Just need to make sure that I was not overlooking anything seeing how I could not find a solver on their website.

asked 19 Sep '10, 20:41

Berk%20Ustun's gravatar image

Berk Ustun
3071311
accept rate: 0%

edited 27 Sep '10, 19:44

In case you guys were interested, I also recently discovered that you can solve some SPs using MS Solver Foundation 2.1. It's not AMPL but at least it works with Excel :)

http://msdn.microsoft.com/en-us/library/ff524509%28v=VS.93%29.aspx

(27 Sep '10, 19:43) Berk Ustun

My understanding is that Dave Gay is working on the AMPL stochastic programming extensions right now. So, perhaps this will be available before too long!

link

answered 24 Sep '10, 00:46

Bill%20Hart's gravatar image

Bill Hart
361
accept rate: 100%

It's also worth taking a look at the PySP stochastic programming extensions to the Pyomo modeling tool. PySP and Pyomo are part of the Coopr optimization package: see projects.coin-or.org/Coopr or https://software.sandia.gov/trac/coopr . PySP has a native representation (using Pyomo) that can be used directly (e.g. using PySP's progressive hedging solver) or to generate the deterministic equivalent.

(24 Sep '10, 00:53) Bill Hart

Thanks for the PySP link - I think this is as close as it gets right now. Hopefully the AMPL extension will be out soon :)

(25 Sep '10, 23:47) Berk Ustun

I could be wrong, but I had always assumed that AMPL produced the deterministic equivalent and invoked a standard solver.

Does anyone know different?

link

answered 19 Sep '10, 23:08

Matthew%20Saltzman's gravatar image

Matthew Salt... ♦
4.7k310
accept rate: 17%

1

I didn't think the stochastic extensions (which apparently have been in the works since '96, judging by the AMPL web site) were available yet. From what I read in a JoC article (http://joc.journal.informs.org/cgi/content/abstract/21/1/107), I thought the plan was to generate deterministic models for various scenarios and feed them to a standard solver, then synthesize the results in AMPL.

(20 Sep '10, 00:48) Paul Rubin ♦♦

There is the paper about the AMPL stochastic programming specification, StAMPL, but even OptiRisk Systems, who promote AMPL as a modeling environment, have their own modeling tool for stochastic programming. I might be wrong, but this tells me that they think AMPL is not usable for Stochastic Programming. The only widely accepted specification of stochastic programs seems to be SMPS (, although a specific solver could implement its own stochastic programming interface for AMPL using AMPL suffixes).

link

answered 19 Sep '10, 23:25

Hans's gravatar image

Hans
1093
accept rate: 0%

1

Possibly also worth a look is the COIN-OR SMI (Stochastic Modeling Interface) at https://projects.coin-or.org/Smi. It supports SMPS as well as its own modeling tools and any COIN OSI-compatible solver.

(23 Sep '10, 13:56) Matthew Salt... ♦

You can use smpswriter to convert two-stage scenario based stochastic programming problems from deterministic equivalent written in AMPL into SMPS format that is supported by some (most?) SP solvers.

link

answered 11 Feb '14, 20:44

vitaut's gravatar image

vitaut
32718
accept rate: 20%

Your answer
toggle preview

Follow this question

By Email:

Once you sign in you will be able to subscribe for any updates here

By RSS:

Answers

Answers and Comments

Markdown Basics

  • *italic* or _italic_
  • **bold** or __bold__
  • link:[text](http://url.com/ "Title")
  • image?![alt text](/path/img.jpg "Title")
  • numbered list: 1. Foo 2. Bar
  • to add a line break simply add two spaces to where you would like the new line to be.
  • basic HTML tags are also supported

Tags:

×58
×40
×29

Asked: 19 Sep '10, 20:41

Seen: 4,647 times

Last updated: 11 Feb '14, 20:44

OR-Exchange! Your site for questions, answers, and announcements about operations research.