[Clp] number of columns in CLP output do not match with mps file

[John J Forrest]

After a quick look at CoinMpsIO, I can't see anywhere that it will skip
columns - even if all the elements for a column are zero.

I would suggest solving problem e.g. -solve -printing all -solution mx2.sol
and look at mx2.sol to see where first mismatch occurs.  The first line of
column section should be
0 C0000001 (or maybe C0000000)
and do a binary search to see where it first goes wrong.  Then look at you
input file round that column.

If you have no luck then compress the file and send it to me.

John Forrest


                                                                                                                              
  From:       Mehmet Ferhat Candas <mfcandas at gmail.com>                                                                       
                                                                                                                              
  To:         clp at list.coin-or.org                                                                                            
                                                                                                                              
  Date:       01/15/2010 05:34 PM                                                                                             
                                                                                                                              
  Subject:    [Clp] number of columns in CLP output do not match with mps file                                                
                                                                                                                              
  Sent by:    clp-bounces at list.coin-or.org                                                                                    
                                                                                                                              





Hi,

We are having difficulty to match the CLP variable results with the
original variable names. CLP reports smaller number of columns at the
beginning of the run (in readMPS process) than what we have in the MPS
file. Such as:

Clp output:

Clp:import mpmod_mx2.mps
At line 1 NAME          BLANK
At line 2 ROWS
At line 282052 COLUMNS
At line 2124705 RHS
At line 2127427 BOUNDS
At line 2134118 ENDATA
Problem BLANK has 282048 rows, 649486 columns and 2904831 elements
Model was imported from .\mpmod_mx2.mps in 12.031 seconds



Mps file (having 726804 columns):
...
    C0726797  R0014563  1.             R0023421  1.
    C0726798  R0014564  1.             R0023422  1.
    C0726799  R0014565  1.             R0023915  1.
    C0726800  R0014566  1.             R0023916  1.
    C0726801  R0014567  1.             R0023917  1.
    C0726802  R0282045  1.
    C0726803  R0282046  1.
    C0726804  R0282047  1.

Need to match the original variable names with the results. So what do you
think causing this difference? Is this legitimate? If yes, what should we
do?
_______________________________________________
Clp mailing list
Clp at list.coin-or.org
http://list.coin-or.org/mailman/listinfo/clp

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/clp/attachments/20100118/45afff83/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: graycol.gif
Type: image/gif
Size: 105 bytes
Desc: not available
URL: <http://list.coin-or.org/pipermail/clp/attachments/20100118/45afff83/attachment.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ecblank.gif
Type: image/gif
Size: 45 bytes
Desc: not available
URL: <http://list.coin-or.org/pipermail/clp/attachments/20100118/45afff83/attachment-0001.gif>