Gym interface

.gitignore

@@ -0,0 +1,118 @@
+# Pycharm
+.idea
+
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+.hypothesis/
+.pytest_cache/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+.python-version
+
+# celery beat schedule file
+celerybeat-schedule
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/

gym_example.py

@@ -1,88 +1,34 @@
-from rover_domain_core_gym import RoverDomainCoreGym
-from mods import *
-import datetime
-from code.world_setup import * # Rover Domain Construction 
-from code.agent_domain_2 import * # Rover Domain Dynamic  
-from code.trajectory_history import * # Agent Position Trajectory History 
-from code.reward_2 import * # Agent Reward 
-from code.reward_history import * # Performance Recording 
-from code.ccea_2 import * # CCEA 
-from code.save_to_pickle import * # Save data as pickle file
-import random
+"""
+An example using the rover domain gym-style interface and the standard, included CCEA learning algorithms.
+This is a minimal example, showing the minimal Gym interface.
+"""
+from rover_domain_core_gym import RoverDomainGym
+import code.ccea_2 as ccea
+import code.agent_domain_2 as domain
 
+episodeCount = 1000  # Number of learning episodes
 
-stepCount = 5
-trainCountXEpisode = 3
-testCountXEpisode = 1
-episodeCount =  20
+sim = RoverDomainGym()
 
-# NOTE: Add the mod functions (variables) to run to modCol here:
-modCol = [
-    globalRewardMod,
-    differenceRewardMod,
-    dppRewardMod
-]
+ccea.initCcea(input_shape=8, num_outputs=2, num_units=32)(sim.data)
 
-i = 0
-while True:
-    print("Run %i"%(i))
-    random.shuffle(modCol)
-    for mod in modCol:
-        sim = RoverDomainCoreGym()
-        mod(sim)
-
-        #Trial Begins
-        createRewardHistory(sim.data)
-        initCcea(input_shape= 8, num_outputs=2, num_units = 32)(sim.data)
-        sim.data["Steps"] = stepCount
-
-        for episodeIndex in range(episodeCount):
-            sim.data["Episode Index"] = episodeIndex
-
-            # Training Phase
-
-            obs = sim.reset('Train', True)
-
-            for worldIndex in range(trainCountXEpisode):
-                sim.data["World Index"] = worldIndex
-                obs = sim.reset('Train', False)
-                assignCceaPolicies(sim.data)
-
-                done = False
-                stepCount = 0
-                while not done:
-                    doAgentProcess(sim.data)
-                    jointAction = sim.data["Agent Actions"]
-                    obs, reward, done, info = sim.step(jointAction)
-                    stepCount += 1
-
-                rewardCceaPolicies(sim.data)
-
-
-            # Testing Phase
-
-            obs = sim.reset('Test', True)
-            assignBestCceaPolicies(sim.data)        
-
-            for worldIndex in range(testCountXEpisode):
-                sim.data["World Index"] = worldIndex
-                obs = sim.reset('Test', False)
-
-                done = False
-                stepCount = 0
-                while not done:
-                    doAgentProcess(sim.data)
-                    jointAction = sim.data["Agent Actions"]
-                    obs, reward, done, info = sim.step(jointAction)
-                    stepCount += 1
-
-            evolveCceaPolicies(sim.data)
-            updateRewardHistory(sim.data)
-
-            # Trial End
-            saveRewardHistory(sim.data)
-            saveTrajectoryHistories(sim.data)
-
-
-
-    i += 1
+for episodeIndex in range(episodeCount):
+    sim.data["Episode Index"] = episodeIndex
+
+    obs = sim.reset()
+    ccea.assignCceaPolicies(sim.data)
+
+    done = False
+    stepCount = 0
+    while not done:
+        # Select actions and create the joint action from the simulation data
+        # Note that this specific function extracts "obs" from the data structure directly, which is why obs is not
+        # directly used in this example.
+        domain.doAgentProcess(sim.data)
+        jointAction = sim.data["Agent Actions"]
+        obs, reward, done, info = sim.step(jointAction)
+        stepCount += 1
+
+    ccea.rewardCceaPolicies(sim.data)
+    ccea.evolveCceaPolicies(sim.data)
+    print(sim.data["Global Reward"])