Error log files - v. 06/22/22 Temoa - Utopia scenario

[Romain Chaffanjon]

Dear all, 

I have some difficulties with the Calculating reporting variables and formatting results of Temoa. 

The simulation parameters is as indicated in the attached document, and the version of temoa used is the last version from the Temoa project branch of Github repository.

About the results, all the outputs variable are calculated and filled in the sqlite file, expected the output_dual. By the way, I'm not sure about what it is, is it the solution of dual variables ?

Thank you, 

Regards, 

Romain Chaffanjon

[Temoa Project]

Hi Romain,

Could you try pulling the most recent commit to the repository and testing again? I made some changes to address what might be recent updates to one of the packages we use.

If you want the Output_Duals table to be populated as well, you can make the change in the config file: uncomment the line #--saveDUALS in the config file. This stores the dual values of the constraints in the model (shadow prices) , as reported by the solver that you use.

Best,

Aranya

[Romain Chaffanjon]

Hi Aranya, 

I just used the last version but I have a new error now (see attached). Anyway, it is not that important, I can comment #--saveexcel and everything is running correctly. Thank you for your help !

That's what I did to get shadow prices ! What I don't really figure out is why I get negative values for DemandConstraint, so they are not printed in output_duals. 

I'm not sure how the dual is implemented, but in TIMES-Markal this specific dual value is indicating the marginal price of commodities.

Without price demand equilibrium, I cannot understand a negative value (I'm still using utopia scenario btw)

I slightly modified the pformat_results.py by keeping all dual values, and without transformation (see attached), and now I get positive values for all DemandConstraint. 

If you have any documentation or explanation about this it would be so helpful !

Thank you, 

Romain

[Aranya Venkatesh]

Hi Romain,

When I run temoa_utopia without any modifications to pformat_results.py with the Gurobi solver, I see positive values for the DemandActivity constraints in the Output_Duals table, as expected.

(you'll see in pformat_results that we save the negative value of the duals reported by the solver in the Output_Duals table - any positive values reported by the solver can be ignored, since it can be assumed that those constraints are not binding). 

Could you let me know which solver you are using? 

Best,

Aranya

--
You received this message because you are subscribed to the Google Groups "Temoa Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email to temoa-projec...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/temoa-project/2f4e4136-73d1-4f2a-b246-c9b499afd542n%40googlegroups.com.

[Temoa Project]

Hi Romain,

After some internal discussions, we made a few changes to pformat_results.py, where we now allow for all dual variables (with an absolute value of greater than 1e-5) to be reported to the Output_Duals table. Our earlier decision to report the duals with negative values only was based our interest in a few key constraints that were set up as <= constraints specifically (e.g. the EmissionLimit constraints).

You'll see from the Gurobi documentation here how the dual values are expected to change based on the type of constraint specified. Given that the sign of the dual values depend on how each individual constraint is formulated, we now output all the dual values directly reported by the solver to the Output_Duals table. I believe this is now consistent with the modification that you had tested earlier. 

I hope this helps. 

Best,

Aranya