AMPL Modeling Language Google Group

RE: [AMPL 1873] Running Benders Decomposition on Cluster

4...@ampl.com (Robert Fourer) — Thu, 24 Jul 2008 17:55:33 UT

For some examples of Benders Decomposition implemented through AMPL scripts,
see [link].
An example of how an AMPL script was set up to solve multiple subproblems in
parallel is given in
Elizabeth D. Dolan, Robert Fourer, Jean-Pierre Goux,
Todd S. Munson, and Jason Sarich, "Kestrel: An Interface

Re: [AMPL 1874] Modeling Non-Integer Discrete Variables in AMPL

demirel.suley...@gmail.com (Suleyman Demirel) — Wed, 23 Jul 2008 19:00:29 UT

Thanks a lot!

RE: [AMPL 1874] Modeling Non-Integer Discrete Variables in AMPL

4...@ampl.com (Robert Fourer) — Wed, 23 Jul 2008 18:02:41 UT

You could write
var x in {0, 0.1, 0.17, 0.29, 0.77, 1.02};
and let AMPL convert to an integer program. Or you could do the conversion
yourself:
set Sx; # possible values for x
var x;
var xv {Sx} binary; # x = i iff xv[i] = 1
subj to AtMostOneValueForX:
sum {i in Sx} xv[i] = 1;

Modeling Non-Integer Discrete Variables in AMPL

demirel.suley...@gmail.com (Suleyman Demirel) — Wed, 23 Jul 2008 15:59:39 UT

Hello Everybody,

Does anyone know how to model non-integer discrete variables in AMPL?
For example, x takes values in the set {0, 0.1, 0.17, 0.29, 0.77,
1.02}.

For a well behaved discrete set, e.g. {0, 0.25, 0.50, 0.75, 1}, this
is easy by letting y=4x and setting y as an integer variable. For

Running Benders Decomposition on Cluster

mr.ashutosh.sha...@gmail.com (sharmaa@hope.ac.uk) — Tue, 22 Jul 2008 11:31:26 UT

Anybody knows how to run Benders Decomposition on Cluster computer
under Linux or Windows ?

Thanks

Ash

Installation Problem

stuart.nettle...@optusnet.com.au (Stuart Nettleton) — Sun, 20 Jul 2008 12:13:43 UT

Hi, I am getting the following errors in Suse 11 when running make.
The BLAS package is installed but it doesn't seem to have lcblas
Any help in a workaround would be appreciated.
Thanks,
Stuart
~/drampl> make
gcc -std=c99 -Wall -Wstrict-prototypes -pthread -fPIC -Bdynamic
-fno-strict-aliasing -g -I/home/stuart/libampl/Src/sol vers

AMPL/CPLEX vs Java/CPLEX performance question

ronald.men...@jda.com (Ronald Menich) — Thu, 10 Jul 2008 20:15:53 UT

We have an AMPL/CPLEX prototype model using the substout option for
which the actual CPLEX solve executes in x minutes.

For a variety of reasons, we need to have a Java/CPLEX (Concert)
software implementation. We've created that and we get exactly the
same results between AMPL/CPLEX and the Java/CPLEX (Concert), out to

Re: Strange results from AMPL

mittelm...@asu.edu (Hans Mittelmann) — Thu, 10 Jul 2008 04:18:14 UT

The first problem was solved the second not really.

Re: [AMPL 1867] fix command

olin...@engr.smu.edu (Eli Olinick) — Tue, 08 Jul 2008 20:09:42 UT

The fix command can be indexed like a variable declaration or constraint.
So, you can use a command such as the following:
fix {v in V} x[118,40,v] := 0;
Eli
------------------------------ ------------------
Eli V. Olinick
Associate Professor
Department of Engineering Management, Information, and Systems

fix command

jabra...@gmail.com (Mushi) — Tue, 08 Jul 2008 20:03:34 UT

I have been using the fix command in my data file to set certain
binary variables to zero. For example
fix x[118,40,1]:= 0;
and suppose var x{i in N, j in N, v in V}binary; (v are vehicles)

I would like to be able to put something like
fix x[118,40,*]:= 0;
so that I could say for any vehicle, 118-40 is zero, without listing

Re: Strange results from AMPL

erichuan...@gmail.com (Eric) — Tue, 08 Jul 2008 20:00:54 UT

Thank you Paul for your suggestions. I will try and see.

Re: Strange results from AMPL

ru...@msu.edu (Paul) — Tue, 08 Jul 2008 14:09:44 UT

Eric wrote:

In my experience, this sort of anomaly usually traces back to a
mixture of poor scaling in the original problem and/or rounding errors
and/or the particular tolerances being used. The fact that you only
needed one b-&-b node after you added the extra bounding constraint
makes me wonder if either of the presolvers is a factor here. Try

Strange results from AMPL

erichuan...@gmail.com (Eric) — Mon, 07 Jul 2008 23:28:49 UT

Hi there,

I solved a mixed integer problem, say, A (minimization problem), I got
following results:

CPLEX 11.0.0: optimal integer solution; objective 15043258.48
324295 MIP simplex iterations
34 branch-and-bound nodes
1 cover cut
412 flow-cover cuts
1 Gomory cut
7314 implied-bound cuts

Re: hooking minos to ampl

mittelm...@asu.edu (Hans Mittelmann) — Fri, 04 Jul 2008 20:45:32 UT

[link]

hooking minos to ampl

chrisw...@gmail.com (Chris Wu) — Thu, 03 Jul 2008 15:20:45 UT

hi,
Recently I bought a copy of my own minos 5.5. There is AMPL installed on one
server I have access to.
Now I would like to install minos on this server under my home directory and
link it to AMPL so that I can use minos solver inside AMPL. Could anybody
tell me how to do this? Many thanks!
by the way, my linux system information is