Populate step info in environment, and finish notebook

src/primaite/config/_package_data/example_config.yaml

@@ -31,7 +31,7 @@ game:
   - UDP
 
 agents:
-  - ref: client_1_green_user
+  - ref: client_2_green_user
     team: GREEN
     type: GreenWebBrowsingAgent
     observation_space:

src/primaite/config/_package_data/example_config_2_rl_agents.yaml

@@ -25,7 +25,7 @@ game:
   - UDP
 
 agents:
-  - ref: client_1_green_user
+  - ref: client_2_green_user
     team: GREEN
     type: GreenWebBrowsingAgent
     observation_space:

src/primaite/game/game.py

@@ -113,7 +113,7 @@ class PrimaiteGame:
         self.update_agents(sim_state)
 
         # Apply all actions to simulation as requests
-        self.apply_agent_actions()
+        agent_actions = self.apply_agent_actions()  # noqa
 
         # Advance timestep
         self.advance_timestep()
@@ -131,12 +131,15 @@ class PrimaiteGame:
 
     def apply_agent_actions(self) -> None:
         """Apply all actions to simulation as requests."""
+        agent_actions = {}
         for agent in self.agents:
             obs = agent.observation_manager.current_observation
             rew = agent.reward_function.current_reward
             action_choice, options = agent.get_action(obs, rew)
+            agent_actions[agent.agent_name] = (action_choice, options)
             request = agent.format_request(action_choice, options)
             self.simulation.apply_request(request)
+        return agent_actions
 
     def advance_timestep(self) -> None:
         """Advance timestep."""

src/primaite/session/environment.py

@@ -29,7 +29,7 @@ class PrimaiteGymEnv(gymnasium.Env):
         # make ProxyAgent store the action chosen my the RL policy
         self.agent.store_action(action)
         # apply_agent_actions accesses the action we just stored
-        self.game.apply_agent_actions()
+        agent_actions = self.game.apply_agent_actions()
         self.game.advance_timestep()
         state = self.game.get_sim_state()
 
@@ -39,7 +39,7 @@ class PrimaiteGymEnv(gymnasium.Env):
         reward = self.agent.reward_function.current_reward
         terminated = False
         truncated = self.game.calculate_truncated()
-        info = {}
+        info = {"agent_actions": agent_actions}  # tell us what all the agents did for convenience.
         if self.game.save_step_metadata:
             self._write_step_metadata_json(action, state, reward)
         return next_obs, reward, terminated, truncated, info
@@ -172,7 +172,7 @@ class PrimaiteRayMARLEnv(MultiAgentEnv):
         # 1. Perform actions
         for agent_name, action in actions.items():
             self.agents[agent_name].store_action(action)
-        self.game.apply_agent_actions()
+        agent_actions = self.game.apply_agent_actions()
 
         # 2. Advance timestep
         self.game.advance_timestep()
@@ -186,7 +186,7 @@ class PrimaiteRayMARLEnv(MultiAgentEnv):
         rewards = {name: agent.reward_function.current_reward for name, agent in self.agents.items()}
         terminateds = {name: False for name, _ in self.agents.items()}
         truncateds = {name: self.game.calculate_truncated() for name, _ in self.agents.items()}
-        infos = {}
+        infos = {"agent_actions": agent_actions}
         terminateds["__all__"] = len(self.terminateds) == len(self.agents)
         truncateds["__all__"] = self.game.calculate_truncated()
         if self.game.save_step_metadata:

tests/assets/configs/bad_primaite_session.yaml

@@ -19,7 +19,7 @@ game:
   - UDP
 
 agents:
-  - ref: client_1_green_user
+  - ref: client_2_green_user
     team: GREEN
     type: GreenWebBrowsingAgent
     observation_space:

tests/assets/configs/eval_only_primaite_session.yaml

@@ -23,7 +23,7 @@ game:
   - UDP
 
 agents:
-  - ref: client_1_green_user
+  - ref: client_2_green_user
     team: GREEN
     type: GreenWebBrowsingAgent
     observation_space:

tests/assets/configs/multi_agent_session.yaml

@@ -29,7 +29,7 @@ game:
   - UDP
 
 agents:
-  - ref: client_1_green_user
+  - ref: client_2_green_user
     team: GREEN
     type: GreenWebBrowsingAgent
     observation_space:

tests/assets/configs/test_primaite_session.yaml

@@ -27,7 +27,7 @@ game:
   - UDP
 
 agents:
-  - ref: client_1_green_user
+  - ref: client_2_green_user
     team: GREEN
     type: GreenWebBrowsingAgent
     observation_space:

tests/assets/configs/train_only_primaite_session.yaml

@@ -23,7 +23,7 @@ game:
   - UDP
 
 agents:
-  - ref: client_1_green_user
+  - ref: client_2_green_user
     team: GREEN
     type: GreenWebBrowsingAgent
     observation_space: