Published December 2, 2020 | Version v1
Software Open

Code and experiment data for the AAAI 2021 paper "Saturated Post-hoc Optimization for Classical Planning"



The file contains an extended version of the Fast Downward planning system ( The code for saturated post-hoc cost partitioning can be found in "src/search/operator_counting/pho_abstraction_constraints.{h,cc}". Please see for detailed instructions on how to compile the planner. Here is the short version for building the planner and running saturated post-hoc optimization:

./ PDDL_TASK --search "astar(operatorcounting([pho_abstraction_constraints([projections(hillclimbing(max_generated_patterns=200, random_seed=0)), projections(systematic(2)), cartesian([landmarks(order=random, random_seed=0), goals(order=random, random_seed=0)])], saturated=true, forbid_useless_operators=true)]))"

The latest version of the code is available at


The file contains the STRIPS PDDL benchmarks from sequential optimization tracks of IPC 1998-2018.

Experiment data

The remaining zipfiles contain the raw experiment data, parsed values and basic reports for the experiments in the paper. For each experiment there are two files. The first contains the raw data of all experiment runs. The code directories and benchmark files have been removed to avoid duplication and to save space. The second file (* contains a "properties" file with all parsed values and an HTML report.


Files (148.0 MB)

Name Size Download all
3.5 MB Preview Download
128.9 MB Preview Download
13.1 MB Preview Download
2.5 MB Preview Download

Additional details


TAILOR – Foundations of Trustworthy AI - Integrating Reasoning, Learning and Optimization 952215
European Commission