Traveling Umpire Problem Solver
Overview
This program is based on the paper Solving the Travelling Umpire Problem with Answer Set Programming . The goal is to solve instances of traveling umpire problem (TUP), the definition of TUP can be viewed here . This program uses a Python script to generate source code files in Clingo language which are used together with the Clingo solver to find optimal solutions for the instances of TUP.
Dependencies
- Python version 3.8.3
- Clingo - easiest way to install is
conda install -c potassco clingo, otherwise see release
Running the Project
- Download TUP instance files and add them to the
/instancesdirectory - (Optional) Compile your own program in Clingo language
- Open
Anaconda Prompt - (Optional) Change configuration of experiments in
run.py - Run
python run.py [-h] [-t [TIME_OUT]] [-n [NUM_PROC]] [--src [SRC]] [--inst [INST]]script - View results in the
/outdirectory
Optional command line arguments for the TUP solver:
- -h, --help Show the help message and exit
- -t [TIME_OUT], --time-out [TIME_OUT] Time out for the Clingo solver specified in seconds,
default=300 - -n [NUM_PROC], --num-proc [NUM_PROC] Number of parallel processes of the Clingo solver,
default=1 - --src [SRC] Path to the directory that will contain generated '.lp' source files in Clingo language,
default='src_asp' - --inst [INST] Path to the directory containing TUP instances
default='instances'
If you want to run individual ASP programs with your own parameters you can use Clingo directly:
clingo [SRC] --opt-mode=optN --time-limit=[T] --parallel-mode=[NUM_PROC],[MODE] -q
If you want to view the models that are computed, don't run Clingo with the paremeter -q.
Be midful that the sheer volume of output is rather confusing. If you only want to view the
setting of specific atoms you need to modify the ASP program with #show A/n directive,
whereA is the name of the atom and n is its arity.