[Coin-discuss] Non-integer result from an integer problem

[John J Forrest]

Could be an error in the mps reader.  Won't be able to look at it until
Friday.

1.0e30 might be magically bad number - does it go away with UI 1.0e6 for
instance.

John Forrest


                                                                           
             Esben Mose Hansen                                             
             <esben at ange.dk>                                               
             Sent by:                                                   To 
             coin-discuss-boun         coin-discuss at list.coin-or.org       
             ces at list.coin-or.                                          cc 
             org                                                           
                                                                   Subject 
                                       [Coin-discuss] Non-integer result   
             28.11.2006 05:56          from an integer problem             
                                                                           
                                                                           
             Please respond to                                             
             Discussions about                                             
                open source                                                
               software for                                                
                Operations                                                 
                 Research                                                  
             <coin-discuss at lis                                             
              t.coin-or.org>                                               
                                                                           
                                                                           




Hi,

I seem to have encountered a bug in coin-or. The attached problem (in which

all variables are UI, i.e. integers) gives a non-integer result for a few
variables, (8--13). As a sanity check, I ran the same problem through glpk,

which returns a integer solution:

     8 C0000007     *              0             0         1e+30
     9 C0000008     *              1             0         1e+30
    10 C0000009     *              0             0         1e+30
    11 C0000010     *              0             0         1e+30
    12 C0000011     *              1             0         1e+30
    13 C0000012     *              1             0         1e+30

below is a an except from a cbc session. The model is attached.

Coin Cbc and Clp Solver version 1.01.00, build Nov  2 2006
CoinSolver takes input from arguments ( - switches to stdin)
Enter ? for list of commands or help
Coin:import coin_oct_dosave.mps
At line 1 NAME          BLANK
At line 2 ROWS
At line 427 COLUMNS
At line 1098 RHS
At line 1220 BOUNDS
At line 1415 ENDATA
Problem BLANK has 423 rows, 194 columns and 1172 elements
Model was imported from ./coin_oct_dosave.mps in 0.004 seconds
Coin:solve
-- snip --
Result - Finished objective 4156.21 after 0 nodes and 10 iterations - took
0.052003 seconds
Coin:solut $
--snip --
      8 C0000008     0.57142857          5.5542483e-17
      9 C0000009     0.57142857          5.5542483e-17
     10 C0000010     0.57142857          5.5542483e-17
     11 C0000011            0.5          5.5542483e-17
     13 C0000013              1                      0

-- snip --

--
kind regards, Esben
[attachment "coin_oct_dosave.mps" deleted by John J Forrest/Watson/IBM]
_______________________________________________
Coin-discuss mailing list
Coin-discuss at list.coin-or.org
http://list.coin-or.org/mailman/listinfo/coin-discuss