alviano / wasp Goto Github PK

License: Apache License 2.0

C++ 0.57% Python 98.23% Shell 0.01% Makefile 0.01% Classic ASP 1.20%

wasp's Introduction

Wasp

Wasp is an ASP solver handling disjunctive logic programs under the stable model semantics. Wasp implements techniques originally introduced for SAT solving combined with optimization methods that have been specifically designed for ASP computation, such as source pointers enhancing unfounded sets computation and efficient techniques for stable model checking.

Wasp takes as input logic programs in the GRINGO format and SAT/MaxSAT formulas in DIMACS format.

Compilation

In order to compile wasp boost (http://www.boost.org/) and g++-4.6 (or more recent) are required.

If all libraries have been installed just type:

make

and pray!

Other compiling options are also available:

make BUILD=stats enables the usage of statistics
make BUILD=trace enables the possibility to trace the internal behavior of wasp
make BUILD=debug compiles wasp for debugging

Usage

In order to use wasp you need a grounder. You can use either GRINGO (http://potassco.sourceforge.net/) or I-DLV (https://github.com/DeMaCS-UNICAL/I-DLV/wiki).

If the grounder has been downloaded just type:

./gringo filename | ./wasp

or

./i-dlv filename | ./wasp

Team

Mario Alviano, University of Calabria
Carmine Dodaro, University of Calabria (main developer)
Francesco Ricca, University of Calabria

For any doubt or suggestion feel free to contact us at [email protected].

wasp's People

Contributors

Stargazers

Watchers

Forkers

gaste dbaeck yarrick13 knowlp bernardocuteri danhlephuoc mourad-hassani mahi045

wasp's Issues

Can't set up the python API

Is there a tutorial on how to use the python API of Wasp ?

Variable Name

Remove string name from class Variable.

Aging

void
BerkminHeuristic::onLearning()
{
numberOfConflicts++;
//TODO: Implement aging.
}

Compute anytime cautious reasoning

I am using wasp to compute cautious reasoning.
What is the command to compute anytime cautious reasoning using wasp? Is it enough to use the params --minimization-cautious-enum or --minimization-cautious with ./wasp command?

Thanks in advance.

Cant get WASP to work with a grounder

Hi,

When I try to use WASP with a grounder (gringo or idlv) it does not seem to work. It says that the output provided by the grounders is "unsupported rule type". I hope someone could help me fix this issue.

Sincerely,

Segfault on newer GCC versions

WASP produces a segfault when compiled with GCC 5.3.1 (on Fedora 23). This has been tested using graph coloring instances from ASPCOMP 2014, but it also fails most of the tests included with WASP.

The bug occurs when WASP is compiled with -O2 and above, but not with -O1. Compiling with -fno-strict-aliasing and -fsanitize=undefined sheds some light on the issue. As far as I understand, the call to memcpy in src/stl/Vector.h:83 is given a null pointer at some point. GCC assumes (in accordance to the standard, see C11 standard §7.1.4), that this function must never be passed a null pointer and optimizes the checks away at high enough optimization levels, eventually resulting in a segfault.

Some research suggests that the bug might occur with any GCC version > 4.9. See the porting guide for 4.9, in particular the section "Null pointer checks may be optimized away more aggressively". Indeed, the code does work on GCC 4.8.4.

I've attached an instance from ASPCOMP 2014 which produces the error for me.

graphcol.txt

Executing Wasp in stream reasoning mode

Hi,

The code available at http://distributed-stream-reasoner.ainf.at/reconf/ (which is very similar to the branch "streamreasoning" in this repo) is not working. Is there a working implementation of the paper: "Managing caching strategies for stream reasoning with reinforcement learning" ? Is there a way to execute Wasp for stream reasoning?

Thank you.

Berkmin with cache

There is a new class BerkminHeuristicWithCache. Can you introduce a command.line option and run the experiment?