(max_depth=2)
# Use a the malfunction generator to break agents from time to time
-stochastic_data = {'prop_malfunction': 0.0, # Percentage of defective agents
- 'malfunction_rate': 30, # Rate of malfunction occurence
- 'min_duration': 3, # Minimal duration of malfunction
- 'max_duration': 20 # Max duration of malfunction
+stochastic_data = {'malfunction_rate': 80, # Rate of malfunction occurence of single agent
+ 'min_duration': 15, # Minimal duration of malfunction
+ 'max_duration': 50 # Max duration of malfunction
}
# Custom observation builder
@@ -59,10 +59,10 @@ env = RailEnv(width=x_dim,
seed=1, # Random seed
grid_mode=False,
max_rails_between_cities=2,
- max_rails_in_city=2),
+ max_rails_in_city=4),
schedule_generator=sparse_schedule_generator(speed_ration_map),
number_of_agents=n_agents,
- stochastic_data=stochastic_data, # Malfunction data generator
+ malfunction_generator_and_process_data=malfunction_from_params(stochastic_data),
obs_builder_object=TreeObservation)
env.reset()
diff --git a/torch_training/training_navigation.py b/torch_training/training_navigation.py
index daac62d..fe4905f 100644
--- a/torch_training/training_navigation.py
+++ b/torch_training/training_navigation.py
@@ -2,9 +2,11 @@ import getopt
import random
import sys
from collections import deque
-
# make sure the root path is in system path
from pathlib import Path
+
+from flatland.envs.malfunction_generators import malfunction_from_params
+
base_dir = Path(__file__).resolve().parent.parent
sys.path.append(str(base_dir))
@@ -40,10 +42,9 @@ def main(argv):
# Use a the malfunction generator to break agents from time to time
- stochastic_data = {'prop_malfunction': 0.0, # Percentage of defective agents
- 'malfunction_rate': 30, # Rate of malfunction occurence
- 'min_duration': 3, # Minimal duration of malfunction
- 'max_duration': 20 # Max duration of malfunction
+ stochastic_data = {'malfunction_rate': 8000, # Rate of malfunction occurence of single agent
+ 'min_duration': 15, # Minimal duration of malfunction
+ 'max_duration': 50 # Max duration of malfunction
}
# Custom observation builder
@@ -65,7 +66,8 @@ def main(argv):
max_rails_in_city=3),
schedule_generator=sparse_schedule_generator(speed_ration_map),
number_of_agents=n_agents,
- stochastic_data=stochastic_data, # Malfunction data generator
+ malfunction_generator_and_process_data=malfunction_from_params(stochastic_data),
+ # Malfunction data generator
obs_builder_object=TreeObservation)
# After training we want to render the results so we also load a renderer
--
GitLab