[Coin-lpsolver] problems with BInvCol and BInvRow

[John J Forrest]

Clp stores slacks as -1.0 (i.e. not really slack but row activity).  There 
is a comment about this in OsiClpsolverInterface but something is 
obviously wrong.  CglGomory.cpp uses CoinFactorization directly so does 
not have this problem.

I will try and look at it this weekend.

John Forrest



Anureet Saxena <anureet at yahoo.com> 
Sent by: coin-lpsolver-bounces at list.coin-or.org
09/15/2005 05:34 PM

To
coin-lpsolver at list.coin-or.org
cc

Subject
[Coin-lpsolver] problems with BInvCol and BInvRow






hi,

There seems to be some trouble with the 
OsiSimplexInterface::getBInvCol() and
OsiSimplexInterface::getBInvRow() functions.

In particular, constructing the basis inverse via these two functions 
gives
different results (after transposing). Furthermore, the problem is 
specifically
in those rows whose corresponding basic variable is a slack variable. 

A small example (3 rows and 5 variables) and relevent code can be 
downloaded
from the following link:

http://www.andrew.cmu.edu/user/anureets/tmp/
File Name: BInvCol_case.tgz

It will be very nice if someone could point out and/or rectify the error.
 
On a different note, OsiSimplexInterface has a function:
 
enableSimplexInterface(bool doingPrimal).
 
The role of the parameter "doingPrimal" is not adequately described in the 
doxygen documentation. Could someone please give some idea as to what this 
parameter accomplishes. 


regards,
Anureet Saxena
ACO Doctoral Student,
Tepper School of Business,
Carnegie Mellon University.




Anureet Saxena
anureet at cmu.edu
anureet at yahoo.com

--This mail was delivered to you on 100% recyclable electrons.
Yahoo! for Good
Click here to donate to the Hurricane Katrina relief effort. 
_______________________________________________
Coin-lpsolver mailing list
Coin-lpsolver at list.coin-or.org
http://list.coin-or.org/mailman/listinfo/coin-lpsolver

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/clp/attachments/20050916/54dab27b/attachment.html>