Error in CglRedSplit::generateCuts(): rstat[557]: 0 and wrong result #105
Comments
|
It looks as if the problem is that the solver is passed in to CglRedSplit as const, but the code casts away the const. It looks as if the code returns early because of a problem e.g. rstat[557] and not everything is reset. The next generator gets confused and gives a bad cut. I could modify master, but maybe someone else can come up with a better solution. So try modifying code as
|
|
I tried that, leads for me to even more ERROR: CglRedSplit::generateCuts(): rstat[557]: 0but the result was then ok as 8315709965526. |
|
|
Probably a mistake to say error (ifndef NDEBUG). It seems to mean that the code thinks that it is a free row and it does not know what to do - so returns leaving model in a mess. It is not a free row ? it is a cut with a lower rhs of -2.0e15 and upper of infinity! I found out that that cut was being generated in CglProbing so will stop that happening in master. |
|
I tried it with the latest master today, there it looks already better, thanks a lot: Unfortunately another of my solver regression tests then runs into an error with that revision. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
With
value_8315709965526.lp.txt
and the driver options
I run into
The text was updated successfully, but these errors were encountered: