Re: [eclipse-clp-users] small problem

From: Thorsten Winterer <thorsten_winterer_at_web.de>
Date: Wed, 16 Nov 2011 13:28:46 +0100
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  <mailto:ECLiPSe-CLP-Users_at_lists.sourceforge.net>
> https://lists.sourceforge.net/lists/listinfo/eclipse-clp-users
>
Received on Wed Nov 16 2011 - 12:28:54 CET

This archive was generated by hypermail 2.2.0 : Thu Feb 02 2012 - 02:31:58 CET