Dear all,
this is a message about the forthcoming 2024 XCSP3 competition, with 3 main points:
1) Competition Call and Deadline
We are happy to announce that
the 2024 XCSP3 competition will be conducted during the first semester
of 2024 (with results envisioned to be published during CP'24).
The most important details are:
- as usual, 4 tracks related top COP (Optimization) and 2 tracks related to CSP (satisfaction)
- Deadline for participating: 30 May 2024
- around 24 classical constraints (XCSP3-core) ; modifications wrt 2023 are described in the call and are quite minor
You have the possibility of submitting benchmarks too, using for example the library PyCSP3 (see
pycsp.org).
2) About the Deadline
If you think that it may be difficult to submit before the specified deadline (30 May),
please contact us (but, of course, we offer no guarantee).
3) Checking Solutions
Here is a link for the tool SolutionChecker :
You can check solutions and bounds (in case of optimisation problems). For example:
java ace Drinking-101.xml -npc
outputs:
v <instantiation id='sol1' type='solution' cost='10'>
<list> x[] y[] </list> <values> 0x5 1 0x8 1 0x10 1 0x9
1 0x9 1 0x9 1 0x9 1 0x9 1 0x8 1 0x10 1 0x5 * 0x5 1x8 0 1x8 0x3 1x8 0 0
1x8 0 0 1x8 0 0 1x8 0 0 1x8 0 0 1x8 0 1x8 0x3 1x5 </values>
</instantiation>
Then, you can run:
java -jar xcsp3-solutionChecker-2.5.jar Drinking-101.xml
"<instantiation id='sol1' type='solution' cost='10'> <list>
x[] y[] </list> <values> 0x5 1 0x8 1 0x10 1 0x9 1 0x9 1 0x9 1
0x9 1 0x9 1 0x8 1 0x10 1 0x5 * 0x5 1x8 0 1x8 0x3 1x8 0 0 1x8 0 0 1x8 0 0
1x8 0 0 1x8 0 0 1x8 0 1x8 0x3 1x5 </values>
</instantiation>" -b=10
and the tool indicates that everything is ok
If '-b=10' is replaced by '-b=11' you get an error.
With
the option -cm (competition mode) it can even be automatic when using
directly the output of the solver. For example, after executing:
java ace Drinking-101.xml -npc > out
we can run:
java -jar xcsp3-solutionChecker-2.5.jar Drinking-101.xml out -cm
so that the last value in a 'o ...' line is compared with bound computed for the lsats solution.
Best,
Gilles, Emmanuel and Christophe