mclpif

Maximal Covering Location Problem with Interconnected facilities

Detailed results of our computational experiments for the manuscript "Continuous maximal covering location problems\with interconnected facilities" by Víctor Blanco y Ricardo Gázquez.

In particular:

• Instances: based on the 50-points data set in (Eilon, S., Watson-Gandy, C. D.~T., and Christofides, N. (1971). Distribution management : mathematical modelling and practical analysis. Folder "instances" in the form eilonN_i.csv where N is the number of demand points (N in {10, 20, 30, 40, 50} and i the instance (i in {1,2,3,4,5}).
• Results: Organized by graph structure. GraphType in {Cycle, Line, Complete, Matching, Ring, RingStar} in files GraphType_Results.csv with columns:
  • instance: file
  • GraphType
  • n: number of demand points
  • p: centers to be located
  • R: Coverage radius
  • r: limit radius between facilities
  • obj_NL: Objective value for the MINLP formulation.
  • Time_NL: CPU Time for the MINLP formulation.
  • Relax_NL: Root node relaxation of the MINLP formulation.
  • Gap_NL: MIPGap for the MINLP formulation within the time limit (3600 secs)
  • C_NL: Clusters of points obtained with the MINLP formulation.
  • obj_IP1: Objective value for the Incomplete formulation (INC1).
  • Time_IP1: CPU Time for solving the IP problem with the Incomplete formulation (INC1).
  • Relax_IP1: Root notde relaxation for the Incomplete formulation (INC1).
  • LC_IP1: # of Linear constraints for the Incomplete formulation (INC1).
  • numcuts_IP1: # of Cuts in the B-&-C skeme for the Incomplete formulation (INC1).
  • TConstrGen_IP1: CPU time required to generate constraints for the Incomplete formulation (INC1).
  • TotalT_IP1: Total CPU time for the Incomplete formulation (INC1) to solve the problem.
  • GapIP1: MIPGap for the Incomplete formulation (INC1) within the time limit (3600 secs).
  • C_IP1: Clusters of points obtained with formulation (INC1).
  • obj_IP2: Objective value for the Incomplete formulation (INC2).
  • Time_IP2: CPU Time for solving the IP problem with the Incomplete formulation (INC2).
  • Relax_IP2: Root notde relaxation for the Incomplete formulation (IN2).
  • LC_IP2: # of Linear constraints for the Incomplete formulation (INC2).
  • numcuts_IP2: # of Cuts in the B-&-C skeme for the Incomplete formulation (INC2).
  • TConstrGen_IP2: CPU time required to generate constraints for the Incomplete formulation (INC2).
  • TotalT_IP2: Total CPU time for the Incomplete formulation (INC1) to solve the problem.
  • GapIP2: MIPGap for the Incomplete formulation (INC2) within the time limit (3600 secs).
  • C_IP2: Clusters of points obtained with formulation (INC2).

We also report a file with the results obtained with the compact formulation (file CompactIP_Results.csv) with the following information:

• obj_IP: objective value.
• IP_Time: CPU time required to solve the IP problem.
• COntrGen_Time: CPU Time required to incorporate the constraints to the model.
• Tot_Time: Total CPU time (IP_Time + COnctr_Time)
• #Ball_Ctrs: Number of ball-type constraints.
• #O_Ctrs: Number of O-sets constraints.
• #All_Ctrs: Total number of constraints.
• Gap: MIP Gap for the IP formulation within the time limit (3600 secs)
• OoM: 1 if the problem got a Out Of Memory Flag in Gurobi; 0 otherwise.