Efficient and Complete FD-solving for extended array constraints

Authors: Quentin Plazar, Mathieu Acher, Sébastien Bardin, Arnaud Gotlieb

IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments show that the proposed solver significantly outperforms FD solvers, and successfully competes with the best SMT-solvers.
Researcher Affiliation Academia INRIA Bretagne-Atlantique, Rennes, France 2 CEA, LIST, Gif-sur-Yvette, F-91191, France 3 Simula Research Laboratory, Certus Center, Lysaker, Norway
Pseudocode Yes Algorithm 1: Propagator for consistent(L1, L2)
Open Source Code No The paper mentions implementing the technique inside fdcc, but does not provide a specific link or explicit statement about the availability of the source code for their implementation.
Open Datasets Yes First we take the 550 array formulas from the SMTCOMP benchmark, the standard competition in SMT solving with formulas coming essentially from hard verification-oriented industrial and academic case-studies.
Dataset Splits No The paper does not explicitly provide details about training/validation/test splits for the formulas used from the SMT-COMP benchmark, nor does it specify a cross-validation setup beyond using the benchmark formulas directly for evaluation.
Hardware Specification Yes Experiments were run on an Intel(R) Core(TM) i7-5600U (2,6 GHz) (2 cores), 16GB RAM, running Linux Fedora 22.
Software Dependencies No The paper mentions 'fdcc [Bardin and Gotlieb, 2012]', 'SICStus clpfd', 'Yices [Dutertre, 2014]', 'Math SAT [Cimatti et al., 2013]', 'CVC4 [Barrett et al., 2011]', and 'Z3 solver [De Moura and Bjørner, 2008]'. While it names software, it does not provide specific version numbers for these tools or libraries beyond the publication year of their foundational papers.
Experiment Setup Yes We compare each solver on the number of successfully resolved formulas... with a timeout set to 30 seconds... The results for a timeout of 120 seconds are also shown without discussion, since there is only very little difference.