[Clp] CLP solve a linear programming model with 35, 000 dec. variables and 100, 000 constraints.

usa usa usact2012 at gmail.com
Mon May 9 17:51:48 EDT 2016


Dear Julian,

In  O(m^2), m is max (row, column) ?

Thanks






On Mon, May 9, 2016 at 5:40 PM, OR MSc <ormsc at ed.ac.uk> wrote:

> Dear Usa
>
> What if I find a machine with memory large enough to hold the matrix ?
>>
>> Are there some internal limits about the matrix size inside CLP
>> implementation ?
>>
>
> The only limit I can think could be reached occurs when the number of
> nonzeros reaches what can be stored in a 4-byte integer
>
> Especially, the limits about matrix inversion or other matrix operations ?
>>
>
> No: so long as there is enough memory, Clp will allocate O(m^2) memory to
> store the LU factors of the basis.
>
> Marginally, the number of updates which the simplex method will perform
> may be limited by  the available memory, increasing the frequency with
> which refactorization takes place but decreasing the iteration speed.
>
> With a very dense problem the sparsity-exploiting numerical linear algebra
> may be
>
> Of course, I didn't write Clp, but I've written a comparable system so
> know of the issues involved.
>
> Julian
>
> PS Sorry if you've already received this reply!
>
>
> On Mon, May 9, 2016 at 4:51 AM, Julian Hall <jajhall at ed.ac.uk
>> <mailto:jajhall at ed.ac.uk>> wrote:
>>
>>     Dear Usa
>>
>>     Not a hope. With no sparsity to exploit, your problem has 3.5G
>>     values, each needing 8B of memory. To solve the problem with Clp
>>     probably needs about 80G memory.
>>
>>     Both your problem dimensions are at least half an order of magnitude
>>     too large.
>>
>>     Julian
>>
>>
>>     On 9 May 2016 04:56:45 BST, usa usa <usact2012 at gmail.com
>>     <mailto:usact2012 at gmail.com>> wrote:
>>
>>         Hi,
>>
>>         I would like to solve a linear programming model by CLP on a
>>         laptop with 8 GB memory.
>>
>>         It has 35,000 dec. variables and 100,000 constraints.
>>
>>         The constraint matrix is very dense.
>>
>>         Is it possible to solve this size of LP model by CLP ?
>>
>>         Any advices would be appreciated.
>>
>>         thanks !
>>
>>
>> ------------------------------------------------------------------------
>>
>>         Clp mailing list
>>         Clp at list.coin-or.org  <mailto:Clp at list.coin-or.org>
>>         http://list.coin-or.org/mailman/listinfo/clp
>>
>>
>>     --
>>     Sent from my Android device with K-9 Mail. Please excuse my brevity.
>>
>>     The University of Edinburgh is a charitable body, registered in
>>     Scotland, with registration number SC005336.
>>
>>
>>
> --
> Julian Hall (Postgraduate Taught Programme Coordinator - Operational
> Research MSc Programme Director)
> --
> Dr. J. A. Julian Hall, Senior Lecturer, School of Mathematics,
> University of Edinburgh, James Clerk Maxwell Building,
> Peter Guthrie Tait Road, EDINBURGH, EH9 3FD, UK.
> Room: 6221   Phone: [+44](131) 650 5075   Email: J.A.J.Hall at ed.ac.uk
> Web: http://www.maths.ed.ac.uk/people/show/person/47
>
>
> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/clp/attachments/20160509/3731e335/attachment.html>


More information about the Clp mailing list