Skip to content

README.md

Latest commit

 

History

History
148 lines (112 loc) · 5.15 KB

README.md

File metadata and controls

148 lines (112 loc) · 5.15 KB
Raw

jupyddl — Python PDDL

✨ A dependency-free, pure-Python PDDL planning framework: parser, grounder, classical-to-SOTA planners, heuristics and a benchmarking harness. ✨

tests build GitHub license

jupyddl used to be a thin wrapper around the Julia PDDL.jl parser. It has been rewritten from scratch as a pure-Python framework — no Julia, no native dependencies, just the standard library — so it is trivial to install, embed and build on top of for research and teaching.

Features 🌱

  • 🧩 Hand-written PDDL parser and grounder (typing, negative preconditions, equality, action costs, and forall/when conditional effects).
  • 🔎 Uninformed planners: BFS, DFS, Iterative Deepening.
  • 🧭 Informed planners: Dijkstra (uniform cost), Greedy Best-First, A*, Weighted A*, IDA*, and Enforced Hill Climbing (FF-style).
  • 📊 Heuristics from classical to SOTA: blind, goal-count, h_max, h_add, FF (h_ff), critical-path h^m (h1/h2), and LM-cut.
  • ⚖️ Benchmarking harness for comparative analysis with CSV export and plots.
  • 🧱 Extensible by design: planners and heuristics live behind simple registries; add your own in a few lines.
  • ✅ Zero runtime dependencies and a comprehensive test suite.

Install 💾

Requires Python ≥ 3.9. Using uv (recommended):

uv venv
uv pip install -e ".[dev,viz]"   # 'viz' pulls matplotlib for benchmark plots

or with plain pip:

python -m pip install -e ".[dev,viz]"

Command line ⚔️

# Solve a single instance
jupyddl solve pddl-examples/tsp/domain.pddl pddl-examples/tsp/problem.pddl \
    --search astar --heuristic lmcut

# Compare planners over a folder of <name>/{domain,problem}.pddl instances
jupyddl benchmark pddl-examples \
    --planners bfs,dijkstra,astar,gbfs,ehc --heuristic hff \
    --csv results.csv --plot comparison.png

Library usage 📑

from jupyddl import solve, build_task, solve_task, validate_plan

# One-shot solve
result = solve("pddl-examples/tsp/domain.pddl",
               "pddl-examples/tsp/problem.pddl",
               search="astar", heuristic="lmcut")
print(result.solved, result.cost, result.plan_names())

# Ground once, try several configurations
task = build_task("pddl-examples/blocksworld/domain.pddl",
                  "pddl-examples/blocksworld/problem.pddl")
for cfg in [("astar", "lmcut"), ("gbfs", "hff"), ("bfs", None)]:
    r = solve_task(task, cfg[0], cfg[1])
    assert validate_plan(task, r.plan)
    print(cfg, r.cost, r.stats.expanded, "expanded")

Available planners & heuristics

Planners Heuristics
bfs, dfs, iddfs, dijkstra, gbfs, astar, wastar, idastar, ehc blind, goalcount, hmax, hadd, hff, h1, h2/hm, lmcut

A*/IDA* are cost-optimal with an admissible heuristic (blind, hmax, h1/h2, lmcut); dijkstra and bfs are optimal without a heuristic.

Supported PDDL subset

STRIPS, :typing (with hierarchy), :negative-preconditions, :equality, :action-costs ((increase (total-cost) k)), universally-quantified goals, and forall/when conditional effects. Numeric fluents beyond total-cost and :durative-actions are out of scope and rejected with a clear error.

Note: on domains with conditional effects (e.g. flip), the delete-relaxation heuristics (hadd, hff, lmcut, h^m) are satisficing only — admissibility is not guaranteed. Use bfs, dijkstra, or astar/idastar with blind for guaranteed-optimal plans on such domains.

Architecture 🏗️

jupyddl/
  parser/       tokenizer + AST + recursive-descent parser
  grounding.py  Domain+Problem -> grounded Task (typing, PNF, static pruning)
  task.py       grounded STRIPS(+conditional-effects) task & operators
  search/       planners + shared best-first engine + registry
  heuristics/   heuristics + delete-relaxation machinery + registry
  benchmark.py  comparative benchmarking (CSV + plots)
  cli.py        `jupyddl solve` / `jupyddl benchmark`

Add a planner by subclassing jupyddl.search.Planner (or reusing best_first) and registering it in jupyddl.search.PLANNERS; add a heuristic by subclassing jupyddl.heuristics.Heuristic and registering it in jupyddl.heuristics.HEURISTICS.

Development 🛠️

git submodule update --init      # fetch the pddl-examples used by the tests
uv pip install -e ".[dev,viz]"
pytest --cov=jupyddl             # run the test suite
flake8 jupyddl tests             # lint

Cite 📰

@misc{https://doi.org/10.13140/rg.2.2.22418.89282,
  doi = {10.13140/RG.2.2.22418.89282},
  url = {http://rgdoi.net/10.13140/RG.2.2.22418.89282},
  author = {Erwin Lejeune},
  title = {Jupyddl, an extensible python library for PDDL planning and parsing},
  year = {2021}
}

Maintainers Ⓜ️

  • Erwin Lejeune
  • Sampreet Sarkar