John Forrest

Hopefully fixed. Reduce-and-split gets confused on odd row types. Clp thought a cut (from another cut generator was odd) and confused reduce-and-split. A lot of the cuts were "interesting" e.g....

I am unable to reproduce the error. Your log says - Cgl0003I 0 fixed, 148 tightened bounds, 13 strengthened rows, 0 substitutions Mine says - Cgl0003I 0 fixed, 152 tightened...

As parallel cbc is normally non-deterministic, this does not matter bug-wise as it just means cuts may or may not get generated (but they will be valid ones). But it...

If there are not many duplicate columns then checking for them can be switched off in preprocessing. If you are using stand-alone cbc (or calling CbcMain as in some of...

Not really at home in CoinLpIO - not as easy to clean up messages as I would like. A quick hack gives - on a messed up problem. command line...

On 05/09/2022 10:18, Jo Devriendt wrote: > > Over at Exact we're using CoinUtils as parser for .mps and .lp files. > This works most of the time, but for...

On 05/09/2022 14:37, Jo Devriendt wrote: > > The ideal case would be to have these "indicators" in fitting data > structures, but I can imagine this breaks a bit...

Not wishing to disturb the code too much, I think the best thing to do is leave maximum at 8 characters when adding a row/column without a name. I will...

Put change into master. The names are not elegant - I use a - z and pad out with 0.

The file is not in the correct format, but the error message is wrong. The correct format for the Cplex extension to MPS format has MIN or MAX on the...