Debugging State and State Transitions

Overview

When developing new states or troubleshooting state transitions, you can create a debug configuration YAML file to test specific scenarios with customized working memories.

Example: Testing ChatingState

Let's examine how to debug the ChatingState class:

@register(name='CHATING', type="state")
class ChatingState(BaseState):
    def __init__(self, character: Character, 
                 character_list: CharacterList, 
                 building_list: BuildingList,  
                 on_change_state, 
                 followed_states=[CharacterState.SUM],
                 main_prompt = PromptType.CHATING,
                 state_name = CharacterState.CHATING,
                 arbitrary_wm=dict()
    )    

Creating a Debug Configuration

You can customize state initialization arguments and working memory using a debug configuration file:

# config/debug_chat.yaml
States:
  Jack Brown:  
    ACT:
        followed_states: 
            - CHATING 
    CHATING: 
        arbitrary_wm: 
            act_obj: Emma Smith
            init_conversation: Hi, how are you?  
  Emma Smith:
    IDLE: {}

Configuration Components

1. Character Configuration: Define states for each character under their name
2. State Parameters: Customize state-specific settings
3. followed_states: Define valid state transitions
4. arbitrary_wm: Set initial working memory values

Running the Debug Configuration

Execute your simulation with the debug configuration:

DEBUG=1 python main.py configs/debug_chat.yaml

Best Practices

1. Working Memory Testing
2. Use arbitrary_wm to simulate different memory states
3. Test edge cases and boundary conditions

4. Verify memory persistence across state transitions

5. State Transitions

6. Configure followed_states to test transition paths

7. Verify state entry and exit conditions

8. Character Interactions

9. Set up multiple characters to test interactions
10. Verify message passing between states
11. Test concurrent state transitions