numeric-pddl-gym 1.2.0

Convert a Numeric PDDL domain into an Gymnasium environment

Numeric PDDLGym

A framework for automatically translating Numeric PDDL domains with the Gymnasium API. This allows Numeric PDDL planning problems to be solved using standard RL algorithms (e.g., PPO via RLlib).

The environment converts states, actions, and goals from Numeric PDDL into fixed-size numeric vectors, enabling direct integration with deep RL libraries.

✨ Features

• Supports Numeric PDDL domains and problems
• Automatic grounding and vectorization of:
  • Predicates
  • Numeric fluents
  • Goal conditions
• Compatible with the Gymnasium API
• Designed for RLlib (Ray) integration
• Works with standard deep RL algorithms (e.g., PPO)

⚙️ Configuration

The environment is configured via a dictionary (env_config).

Parameter	Type	Relevant Env	Description
domain_path	Path	All	Path to the PDDL domain file
problems_list	List[Path]	All	List of problem files
max_steps	int	All	Max steps per episode
executing_algorithm	str	All	Name of RL algorithm (e.g., "PPO")
masking_strategy	"pre" or "post"	PDDLMaskedEnv	"pre": Filters invalid actions before execution using PDDL checks. "post": Learns invalid actions after execution from feedback.
count_inapplicable	bool	PDDLMaskedEnv	Relevant when using "post" masking. If true, inapplicable actions count toward the episode length; if false, they are ignored.
map_size	int	MinecraftEnv	Grid size, e.g., 6, 10, 15.

🚀 Installation

pip install numeric_pddl_gym

With RL (RLlib + Torch):

pip install numeric_pddl_gym[rl_agents]

▶️ Run Example

RL Agent

python rl_agents/ppo_pddl_rllib_agent.py

Planning Agent

See example for a planning-based agent in:

test/test_fixed_script_agent.py

⚠️ Limitations

1. Can't encode complex goal conditions.
2. Agents must be retrained if the problem has a different number of fluents, predicates, or goal conditions.
3. Currently does not support manual interaction with the environment.
4. Designed for fixed-structure problems (no variable-sized domains).
5. Enabling action applicability checking leads to slow runtime in large-scale problems.

📚 Citations

Coming soon