Xpress-MP Google Group

The Xpress community has moved

chris.oak...@dashoptimization.com (ChrisOakley) — Tue, 06 June 2009 14:15:13 UT

This Xpress community forum has been relocated to the FICO Decision
Management Community.

We hope that you will join us there to post questions, comments, ideas
and engage in discussion related to Xpress, Optimization and the wider
area of Decision Management.

Stochastic Programming Module

eejriets...@gmail.com (Eelco) — Sun, 06 June 2009 17:07:41 UT

Hi all,

I have been using the "mmsp" module for solving a multistage recourse
model. I have been searching on the website to find the underlying
algorithm(s) that are used by this module. Can anyone help?
I believe that the recourse model is reset as a large scale LP
equivalent en solved accordingly. Is this correct?

Re: XpressMP on Leopard

vimek...@gmail.com (vimek vanamali) — Thu, 06 June 2009 19:59:20 UT

install and keep another OS just to run Xpress"
yes i agree with u 100%..But...,,Try Vmware in linxu so that u can switch
into WIN/Linux at the same time....To use Vmware u will have to enable the
VT option in your motherboard.

Re: XpressMP on Leopard

dacat...@ulb.ac.be (Daniele Catanzaro) — Wed, 06 June 2009 14:40:02 UT

Dear Chris and Vimek

Thanks for yours answers. I confess it's a bit uncomfortable to
install and keep another OS just to run Xpress, but since I love
Xpress and
since it is much faster than Cplex 11.0 for Mac OS X, I have no choice.
All the best,

Daniele

--
Dr. Daniele Catanzaro,

Re: XpressMP on Leopard

vimek...@gmail.com (vimek vanamali) — Wed, 06 June 2009 14:08:33 UT

USE WINE TO INSTALL WIN APPLICATION INTO LINUX

Re: XpressMP on Leopard

chris.oak...@dashoptimization.com (ChrisOakley) — Wed, 06 June 2009 13:29:24 UT

There are no plans to port Xpress to the Mac, although we would
certainly consider it if there were enough commercial interest.

Chris.

E-1013 Infeasible constraint

sfonne...@gmail.com (Sebastian) — Tue, 05 May 2009 17:39:55 UT

I'm getting the error E-1013 Infeasible constraint in my model with
code on line 78: "sum(l in T)M(i,l,1)*l+sum(j in E,l in T)P(i,j,l,1)
*l>=Es(i)" How can I solve this?

This is the code for my model=

model Carpool
uses "mmxprs"; !gain access to the Xpress-Optimizer solver

!optional parameters section

XPRS_TOP and XPRS_LIN

anaramo...@gmail.com (Ana Ramos) — Sun, 05 May 2009 17:08:28 UT

Hello

What is the difference between XPRS_TOP and XPRS_LIN?

Thanks

XpressMP on Leopard

dacat...@ulb.ac.be (Daniele Catanzaro) — Fri, 05 May 2009 14:07:37 UT

Dear Madam, Sir

may I ask you whether there is any chance to see XpressMP running on
Mac Os X in the next months?

I do not refer explicitly about the porting of Xpress-Ive (which
however make the development really easy), what is
really important to me is the possibility to use XpressMP on mac os x

Converting a string to numeric?

zuhair.abdu...@gmail.com (Zee) — Mon, 05 May 2009 18:32:33 UT

Hi guys

I need some help. I have a file with both numeric and character data.
and i imported these to a 2d array(of type string) in mosel.

Now, this is what I wanna try...

Can I access a column from this array and then extract the numeric
value (in effect i wanna convert the string type info to numeric data

Re: PLZZZZZZZZZ HELP

ov.mig...@googlemail.com (Miguel O-V) — Mon, 05 May 2009 17:23:06 UT

I am not sure if I did understand the question... but if what you want
is to load an array from a dat file you can do it by means of for
loops as:

forall(a in 1..A,b in 1..B) do
read(D(a,b))
end-do

You do not need to strictly start from 1.

Re: xPress Mosel IVE!

m.margreit...@gmx.at (Mags) — Tue, 04 Apr 2009 08:24:14 UT

Hey!

Please, if somebody knows anything to this problem, help me!

Best regards,

Mags

Interpretation of the Solution

peterschr...@googlemail.com (peterschrube@googlemail.com) — Wed, 04 Apr 2009 17:01:02 UT

Hi,

I have used Xpress in my dissertation and must interpret my results
properly. In the book as well as in the net I could find nothing the
fitting, for this reason the last attempt is to write here.

Please, explain me how to interpret the Rows (constrains), Columns
(variable), Non-zero elements as well as Simplex iterationens.

Interpretation of the Solution

peterschr...@googlemail.com (peterschrube@googlemail.com) — Wed, 04 Apr 2009 16:59:59 UT

Hi,

I have used Xpress in my dissertation and must interpret my results
properly. In the book as well as in the net I could find nothing the
fitting, for this reason the last attempt is to write here.

Please, explain me how to interpret the Rows (constrains), Columns
(variable), Non-zero elements as well as Simplex iterationens.

RE: Optimization Libraries Student Version

chrisoak...@fairisaac.com (Oakley, Chris) — Wed, 04 Apr 2009 10:20:34 UT

The library APIs are currently only available through the Academic Partner Program which grants full licenses of the software to degree awarding institutes that teach at least one course with Xpress.

On Windows, some access to the solver API is provided as part of the Microsoft Solver Foundation v1.2 which will be released this week.

xPress Mosel IVE!

m.margreit...@gmx.at (Mags) — Tue, 04 Apr 2009 13:00:24 UT

Hello!

I'm a greenhorn in programming and modelling and I hope somebody could
help me!

I have the following problem with xPress Mosel IVE:

I have a linear program with the following side condition: (Transport
condition)

y1l >= LB1 or y1l = 0

l ... stands for distribution center or consumer

Re: Optimization Libraries Student Version

deneve.dil...@gmail.com (laradeneve) — Sat, 04 Apr 2009 08:26:25 UT

Hi,

I was going to ask the same thing for Windows.

Lara.

Re: How to know when the B&B has found an Integer solution

joaop...@gmail.com (jpita) — Mon, 04 Apr 2009 14:05:04 UT

Hello!

I'm a PhD student and I don't how to see if there are alternative
solutions to a given integer solution.

I mean...a run the optimizer it gives a couple of solutions but maybe
there are some equal solutions and I can't check it. Do you know how
to do it?

Thanks!

Optimization Libraries Student Version

balan...@googlemail.com (Maximilian Balandat) — Thu, 04 Apr 2009 20:43:51 UT

Hi there,

I was wondering if it is possible to obtain the Xpress Optimization
library for UNIX operating systems as a student version? I just need
the optimizer (and a mex interface), no fancy IDE.

Thanks,

Max

Xpress MP student edition can do sensitivity analysis?

firmin...@gmail.com (Andre Soares) — Tue, 03 Mar 2009 15:36:50 UT

Good afternoon
I have a doubt about the student version of xpress MP. I want to know
if the student version of xpress MP can do sensitivity analysis?
Best regards
André Soares

sensitivty analysis on the student edtion

firmin...@gmail.com (firmino82@gmail.com) — Mon, 03 Mar 2009 17:14:00 UT

Good afternoon

I have a doubt about the student version of xpress MP. I want to know
if the student version of xpress MP can do sensitivity analysis?

Best regards
André Soares

Information about ucbdata

anaramo...@gmail.com (Ana Ramos) — Tue, 03 Mar 2009 11:05:07 UT

Dear all,
We are trying to lead the branching variable in the B&B performed by
XPRESS.
To this end, we have found the callback function XPRS_CB_CHGBRANCH
that in
principle allows to choose the new branching var. by means of the
ucbdata
procedure.

Our question is about the parameter of this function. In principle,

Re: LP gone horribly wrong

eejriets...@gmail.com (Eelco Rietsema) — Mon, 03 Mar 2009 07:31:05 UT

Hi Patrizia,

I think the studentversion of Xpress could deal with this size, but I am
unsure. Please visit dashoptimization.com to find the dimensions of the
student version.

Furthermore, it is unclear how many products can be produced by each
factory. Is there only one product that each factory produces or are there

LP gone horribly wrong

patrizia...@gmail.com (Patrizia) — Sun, 03 Mar 2009 17:41:03 UT

Hi there!
I'm trying to solve the following problem: I've got five production
plants and numerous customers and I have to decide what amount of a
product should be produced by which plant to serve the needs of the
customers.

As I'm new to the XPress MP (Student version) I implemented that in

Re: How to know when the B&B has found an Integer solution

unicamp.pe...@gmail.com (Pedro Hokama) — Wed, 03 Mar 2009 23:27:47 UT

I'm not sure about that but:

int XPRS_CC myCut(XPRSprob prob, void* myData)
{
int IntInf;

XPRSgetintattrib(prob, XPRS_MIPINFEAS, &IntInf);
if (IntInf>0) return 0;

...

On Tue, Mar 3, 2009 at 7:26 AM, pmayadu...@gmail.com

--
Atenciosamente,
-Hokama

How to know when the B&B has found an Integer solution

pmayadu...@gmail.com (pmayaduque@gmail.com) — Tue, 03 Mar 2009 10:26:50 UT

I am working in a Branch and Cut algorithm.

I have to apply a procedure to add cuts when but only when the node is
an integer solution. Somebody knows how to determine if one node is
and Integer solution?

Kind regards

Stacked column

klokkp...@gmail.com (oystek) — Fri, 02 Feb 2009 14:58:29 UT

Is there anybody that can help me graphing stacked columns (thats what
they call it in MS excel) in Xpress (IVE)? It is a power modelling
problem where I want total power production in each period as columns,
and the different power generators as different colors in the columns.
For example if power production in period 1 is 20MWh coal and 30MWh

Re: help with removing objects in a set

david.ondra...@gmx.de (morgoth) — Fri, 01 Jan 2009 11:26:02 UT

Hi,

I think what you are looking for is a list, not a set. In a list you
may add and remove elements as you wish. I hope this helps a bit.

Greets,

David

Re: Help needed in some modelling issues

david.ondra...@gmx.de (morgoth) — Fri, 01 Jan 2009 11:21:48 UT

Hi,

for your first problem:
By defining Z(q,0)=Z0(q) you do not set an initial solution, but
create a new constraint, in your case it means that Z(q,0) has to be
[1,1,0]. Literaly, it does not say "Z(1,0) equals 1", but "In a
feasible solution, Z(1,0) has to be equal to 1". Probably, your model

Re: MAX_COMPUTATION_TIME does not work?

david.ondra...@gmx.de (morgoth) — Thu, 01 Jan 2009 21:16:55 UT

First, thank you for your answer.

Actually, I wanted to create files in a benchmark-session, like this:

setparam("XPRS_MAXTIME",1000)
tstart:=gettime
minimize( height )
tstop:=gettime
writeln("instance: ",datainpack)
writeln("heuristic: ",mH)
writeln("best UB: ", getobjval)
writeln("LB:",getparam("XPRS_B ESTBOUND"))

Re: MAX_COMPUTATION_TIME does not work?

amirah.rah...@gmail.com (amirah) — Thu, 01 Jan 2009 20:45:02 UT

The time should show up in the logfile, at the end. Here's what I use to
print out the logfile:
setparam("XPRS_MIPLOG",-1)
setparam("XPRS_CPUTIME",1)

Is this what you had in mind?

Re: MAX_COMPUTATION_TIME does not work?

david.ondra...@gmx.de (morgoth) — Thu, 01 Jan 2009 12:15:57 UT

Ok, another question:
Can I somehow read out the time taken to find the optimal solution? I
mean the cpu time, which appears when VERBOSE is set to true. Of
course, I could do a time measurement by myself, but that would be the
real time, not the CPU time, right? So.. is there a control to find

getsolatnode

eejriets...@gmail.com (Eelco) — Wed, 01 Jan 2009 09:10:51 UT

Hi all,

I have a multistage mixed integer recourse model in xpress. In order
for me to evaluate my solutions I would like to have a print function
or whatsoever, to get my results at every node of the scenario tree
for different decision variables.
In order to do so I can call some solutions with the getsolatnode()

Re: MAX_COMPUTATION_TIME does not work?

david.ondra...@gmx.de (morgoth) — Tue, 01 Jan 2009 00:19:51 UT

Thank you a lot, MAXTIME works well, just as intended.

David

Re: MAX_COMPUTATION_TIME does not work?

amirah.rah...@gmail.com (amirah) — Mon, 01 Jan 2009 23:38:06 UT

Hi there

Try using this instead.
setparam("XPRS_MAXTIME",-1800)

You can find this by choosing "Wizards" from the menu and then "Results &
tuning". A new window will pop up and under "Parameter" you'll be able to se
"MAXTIME". Click on it and it'll give you the description on the right. Here

MAX_COMPUTATION_TIME does not work?

david.ondra...@gmx.de (morgoth) — Mon, 01 Jan 2009 23:09:25 UT

Hi,

I have a mosel model which I want to test on several instances of a
problem. Therefore, I decided to set a time limit to 1800 seconds,
after which the best solution so far shall be put out.
I tried the following code line:
setparam("MAX_COMPUTATION_TIME ", 1800)

but it does not seem to work, the computation time is at 4800s right

Mixed Integer Linear Programing Duality

ekar...@gmail.com (ekarang@gmail.com) — Mon, 01 Jan 2009 18:07:12 UT

Hi all,

I have been using XPress-MP to solve mixed integer linear problems but
unfortunatelly can't get the dual values of the constraints, since
getdual command is not working for some reason with integer problems.
Does anybody know how to handle this?

Thanks,

RE: need advice Xpress-SP

chrisoak...@fairisaac.com (Oakley, Chris) — Mon, 01 Jan 2009 10:11:13 UT

The SP module is now available for free, along with the source.

[link]

No additional licensing is required, thus the module will work with the
student version. However, the student problem dimension limitations will
still apply. The module is better suited to being used with the full

Re: need advice Xpress-SP

eejriets...@gmail.com (Eelco) — Sun, 01 Jan 2009 11:27:15 UT

Hi,

I am also quite new on this, but what I know thusfar is this. There is
a student version available at dashoptimization.com, only it is
limited in the number of nonzero variables and dimensions.
I am also busy with stochastic mixed integer problems and with xpress-
mp you can call the solver mmsp which allows you to create a scenario

need advice Xpress-SP

usamaasi...@gmail.com (hev03) — Sun, 01 Jan 2009 05:34:14 UT

Hi,

I am very new to Xpress-MP. I have a stochastic mixed integer linear
problem to solve. As I read on website, seems like it can be solved
using Xpress-SP (as I need some scenario generation) which is solver
for stochastic problems. But I have no idea how to use Xpress-SP. It
has been mentioned that there is Xpress-SP reference manual but I have

Re: Help needed in some modelling issues

eejriets...@gmail.com (Eelco) — Sat, 01 Jan 2009 20:11:01 UT

Hi all,

I have another question ready for which I do not have a direct
solution. In mathematical terms what I would like to model is the
following:

B_{qt} <= Demand_{qt} + sum(v | v>q) Demand_{qt}

Where v is a subset of fareclasses, only with v>q.

When I try to do something like:

forall(q,v in FareClasses, t in TimeStages) B(q,t) <= Demand(q,t) +

Help needed in some modelling issues

eejriets...@gmail.com (Eelco) — Sat, 01 Jan 2009 13:44:37 UT

Hi all,

I just started in Xpress-MP and during the programming of my MIP I
came accross a few things that I cannot explain. The model runs
without any error messages, but does not produce any results. I think
this is caused by assigning (actually not assigning) initial values to
my variables.

Problema de Optimización con Funcion Objetivo con Logaritmos

apaola112...@gmail.com (PAOLA) — Tue, 01 Jan 2009 03:12:42 UT

Hola!

Quisiera saber si existe alguna manera de que Xpress-MP resuelva
problemas cuya función objetivo sea el logaritmo de la suma de dos
variables de decisión.

Gracias
PAOLA

Re: Conditions for mpvars in constraints, directly or indirectly

as.corazon.no...@gmail.com (Calidosaúrico) — Tue, 12 Dec 2008 04:21:32 UT

Yes; but when you say in the model

"forall(i in 1..100) x(i) is_binary"

you are implying that (X(I) can only take values of 0 or 1). And the
conditions for the X(I) values are expressed in the "PathLoad(l)" and
"PathLoad_A(l)" restrictions. The rest is done by the optimizer.

I hope this can still help you (though I guess you have already

combining Image to Xpress Model

xinhui.zh...@wright.edu (Xinhui Zhang) — Sat, 12 Dec 2008 02:05:19 UT

Hi, folks,

I am building several location models and desires to embed maps within the
Xpress, has any one tried it in IVE ? or do I have to write a GUI separated
and use data file to link the data and solutions?

Please let me know.

Xinhui Zhang, Ph.D.
Assistant Professor of Industrial Engineering

Re: ?45 error

amirah.rah...@gmail.com (amirah) — Sat, 12 Dec 2008 00:34:45 UT

I'm not sure what you mean. I'm quite new at this. Could you explain
further? Thanks for replying.

Re: ?45 error

jaful...@up.com — Fri, 12 Dec 2008 14:53:44 UT

If your temporary directory has a size limit maybe that's it.
What happens if you explicitly set it, saying something like
setparam("xprs_probname","C:/m ydirectory/pb")
?

Amirah

?45 error

amirah.rah...@gmail.com (Amirah) — Thu, 12 Dec 2008 23:16:01 UT

Hi everyone

I've been running into some problems with mosel. I leave my program
running overnight and when I check on it inthe morning, I keep getting
the same error message. Here's the error message with the end part of
the log file:

Node BestSoln BestBound Sols Active Depth Gap

help with removing objects in a set

pieter.tr...@gmail.com (pnt1990) — Thu, 12 Dec 2008 15:55:27 UT

Hi all,

I have been asked to solve the following problem in a better way by
the use of a greedy algorithm. It is called a vertex cover problem,
some of you might have heard of that one. You have to keep removing
elements from a set until it is not a vertex cover anymore. How can I
put this in MOSEL? The Mosel code looks loke this at the moment:

Re: Conditions for mpvars in constraints, directly or indirectly

peter_s...@yahoo.com (p_sopa) — Wed, 12 Dec 2008 08:18:29 UT

Hi, OK I have got the beer.. So I adjust the pathload instead of the
pathcost? But how do I set X(l) the binary variable. I need to say
that for certain conditions X(l) is 1 otherwise is 0.

Peter