AMPL first time usage problems

[Burcu Akyildiz]

Hi all,

Maybe what I will going to ask is so easy for many of you, however this is my first time of using AMPL. Therefore I need help in a basic level. 

I am attaching my Mathematical model in doc format, my AMPL model and data file.

I kindly ask you to help me in the following issues:

• First, am I able to prepare AMPL model in the right way for my mathematical model?
  
• Second, am I able to prepare data file in the right way?
  
• Last, I am not able to solve the model. I try to solve it like below and it gives the error:

ampl: model Facility.mod; data facility.dat; solve;

Can't find  file "Facility.mod"

context:  model  >>> Facility.mod <<< ; data facility.dat; solve;

Can't find  file "facility.dat"

context:  model Facility.mod; data  >>> facility.dat <<< ; solve;

No variables declared.

Thank you in advance!

[ptiwari]

Your error message suggest that AMPL couldn't find the model and data file. Provide the correct and complete path of your model and data file to AMPL.

[Indrajit Sen Gupta]

Looks like you have missed a semi-colon on line 13 after defining the parameter M.

Can you describe the nature of the problem - looks interesting.

[Indrajit Sen Gupta]

There are a bunch of other issues that you will face when trying to run this model. Have you gone through AMPL Modeling Language book: http://ampl.com/resources/the-ampl-book/ ?

Regards,

Indrajit

[Burcu Akyildiz]

Yes I did. I have changed the model and data as attached. Still the same problem.

ampl: model C:\Users\BURCU\Desktop\Tez\AMPL_TEZ\Facility.mod

C:\Users\BURCU\Desktop\Tez\AMPL_TEZ\Facility.mod, line 37 (offset 2332):

syntax error

context:  subject to limit_flow_length {i in DEPT}:  >>> 0=<x <<< 2[i]<=x1[i]<=B_length;

ampl? 

[ptiwari]

You seems to be combining the more than one inequalities constraint in one statement. Rewrite your constraint subject to limit_flow_length {i in DEPT}: 0<=x2[i]<=x1[i]<=B_length; as follows:

subject to limit_flow_length {i in DEPT}: 0<=x2[i]<=B_length;
subject to limit_flow_length1 {i in DEPT}:0<=x1[i]<=B_length;

Thanks,
Paras

[Burcu Akyildiz]

Dear Mr Paras,

I am still in trouble with AMPL. 

You told me to write the f variable in open format in mod file, I did but unfortunately now AMPL says that all d in for f variable are not defined. 

As an example.

ampl: model C:\Users\BURCU\Desktop\Tez\AMPL_TEZ\Facility-ba2.mod

C:\Users\BURCU\Desktop\Tez\AMPL_TEZ\Facility-ba2.mod, line 52 (offset 3272):

d1 is undefined

context:  subject to flow1for01:  >>> f[d1,d1] <<< =0; 

I am attaching the files:mod , dat and word document of my model.

I will be glad if you can help me to come over these problems with AMPL. 

[ptiwari]

You haven't properly defined the index. You need to define 'd1' in your expression f[d1,d1]. You have defined indexing correctly in some expressions. For e.g. subject to limitabsolute1 {i in DEPT,j in DEPT}: dx[i,j]. Look into "Indexing Expression" in page 79 of http://ampl.com/BOOK/CHAPTERS/08-sets1.pdf book.

Thanks,

Paras

[Robert Fourer]

If you refer to d1 in your model, AMPL interprets it as an identifier for a model component, and when it see that no param (or other model component) is named d1, it gives you the "d1 is undefined" message. To refer in a model to the *string* d1 that is a member of set DEPT, put it in quotes:

subject to flow1for01: f["d1","d1"] = 0;

The same comments apply to other strings that are mentioned in constraints in the model.

Bob Fourer
am...@googlegroups.com

=======

[Burcu Akyildiz]

Hi Mr Fourer,

It was the case! Thanks!