How to quickly load data over a loop (different scenarios

GAMS
1

Dear Antho,



Please tell me the best answer you find. Many of us would like to do the same problem, something like running a base model, then looping over scenarios, such as


\

  1.   Alternative prices

  2.   Alternative production costs

  3.   Alternative water supplies/climate scenarios

  4.   Alternative technical advance possibilities

  5.   Alternative government policy proposals




Looping over all those sets is a common need many of us have. It lets us find optimized response patterns to a bundle of future uncertainties.



Frank



From: gamsworld@googlegroups.com [mailto:gamsworld@googlegroups.com] On Behalf Of Anthony
Sent: Thursday, February 04, 2016 12:22 PM
To: gamsworld
Subject: How to quickly load data over a loop (different scenarios)???



Dear Gents,



All I need is to run 1000 scenarios on a model with various values for two parameters (a(i) and b(j)) in each run and save the results of each run separately. In need to set a mechanism so that GAMS read and put each row of the parameter in the model in each run and save the results to end.

For example in the Transport example, I need to run the model in 1000 different scenarios (1000 different problem sizes defined in excel rows) with different value for a(i) and b(j) over a loop and save the Z.L to different rows in a spreadsheet.



/////////////////////////////////////////////////////////

a(i) = /

i1 i2 i3

Scenario1 100 200 300

Scenario2 120 240 270

. . .

. . .

. . .

Scenario1000 115 213 310

///////////////////////////////////////////////////////////////////

b(j)

J1 J2 J3

Scenario1 100 200 300

Scenario2 120 240 270

. . .

. . .

. . .

Scenario1000 115 213 310

//////////////////////////////////////////////////////////////////





Consider the code is like below:



Set sc /Scenario1*Scenario1000/



Parameter A(i),B(j), Total_cost;



Equatios;



Objective Max X =z.l;

Cons1… A(i)x < B(j)



Loop(sc,

A(i) = A(scenario(sc))

B(j) = B(scenario(sc))



Solve model MIP Maximising Total_cost;

Total_cost(sc,“Z.L”)=z.l;

);

Putclose;

Display Total_cost;

Execute_Unload “All_scenarios.gdx”,Total_cost;

Execute ‘GDXXRW.EXE All_scenarios.gdx epsOut=0 par=Total_cost rng=Total_costs!a1 trace=0’



Just wondering how to import each row to the model in each scenario(sc)? Is there any proper trick?



Looking forward to hearing back from you guys.



Cheers,

Antho


\

To unsubscribe from this group and stop receiving emails from it, send an email to gamsworld+unsubscribe@googlegroups.com.
To post to this group, send email to gamsworld@googlegroups.com.
Visit this group at https://groups.google.com/group/gamsworld.
For more options, visit https://groups.google.com/d/optout.

–
To unsubscribe from this group and stop receiving emails from it, send an email to gamsworld+unsubscribe@googlegroups.com.
To post to this group, send email to gamsworld@googlegroups.com.
Visit this group at https://groups.google.com/group/gamsworld.
For more options, visit https://groups.google.com/d/optout.

2

Reply-to: gamsworld@googlegroups.com

Dear Antho; Dear Frank,

I had very similar problems.

And, Frank, I am exactly working from 3-5 in your list. I am working CGE model.

Amsalu

On Friday, 5 February 2016 18:35:32 UTC+1, Frank Ward wrote:

Dear Antho,



Please tell me the best answer you find. Many of us would like to do the same problem, something like running a base model, then looping over scenarios, such as


\

  1.   Alternative prices

  2.   Alternative production costs

  3.   Alternative water supplies/climate scenarios

  4.   Alternative technical advance possibilities

  5.   Alternative government policy proposals




Looping over all those sets is a common need many of us have. It lets us find optimized response patterns to a bundle of future uncertainties.



Frank



From: gams...@googlegroups.com [mailto:gams...@googlegroups.com] On Behalf Of Anthony
Sent: Thursday, February 04, 2016 12:22 PM
To: gamsworld
Subject: How to quickly load data over a loop (different scenarios)???



Dear Gents,



All I need is to run 1000 scenarios on a model with various values for two parameters (a(i) and b(j)) in each run and save the results of each run separately. In need to set a mechanism so that GAMS read and put each row of the parameter in the model in each run and save the results to end.

For example in the Transport example, I need to run the model in 1000 different scenarios (1000 different problem sizes defined in excel rows) with different value for a(i) and b(j) over a loop and save the Z.L to different rows in a spreadsheet.



/////////////////////////////////////////////////////////

a(i) = /

i1 i2 i3

Scenario1 100 200 300

Scenario2 120 240 270

. . .

. . .

. . .

Scenario1000 115 213 310

///////////////////////////////////////////////////////////////////

b(j)

J1 J2 J3

Scenario1 100 200 300

Scenario2 120 240 270

. . .

. . .

. . .

Scenario1000 115 213 310

//////////////////////////////////////////////////////////////////





Consider the code is like below:



Set sc /Scenario1*Scenario1000/



Parameter A(i),B(j), Total_cost;



Equatios;



Objective Max X =z.l;

Cons1… A(i)x < B(j)



Loop(sc,

A(i) = A(scenario(sc))

B(j) = B(scenario(sc))



Solve model MIP Maximising Total_cost;

Total_cost(sc,“Z.L”)=z.l;

);

Putclose;

Display Total_cost;

Execute_Unload “All_scenarios.gdx”,Total_cost;

Execute ‘GDXXRW.EXE All_scenarios.gdx epsOut=0 par=Total_cost rng=Total_costs!a1 trace=0’



Just wondering how to import each row to the model in each scenario(sc)? Is there any proper trick?



Looking forward to hearing back from you guys.



Cheers,

Antho


\

To unsubscribe from this group and stop receiving emails from it, send an email to gamsworld+...@googlegroups.com.
To post to this group, send email to gams...@googlegroups.com.
Visit this group at https://groups.google.com/group/gamsworld.
For more options, visit https://groups.google.com/d/optout.

–
To unsubscribe from this group and stop receiving emails from it, send an email to gamsworld+unsubscribe@googlegroups.com.
To post to this group, send email to gamsworld@googlegroups.com.
Visit this group at https://groups.google.com/group/gamsworld.
For more options, visit https://groups.google.com/d/optout.

3

I probably don’t understand but here my input:



If you have the parameter A and B in your excel sheet as you describe, you just load them using gdxxrw into the parameters A0(scenario,i) and B0(scenario,j) and then just use them in the equations accordingly.



Parameter a,b;



Equation Cons1… Ax
Cc: anthoy.optimizer@gmail.com; fward@nmsu.edu
Subject: Re: How to quickly load data over a loop (different scenarios



Dear Antho; Dear Frank,



I had very similar problems.



And, Frank, I am exactly working from 3-5 in your list. I am working CGE model.



Amsalu


On Friday, 5 February 2016 18:35:32 UTC+1, Frank Ward wrote:

Dear Antho,



Please tell me the best answer you find. Many of us would like to do the same problem, something like running a base model, then looping over scenarios, such as


\

  1.   Alternative prices

  2.   Alternative production costs

  3.   Alternative water supplies/climate scenarios

  4.   Alternative technical advance possibilities

  5.   Alternative government policy proposals




Looping over all those sets is a common need many of us have. It lets us find optimized response patterns to a bundle of future uncertainties.



Frank



From: gams...@googlegroups.com [mailto:gams...@googlegroups.com] On Behalf Of Anthony
Sent: Thursday, February 04, 2016 12:22 PM
To: gamsworld
Subject: How to quickly load data over a loop (different scenarios)???



Dear Gents,



All I need is to run 1000 scenarios on a model with various values for two parameters (a(i) and b(j)) in each run and save the results of each run separately. In need to set a mechanism so that GAMS read and put each row of the parameter in the model in each run and save the results to end.

For example in the Transport example, I need to run the model in 1000 different scenarios (1000 different problem sizes defined in excel rows) with different value for a(i) and b(j) over a loop and save the Z.L to different rows in a spreadsheet.



/////////////////////////////////////////////////////////

a(i) = /

i1 i2 i3

Scenario1 100 200 300

Scenario2 120 240 270

. . .

. . .

. . .

Scenario1000 115 213 310

///////////////////////////////////////////////////////////////////

b(j)

J1 J2 J3

Scenario1 100 200 300

Scenario2 120 240 270

. . .

. . .

. . .

Scenario1000 115 213 310

//////////////////////////////////////////////////////////////////





Consider the code is like below:



Set sc /Scenario1*Scenario1000/



Parameter A(i),B(j), Total_cost;



Equatios;



Objective Max X =z.l;

Cons1… A(i)x < B(j)



Loop(sc,

A(i) = A(scenario(sc))

B(j) = B(scenario(sc))



Solve model MIP Maximising Total_cost;

Total_cost(sc,“Z.L”)=z.l;

);

Putclose;

Display Total_cost;

Execute_Unload “All_scenarios.gdx”,Total_cost;

Execute ‘GDXXRW.EXE All_scenarios.gdx epsOut=0 par=Total_cost rng=Total_costs!a1 trace=0’



Just wondering how to import each row to the model in each scenario(sc)? Is there any proper trick?



Looking forward to hearing back from you guys.



Cheers,

Antho


\

To unsubscribe from this group and stop receiving emails from it, send an email to gamsworld+...@googlegroups.com.
To post to this group, send email to gams...@googlegroups.com.
Visit this group at https://groups.google.com/group/gamsworld.
For more options, visit https://groups.google.com/d/optout.

–
To unsubscribe from this group and stop receiving emails from it, send an email to gamsworld+unsubscribe@googlegroups.com.
To post to this group, send email to gamsworld@googlegroups.com.
Visit this group at https://groups.google.com/group/gamsworld.
For more options, visit https://groups.google.com/d/optout.

–
To unsubscribe from this group and stop receiving emails from it, send an email to gamsworld+unsubscribe@googlegroups.com.
To post to this group, send email to gamsworld@googlegroups.com.
Visit this group at https://groups.google.com/group/gamsworld.
For more options, visit https://groups.google.com/d/optout.