[Cbc] possible bug in displaying solution
Felix Noeske
felix.noeske at math.rwth-aachen.de
Fri Jul 10 04:04:45 EDT 2009
Dear List,
I am using cbc to solve some small-scale IPs which are generated by
a computer algebra system. The cbc executable (compiled from the svn
repository, Revision 1194) is run from within the CAS and the output is
parsed back in.
The problem in the example file I have attached is infeasible. Cbc
recognizes this, as can be seen if the log level is high enough. But
asking cbc for the optimal solution via "solution stdout" still gives an
optimal solution (probably from some preprocessing, I guess), rendering
a parsing of the output line useless.
Should it not say that the problem is infeasible? Or is there some way I
am missing to produce this effect.
Regards
Felix Noeske
--
Dr.rer.nat. Felix Noeske <Felix.Noeske at math.rwth-aachen.de>
Lehrstuhl D fuer Mathematik, RWTH Aachen, Templergraben 64, 52062 Aachen
Tel. +49 241 80 94547 / Fax +49 241 80 92108
"the worst is not,/
So long as we can say, 'This is the worst.'"
-"King Lear" IV.i
-------------- next part --------------
NAME CBC_PROBLEM
ROWS
N obj
L z1
L z2
L z3
L z4
L z5
L z6
L z7
L z8
L z9
L z10
L z11
L z12
L z13
L z14
L z15
L z16
L z17
L z18
L z19
L z20
L z21
L z22
L z23
L z24
L z25
L z26
L z27
L z28
L z29
L z30
L z31
L z32
L z33
L z34
L z35
L z36
L z37
L z38
L z39
L z40
L z41
L z42
L z43
L z44
L z45
L z46
L z47
L z48
L z49
L z50
L z51
L z52
L z53
L z54
L z55
L z56
L z57
L z58
L z59
L z60
L z61
L z62
L z63
L z64
L z65
L z66
L z67
L z68
L z69
L z70
L z71
L z72
L z73
L z74
L z75
L z76
L z77
L z78
L z79
L z80
L z81
L z82
L z83
L z84
L z85
L z86
L z87
L z88
L z89
L z90
L z91
L z92
L z93
L z94
COLUMNS
MARK0000 'MARKER' 'INTORG'
x1 z1 1 z3 1
x1 z4 -1 z30 -1
x1 z40 -1 z41 -1
x1 z42 -1 z45 1
x1 z46 -1 z47 -1
x1 z48 -1 z75 1
x1 z85 1 z86 1
x1 z87 1 z90 -1
x1 z91 1 z92 1
x1 z93 1
x2 z2 1 z3 1
x2 z4 -1 z33 -1
x2 z45 -1 z78 1
x2 z90 1
MARK0001 'MARKER' 'INTEND'
RHS
rhs z1 1 z2 1
rhs z3 1 z4 -1
rhs z75 1 z78 1
rhs z85 1 z86 1
rhs z87 1 z91 1
rhs z92 1 z93 1
BOUNDS
LI BND1 x1 0
LI BND1 x2 0
ENDATA
More information about the Cbc
mailing list