QAP Instance Generator, Constructive and Local Search heuristics, Tabu Search Metaheuristics and Analysis tools
Tested in Julia 1.6.0, should work in the near and long term (Julia 1.6 is a LTS version)
In order to install the required dependencies specified by Project.toml.
One recommendation is to create a startup.jl file in ~\.julia\config and add the following lines
if isfile("Project.toml") && isfile("Manifest.toml")
Pkg.activate(".")
end
this will ensure that Julia will search and install the project automatically every time.
If this is not performed, then the Pkg mode must be accesed by entering ] inside Julia and every dependence manually added
$julia
julia> ]
(@v1.6) pkg> add Distances, ArgParse, Plots, DataFrames, StatsPlots, CSV
In order to execute the PIA in a single file(includes tools for analysis and experimentation), one must run
$julia pia.jl
using the flag of --help:
usage: pia.jl [-h] batchSize instanceSize
positional arguments:
batchSize Number of instances in batch (type: Int64)
instanceSize Size of instances of batch: S, M or L
optional arguments:
-h, --help show this help message and exit
There is also a version of the PIA without TS, in the file accordingly named.
In order to execute the PIA sequentally, one must first run the instance generator
$julia generator.jl
using the flag of --help:
usage: generator.jl [-v] [-h] nLocations nInstances
positional arguments:
nLocations Number of locations of the instance (type: Int64,
default: 0)
nInstances Number of instances (type: Int64, default: 0)
optional arguments:
-v, --verbose Specify verbose output
-h, --help show this help message and exit
Afterwards, in order to run the heuristic constructive, the next must be entered and executed
$julia constructive.jl
using the flag of --help:
usage: constructive.jl [-d] [-v] [-s] [-h] path
positional arguments:
path Path to file or directory
optional arguments:
-d, --dir Specify if a directory is to be read
-v, --verbose Specify verbose output
-s, --save Save solutions to files
-h, --help show this help message and exit
To execute the local search heuristic, the next must be entered and executed
$julia localSearch.jl
using the flag of --help:
usage: localSearch.jl [-d] [-v] [-s] [-h] path
positional arguments:
path Path to file or directory OF SOLUTIONS
optional arguments:
-d, --dir Specify if a directory is to be read
-v, --verbose Specify verbose output
-s, --save Save solutions to files
-h, --help show this help message and exit
Finally, to run the Tabu search metaheuristic, the next must be entered and executed
$julia tabuSearch.jl
using the flag of --help:
usage: tabuSearch.jl [-d] [-v] [-s] [-h] path iters
positional arguments:
path Path to file or directory OF SOLUTIONS
iters Number of iterations before stopping
optional arguments:
-d, --dir Specify if a directory is to be read
-v, --verbose Specify verbose output
-s, --save Save solutions to files
-h, --help show this help message and exit
This code is available online on an online repository. In this repository a Jupyter Lab will be prepared so anyone can access and run the project without having to install Julia on their personal computers. It will take a few days (As of May 23rd 2021) but it should be up and running relatively soon. The simulated annealing metaheuristic and long term memory component of TS are also expected to be completed on the coming months.