[Cbc] Bug in integer infeasible solution reporting

[John J Forrest]

Cbc does report the problem as infeasible but I agree that the solution 
file does not.  The problem had been fixed in trunk.  I have copied over 
the modifications to stable/2.3

John Forrest





[Cbc] Bug in integer infeasible solution reporting

Stuart Mitchell 
to:
cbc
09/25/2009 12:12 AM


Sent by:
cbc-bounces at list.coin-or.org







This may have something to do with the infeasibility being found in 
presolve

but debug.lp
generates
debug.sol
which states it is an optimal solution even though the variables are non 
integer

Stu

-- 
___________________________________
Dr Stuart Mitchell
Research Fellow
Light Metals Research Centre (LMRC)
University of Auckland
Private Bag 92019
Auckland
New Zealand

Ph (wk)  +64 9 3737599 ext 84867
    (ddi) +64 9 9234967
    (fax) +64 9 3737925
    (mb)  +64 21 441331
___________________________________

\* test6 *\
Minimize
OBJ: __dummy
Subject To
c1: x + y <= 5.2
c2: x + z >= 10.3
c3: - y + z = 7.4
Bounds
__dummy = 0
0 <= x <= 4
0 <= z <= 10
-1 <= y <= 1
Generals
x
z
y
End
Optimal - objective value               0
      0 __dummy               0                      1
      1 x                   1.9                      0
      2 y                     1                      0
      3 z                   8.4                      0
[attachment "s_mitchell.vcf" deleted by John J Forrest/Watson/IBM] 
_______________________________________________
Cbc mailing list
Cbc at list.coin-or.org
http://list.coin-or.org/mailman/listinfo/cbc

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://list.coin-or.org/pipermail/cbc/attachments/20090925/46033551/attachment.html