Skip to content

Classical and FOND Planning for Pure-Past Linear Temporal Logic Goals

License

Notifications You must be signed in to change notification settings

whitemech/Plan4Past

Repository files navigation

Python Version PyPI build codecov GitHub


An award-winning tool for Classical and FOND Planning for Pure-Past Linear Temporal Logic (PPLTL) Goals.

Installation

  • from PyPI:
pip install plan4past
  • from source (main branch):
pip install git+https://github.com/whitemech/Plan4Past.git

or, clone the repository and install:

git clone https://github.com/whitemech/Plan4Past.git
cd Plan4Past
pip install .

Quickstart

You can use the plan4past package in two ways: as a library, and as a CLI tool.

As a library

This is an example of how you can encode a PPLTL goal formula into a PDDL domain and problem programmatically.

from pathlib import Path
from pddl.formatter import domain_to_string, problem_to_string
from pddl.parser.domain import DomainParser
from pddl.parser.problem import ProblemParser
from pylogics.parsers import parse_pltl
from plan4past.compiler import Compiler


domain_path = Path("tests/benchmarks/deterministic/BF/blocksworld_ppltl/domain.pddl")
problem_path = Path("tests/benchmarks/deterministic/BF/blocksworld_ppltl/p2.pddl")

formula = "on_b_a & O(ontable_c)"
domain_parser = DomainParser()
problem_parser = ProblemParser()

domain = domain_parser(domain_path.read_text(encoding="utf-8"))
problem = problem_parser(problem_path.read_text(encoding="utf-8"))
goal = parse_pltl(formula)

compiler = Compiler(domain, problem, goal)
compiler.compile()
compiled_domain, compiled_problem = compiler.result

try:
    with open("./new-domain.pddl", "w+", encoding="utf-8") as d:
        d.write(domain_to_string(compiled_domain))
    with open("./new-problem.pddl", "w+", encoding="utf-8") as p:
        p.write(problem_to_string(compiled_problem))
except Exception as e:
    raise IOError(
        "[ERROR]: Something wrong occurred while writing the compiled domain and problem."
    ) from e

By executing the code above, you will obtain a new-domain.pddl as well as a new-problem.pddl in output.

As a CLI tool

The package can also be used as a CLI tool. The supported command is:

plan4past -d DOMAIN_FILE -p PROBLEM_FILE -g "PPLTL_FORMULA" [-m MAPPING] 

where DOMAIN_FILE and PROBLEM_FILE are the PDDL domain and problem files, respectively, and PPLTL_FORMULA is the goal formula expressed in Pure-Past Linear Temporal Logic. The optional MAPPING file is a mapping between the PDDL objects and the PPLTL formula's atomic propositions.

For instance:

plan4past -d examples/pddl/domain.pddl -p examples/pddl/p-0.pddl -g "ontable_c & O(on_b_a)"

Docker & Apptainer images

A Docker image as well as an Apptainer image are available for Plan4Past.

  • To use Docker: docker build -t plan4past . and docker run --rm -it plan4past

  • To use Apptainer:

apptainer build plan4past.sif plan4past.def

# to launch the tool:
./plan4past.sif -d examples/pddl/domain.pddl -p examples/pddl/p-0.pddl -g "ontable_c & O(on_b_a)"

Development

If you want to contribute, set up your development environment as follows:

  • Intall Poetry
  • Clone the repository: git clone https://github.com/whitemech/Plan4Past.git && cd Plan4Past
  • Install the dependencies: poetry shell && poetry install

Docs

To build the docs: mkdocs build

To view documentation in a browser: mkdocs serve and then go to http://localhost:8000

License

Plan4Past is released under the GNU Lesser General Public License v3.0 or later (LGPLv3+).

Copyright 2021 -- 2023 WhiteMech

Citing

If you use Plan4Past in your research, please consider citing the following papers.

  • For deterministic (classical) planning:
@inproceedings{icaps2023bdffgs,
  author       = {Luigi Bonassi and
                  Giuseppe {De Giacomo} and
                  Marco Favorito and
                  Francesco Fuggitti and
                  {Alfonso Emilio} Gerevini and
                  Enrico Scala},
  title        = {{Planning for Temporally Extended Goals in Pure-Past Linear Temporal Logic}},
  booktitle    = {{ICAPS}},
  pages        = {61--69},
  publisher    = {{AAAI} Press},
  year         = {2023}
}
  • For non-deterministic (FOND) planning:
@inproceedings{ecai2023bdffgs,
  author       = {Luigi Bonassi and
                  Giuseppe {De Giacomo} and
                  Marco Favorito and
                  Francesco Fuggitti and
                  {Alfonso Emilio} Gerevini and
                  Enrico Scala},
  title        = {{{FOND} Planning for Pure-Past Linear Temporal Logic Goals}},
  booktitle    = {{ECAI}},
  year         = {2023},
}

Awards 🏆

Plan4Past has been awarded the Best Student Paper Award at ICAPS 2023.

Acknowledgements

Plan4Past has been partially supported by the EU H2020 project AIPlan4EU (No. 101016442), the ERC-ADG WhiteMech (No. 834228), the EU ICT-48 2020 project TAILOR (No. 952215), the PRIN project RIPER (No. 20203FFYLK), and the PNRR MUR project FAIR (No. PE0000013).