cleberoli / mdsp

The minimum distance superset problem: formulations and algorithms

The minimum distance superset problem

This repository contains the code used to recriate the experiments described on the paper The minimum distance superset problem: formulations and algorithms.

Models

Inside the model module we have the five models described in the paper:

• Quadratic formulation (QP)
• Reformulation-linearization technique (RLT)
• Integer formulation (IP)
• Feasible solution search (FEAS)
• Maximization search (MAX)

Instances

In the generator module we have the code responsible for generating the instances tested. According to the paper we divide these instances in four categories:

• Full instances (full)
• Missing distance instances (miss)
• Joint instances (joint)
• Random instances (drand)

Logs

In the logs folder, for each pair of instance type and model there is a folder containing the logs of each instance. For each instance there are four related files (.log, .lp, .sol, .time).

• The .log file has the information exported by the solver (gurobi) during the process
• The .lp file has the (latest) model of the problem for that instance
• The .sol file has the best solution found for the the model
• The .time file contains all the elapsed times for each optimzation, along with their status