Re: [eclipse-clp-users] small problem

Yes. That is the problem. Thank you!

 

From: Thorsten Winterer [mailto:thorsten_winterer_at_web.de] 
Sent: 16 noiembrie 2011 13:29
To: Bogdan Tanasa
Cc: eclipse-clp-users_at_lists.sourceforge.net
Subject: Re: [eclipse-clp-users] small problem

 

Hi,

I think I found the problem.

If you use the reified constraints, then you have a problem when the
automatically generated lists contain the same value twice. In that case,
this value cannot be chosen, because otherwise, two of the Z variables would
become 1, which is forbidden by sum(Z_List) #= 1. This may prevent the
version with the reified constraints from finding the best solution.

Cheers,
Thorsten


Am 16.11.2011 11:45, schrieb Bogdan Tanasa: 

Hi,

 

Thanks for investigating my problem. Try running the code for more cases. It
will happen for sure.

 

I am using Eclipse 6.0 on a windows xp 32bit machine.

 

From: Thorsten Winterer [mailto:thorsten_winterer_at_web.de] 
Sent: 16 noiembrie 2011 11:37
To: eclipse-clp-users_at_lists.sourceforge.net
Subject: Re: [eclipse-clp-users] small problem

 

Hi Bogdan,

can you give more information on the ECLiPSe version that you use?
I tested your program with 6.0#127, and I got identical solutions for both
variants:

Found a solution with cost 63.600314464631339__63.600314464631467
Found a solution with cost 59.30430001273087__59.304300012730991
Found a solution with cost 57.2013985843003__57.201398584300421
Found a solution with cost 49.919935897394716__49.919935897394772
Found a solution with cost 47.885279575251481__47.885279575251538
Found a solution with cost 45.254833995939016__45.254833995939073
Found a solution with cost 43.312815655415406__43.312815655415456
Found a solution with cost 40.951190458886522__40.951190458886572
Found a solution with cost 37.841775856848983__37.841775856849033
Found a solution with cost 34.9714169000914__34.971416900091448
Found a solution with cost 32.419130154894638__32.419130154894681
Found a solution with cost 29.189039038652826__29.189039038652858
Found a solution with cost 24.637369989509825__24.637369989509853
Found no solution with cost 0.0 .. 23.637369989509853
     * Bin Configurations: 
[0, 0, 0, 0, 1]
[0, 0, 0, 1, 0]
[0, 0, 1, 0, 0]
[0, 1, 0, 0, 0]
[0, 0, 1, 0, 0]
[1, 0, 0, 0, 0]
     * Bin Loads: 
[42, 23, 27, 27, 24]
     * Weigth Configurations: 
[14, 33, 24, 49] :: [0, 0, 1, 0]
[41, 49, 9, 27] :: [0, 0, 0, 1]
[6, 49, 10, 12] :: [0, 0, 0, 1]
[41, 9, 23, 45] :: [0, 0, 1, 0]
[49, 19, 15, 51] :: [0, 0, 1, 0]
[47, 12, 42, 47] :: [0, 0, 1, 0]
Found a solution with cost 61.644140029689716__61.644140029689837
Found a solution with cost 57.2013985843003__57.201398584300421
Found a solution with cost 49.919935897394716__49.919935897394772
Found a solution with cost 45.254833995939016__45.254833995939073
Found a solution with cost 43.312815655415406__43.312815655415456
Found a solution with cost 37.841775856848983__37.841775856849033
Found a solution with cost 34.9714169000914__34.971416900091448
Found a solution with cost 32.419130154894638__32.419130154894681
Found a solution with cost 24.637369989509825__24.637369989509853
Found no solution with cost 0.0 .. 23.637369989509853
     * Bin Configurations: 
[0, 0, 0, 0, 1]
[0, 0, 0, 1, 0]
[0, 0, 1, 0, 0]
[0, 1, 0, 0, 0]
[0, 0, 1, 0, 0]
[1, 0, 0, 0, 0]
     * Bin Loads: 
[42, 23, 27, 27, 24]
     * Weigth Configurations: 
[14, 33, 24, 49] :: [0, 0, 1, 0]
[41, 49, 9, 27] :: [0, 0, 0, 1]
[6, 49, 10, 12] :: [0, 0, 0, 1]
[41, 9, 23, 45] :: [0, 0, 1, 0]
[49, 19, 15, 51] :: [0, 0, 1, 0]
[47, 12, 42, 47] :: [0, 0, 1, 0]
Results: 
24.6373699895098__24.6373699895099
24.6373699895098__24.6373699895099


Cheers,
Thorsten


Am 16.11.2011 10:56, schrieb Bogdan Tanasa: 

Hi guys,

 

I have a small version of the bin covering problem which I try to solve.

I have to versions of it, the second one uses reified constraints.

 

It happens that the second one give different results compared with the
first implementation which does not uses reified constraints.

 

Can you please tell me what is wrong_

 

Bogdan.







----------------------------------------------------------------------------
--
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1







_______________________________________________
ECLiPSe-CLP-Users mailing list
ECLiPSe-CLP-Users_at_lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/eclipse-clp-users