DIR TREE

• mcndp : contains mosel models and data for the MCNDP problem

  • cg : column generation approach

  • raw : raw approach

  • data : data file folder

    • easy : easy instances
    • hard : hard instances
    • small : contains a small instance of the MCNDP problem and its network representation to help understanding the .dat file format

  • lib : reusable code folder

• pdf : contains the assignment sheet, the report, the Mosel User Guide and papers on the subject of the MCNDP problem

• sp : contains mosel models and data for the SP problem

• test : mosel examples

  • data : data for the test models
  • mos : test models

• init : quick environment initialization for mosel

• run : shortcut for model execution

USAGE

INIT MOSEL

source ./init

RUN MODEL

./run <model_path> [params..]

EXAMPLES FOR mcndp/cg AND mcndp/raw

time ./run mcndp/cg  < mcndp/data/p33.dat

time ./run mcndp/raw < mcndp/data/p33.dat

BENCHMARK

./bench/run <inp> <out> <*alg>

./bench/run mcndp/data/easy/ bench/data/ mcndp/raw mcndp/cg

BENCHMARK LATEX TABLES

./bench/b2l <inp> <precision> <alg1> <alg2>

./bench/b2l bench/data/ 2 mcndp/raw mcndp/cg

DATA

Data is contained in .dat files inside the data directory (for MCNDP see mcndp/data). Information is splitted in 3 parts:

1. Info on the instance size;
2. Info on source-destination pairs;
3. Info on product supply/demand.

Info on the instance size

The first line contains all the info on the instance size. You can read the 3 following values:

• N = number of nodes;
• M = number of edges;
• K = number of products.

Info on source-destination pairs

Starting from the second line you can read info on source-destination pairs. For each of the M couples we have a first line with values.

•  *j* = destination;
  

•  *i* = source;
  

• f_{ij} = fixed costs;
• u_{ij} = total capacity;

•  *k* = number of products. 
  

After that we have a line for each of the k products with values

•    *k* = product index; 
  

• c^k_{ij} = cost on edge (i,j) for product k;
• b^k_{ij} = bound on edge (i,j) for product k.

Info on product supply/demand

Finally there is the section on product supply and demand. For each product we have a line containing values:

• *k* = product index;
  

• *i* = destination node index;
  

• d^k_i = amount of product k supplied (positive value) or demanded (negative value) by node i.

To have a better understanding see example mcndp/data/small/small.dat and mcndp/data/small/small.pdf.

/!\ Error in mcndp/data/small/small.pdf, u and b swapped.