[Help-glpk] Parsing large, sparse mathprog model takes a (very) long time

[Yaron Kretchmer]

Hi All.

I have a mathprog model which takes a long time to parse:

The model is of a set-to-set routing problem- There is a bipartite graph, and the objective is to establish connections between the two parts such that a length metric is minimized.

There are ~76K nodes in each "half" of the graph (for a total of ~150K nodes), and a total of 500K edges between the nodes.

 

The model has been parsing (i.e. haven't reached presolving stage yet) for more than an hour now.  Memory consumption is increasing, but thus far is quite low (<200M).

 

Is this long parse time reasonable? Are there ways of rewriting which would speed up parsing?

 

Since the model is quite large, I put it on http://foodproj.org/set_to_set_routing.zip

 

 

Thanks and Best Regards

Yaron