WLPlan: Relational Features for PDDL Planning
Project description
WLPlan
WLPlan is a package for generating embeddings of PDDL planning problems for machine learning tasks. It supports both classical and numeric planning problems.
Installation
Python Interface
The Python interface can be installed simply with
sh install.sh
C++ Interface
The C++ interface can be installed in your project by running
./cmake_build.py <path/to/installation>
and adding the following to the root CMakeLists.txt file of your project
list(APPEND CMAKE_PREFIX_PATH "<path/to/installation>")
find_package(wlplan)
...
target_link_libraries(<your_project> PRIVATE wlplan)
Graph Representations
The graph representations of planning tasks implemented thus far are
Name | WLPlan shorthand | Reference |
---|---|---|
Instance Learning Graph (ILG) | ilg |
D. Chen, F. Trevizan, S. Thiébaux: Return to Tradition: Learning Reliable Heuristics with Classical Machine Learning. ICAPS 2024 |
Numeric Instance Learning Graph (νILG) | nilg |
D. Chen, S. Thiébaux: Graph Learning for Numeric Planning. NeurIPS 2024 |
Feature Generators
The feature generators implemented thus far are
Name | WLPlan shorthand | Reference |
---|---|---|
Weisfeiler Leman (WL)/colour refinement | wl |
D. Chen, F. Trevizan, S. Thiébaux: Return to Tradition: Learning Reliable Heuristics with Classical Machine Learning. ICAPS 2024 |
2-WL | kwl2 |
J. Cai, M. Fürer, N. Immerman: An optimal lower bound on the number of variables for graph identification. Combinatorica (1992) |
Local 2-WL (2-LWL) | lwl2 |
C. Morris, K. Kersting, P. Mutzel: Glocalized Weisfeiler-Lehman Graph Kernels: Global-Local Feature Maps of Graphs. ICDM 2017 |
continuous-categorical WL (ccWL) | ccwl |
D. Chen, S. Thiébaux: Graph Learning for Numeric Planning. NeurIPS 2024 |
individualised WL (iWL) | iwl |
coming soon |
normalised iWL (niWL) | niwl |
coming soon |
Usage
Examples for how to use the package include this self-contained notebook and test, as well as the GOOSE framework.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distributions
No source distribution files available for this release.See tutorial on generating distribution archives.
Built Distributions
Close
Hashes for wlplan-1.0.1-cp312-cp312-manylinux_2_34_x86_64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | c2a8af807636c2eb0e507bbbc1f72a444117159fe53e603d40d4c0b6ac666c8b |
|
MD5 | c397e161eeb5e2349e3b5afebe0ccb0a |
|
BLAKE2b-256 | 85a3f7566983071bec8a982c9d3f1d93fdb1c11176a897f75d200541c7a22da0 |
Close
Hashes for wlplan-1.0.1-cp311-cp311-manylinux_2_34_x86_64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 06aa3c338789149d38d156b726a78eb12cd7eff4f62ce699752562b6ae2efc3e |
|
MD5 | 44ccc5f3e1e54f5ddbd7b273045698c5 |
|
BLAKE2b-256 | 634b029c9e90212a089eed8445689115723bef18a8d1593990ac983e0d2d6cf5 |
Close
Hashes for wlplan-1.0.1-cp310-cp310-manylinux_2_34_x86_64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | a4c37be5f838352609b017b174c459e7cbbcf5c8bbe4f55f2673b4b2582cf190 |
|
MD5 | 22dbded075a1f137489b114a91aa0d40 |
|
BLAKE2b-256 | 630a385f092aeb46f802428d54c3d41c955dafb841d688f38878ddf5c259ca0f |