Welcome to pyRDDLGym’s Documentation!

Contents:

• Introduction
  • Purpose and Benefits
  • Status
  • License
  • Citing pyRDDLGym
  • Contributors
• Installation Guide
  • Requirements
  • Installing via pip
  • Installing the Pre-Release Version via git
• Getting Started: Basics
  • Initializing Built-In Environments
  • Initializing Environments from RDDL Files
  • Policies
  • Interacting with an Environment
  • Fixing the Random Seed
  • Spaces
  • Using Built-In Visualizers
  • Using a Custom Visualizer
  • Recording Movies
  • Logging Simulation Data
• Getting Started: Advanced Tools
  • Inspecting the Model
  • Grounding a Domain
  • Vectorized Input and Output
  • Exception Handling
  • Generating Debug Logs
  • Running pyRDDLGym through TCP
• Managing Problems using rddlrepository
  • The Repository Manager
  • Retrieving Specific Problems
  • Loading Environments in pyRDDLGym
  • Registering your Own Problems and Instances
• pyRDDLGym-jax: JAX Compiler and Planner
  • Requirements
  • Installing via pip
  • Installing the Pre-Release Version via git
  • Changing the Simulation Backend to JAX
  • Differentiable Planning: Deterministic Domains
  • Differentiable Planning: Stochastic Domains
  • Running the Basic Example
  • Running from the Python API
  • Writing Configuration Files for Custom Problems
  • Writing Configuration Files for Policy Networks
  • Boolean Actions
  • Constraints on Action Fluents
  • Reward Normalization
  • Utility Optimization
  • Changing the Planning Algorithm
  • Automatically Tuning Hyper-Parameters
  • Dealing with Non-Differentiable Expressions
  • Computing the Gradients Manually
  • Limitations
• pyRDDLGym-gurobi: Gurobi MIP Compiler and Planner
  • Requirements
  • Installing via pip
  • Running the Basic Example
  • Running from the Python API
  • Passing Parameters to the Gurobi Backend
  • Current Limitations
• pyRDDLGym-prost: PROST Planner
  • Installation
  • Building the Image
  • Running the Container
  • Using the Convenience Script
  • Reporting PROST Bugs
• pyRDDLGym-rl: Reinforcement Learning
  • Requirements
  • Installing via pip
  • Installing the Pre-Release Version via git
  • Running the Basic Stable Baselines 3 Example
  • Running the Basic RLlib Example
  • Running Stable Baselines 3 from the Python API
  • Running RLlib from the Python API
  • The Environment Wrapper
  • Limitations
• Symbolic Toolset for pyRDDLGym
  • Installation
  • XADD Compilation of CPFs
  • Visualizing DBNs with XADD
  • Symbolic Dynamic Programming (SDP)
  • Citations
• RDDL Language Description
  • Introduction
  • File Structure and Syntax
  • Functions and Expressions
  • Conditional Expressions
  • Logical Expressions
  • Arithmetic Operators and Aggregations
  • Probability Distributions
  • New Language Features