DEFIANCE Module
ΒΆ
1. Design Documentation
1.1. Overview
1.1.1. Basic Components
1.1.2. Design Criteria
1.1.3. Class diagram
1.2. Customization
1.3. ChannelInterface
1.4. Design of RlApplications
1.4.1. RlApplication
1.4.2. AgentApplication
1.4.2.1. Basic Concept
1.4.2.2. Interaction with other RlApplications
1.4.2.3. Interaction with Python-based learning process
1.4.2.4. Receiving, storing and calculating observations resp. rewards
1.4.2.5. Execution of actions
1.4.2.6. Inference agents vs. training agents
1.4.3. DataCollectorApplication
1.4.4. ActionApplication
1.4.5. General Decisions
1.5. Interface for Multi-Agent RL
1.6. Helper
1.7. Framework expansion options
2. User Documentation
2.1. Overview
2.2. RL-Applications
2.2.1. RlApplication
2.2.1.1. RlApplicationId
2.2.2. AgentApplication
2.2.2.1. Additional Features and Use-Cases
2.2.2.1.1. Configure History Containers
2.2.2.1.2. Provide Extra Info
2.2.2.1.3. Action Delay
2.2.2.1.4. Override initiateAction and initiateActionForApp
2.2.2.1.5. OnRecvFromAgent
2.2.3. ObservationApplication
2.2.3.1. ObservationApplication::RegisterCallbacks
2.2.3.2. ObservationApplication::Send
2.2.4. RewardApplication
2.2.5. ActionApplication
2.2.5.1. ActionApplication::ExecuteAction
2.2.6. Communication between RL-Applications
2.2.6.1. Add interfaces
2.2.6.2. Send
2.2.6.3. AgentApplication Communication
2.2.6.4. RewardApplication Communication and ObservationApplication Communication
2.2.6.5. ActionApplication Communication
2.3. Data History Container
2.4. ChannelInterface
2.4.1. Overview
2.4.2. Usage
2.4.3. SimpleChannelInterface
2.4.4. SocketChannelInterface
2.4.5. Custom Channel Interface
2.5. Helper Classes
2.5.1. RlApplicationHelper
2.5.2. CommunicationHelper
3. References
DEFIANCE Module
Navigation
1. Design Documentation
2. User Documentation
3. References
Related Topics
Documentation overview
Next:
1.
Design Documentation
Quick search