Benchmarks, Code and Data from Mugdan et al, ICAPS 2023
Description
This bundle contains all benchmarks, code, and data used in the paper
"Optimality Certificates for Classical Planning" published at ICAPS 2023.
Benchmarks
Contains two folders:
- original: The original benchmark set, which is a clone of
https://github.com/aibasel/downward-benchmarks. Note that we do not use the
full benchmark set but only a subset used for optimal planning.
- unsolvability-compilation: The compilation to unsolvability we created with
the script "cost_parser.py" (see under "Code").
Code
Contains the following folders and scripts:
- cost-extractor.py: A script to extract optimal cost for tasks from a
properties file created by a Downward Lab experiment.
- cost_parser.py: A script that takes the information generated from
cost-extractor.py and performs the compilation to unsolvability as described
in the paper. (Only supports unit-cost.)
- downward-optimality-certificates: A modified version of Fast Downward
(https://github.com/aibasel/downward/) that we used to create optimality
certificates.
- downward-unsolvability: A clone of
https://github.com/salome-eriksson/downward-unsolvability used for generating
unsolvability certificates.
- helve-optimality: A modified version of Helve (see below) that implements
rules for deriving lower bounds.
- helve-unsolvability: A clone of https://github.com/salome-eriksson/helve/
used for verifying unsolvability certificates.
Data
Contains three folders:
- optimality / unsolvability: A copy of all scripts and logs of the
experiments run with Downward Lab (https://lab.readthedocs.io/en/stable/).
Experiments consist of:
- A script detailing how the experiment is to be conducted (benchmark set,
code, limits, ...)
- custom parsers
- data/<experiment-name>: Contains the logs of each run
- data/<experiment-name>-eval: Contains "properties", a JSON file that
summarizes the experiment based on the parsed attributes.
(For the lmcut experiment only the properties file is present.)
- analysis: Contains html reports and .tex plots, as well as the scripts that
generated them.