This project contains all source files and tests for the essay "Solving Facility Location Problems and Service Modularization Problems using Greedy Adaptive Randomized Search Procedures" for the 19th International Working Seminar on Production Economics (2016)
by Hagen Salewski

The folder "algorithm" holds all source codes to compile a couple of GRASP-based applications which can solve the Simulaneous Multi-Path Service Modularisation Problem (SMP)²

To compile, you need to have version 2.0.1 of ParadisEO (http://paradiseo.gforge.inria.fr/index.php?n=Doc.Install) and all it requirements installed. Then:
- mkdir algorithm/build
- cd algorithm/build
- cmake ..
- make

After the compilation all applications can be found in "algorithms/build/applications"

There are some test applications and some productive applications. For the tests only the application SMP2_TimeToTarget is used, it takes the following command line arguments:
	test problem instance 	(-p="TakeProblemFromHere.dat")
	random seed		(-s=12345)
	result file 		(-o="PutResultsInHere.csv")
	neighborhood definition	(-n=1 for Adjecent Element Flip, -n=2 for Element Flip)
	local search algorithm  (-l=1 for Best Improvement, -l=2 for First Random Better Improvement)
	alpha value		(-a=0.1 or any other number in [0,1] or -a=R for Reactive GRASP od -a=E for random Alpha)
	target value		(-t=44.52)
	number of T2T-trials	(-I=10000)
	max runtime in seconds	(-T=3600 for running for 1 hour and set best result as target value - only important if no target value is proviede)


The folder "tests" is self contained and holds all tests, compiled SMP2_TimeToTarget, R-Scripts and Bash-Scrips required to run the tests. In each test-subfolder you can run the tests by executing the Bash script "Run.sh":
./Run.sh


For question please contact the author:
Dr. Hagen Salewski
Department of Business Administration and Production Management
University of Kaiserslautern
Gottlieb-Daimler-Str. 42
67663 Kaiserslautern
Germany

Email: salewski@wiwi.uni-kl.de