Issue fixes (#39 and #31)

[kemccusker]

• Fixes "global_cons" arg unused in dscim.utils.menu_runs.run_rff() #39
• Fixes convert_old_to_newformat_AR()'s "pulseyrs" arg has a mutable default #31

[brews]

@kemccusker This is marked as "Draft" in the title. Is this ready to review and merge or is another component missing from this PR?

[kemccusker]

@kemccusker This is marked as "Draft" in the title. Is this ready to review and merge or is another component missing from this PR?

Oops sorry @brews, removed the Draft. It's ready for review.

[brews]

LGTM! I think this can merge whenever you're ready, @kemccusker.

[brews]

Should note this PR has breaking changes to dscim.preprocessing.climate.reformat.convert_old_to_newformat_AR() and dscim.utils.menu_runs.run_rff() because it drops the pulseyrs and global_cons arguments.

That's fine but worth mentioning in case someone wants to know why their code no longer works.

[kemccusker]

Thanks for listing the breaking changes, @brews. @davidrzhdu is handling the pulseyrs args in other functions.

I want to confirm from @JMGilbert that we no longer need global_cons for any reason -- it is unused here, but used to be a flag that would force the use of "global consumption" for discounting of domestic damages. I believe (and @JMGilbert can confirm) that we have updated the EconVars and configs to specify explicitly which consumption file to use for discounting and no longer need the flag.

[JMGilbert]

Thanks for listing the breaking changes, @brews. @davidrzhdu is handling the pulseyrs args in other functions.

I want to confirm from @JMGilbert that we no longer need global_cons for any reason -- it is unused here, but used to be a flag that would force the use of "global consumption" for discounting of domestic damages. I believe (and @JMGilbert can confirm) that we have updated the EconVars and configs to specify explicitly which consumption file to use for discounting and no longer need the flag.

I am pretty sure that we no longer use this argument or need it at all.

[JMGilbert]

For future reference:
global_cons was formerly the argument used to save out global_consumption_no_pulse.nc4 files. This has been deprecated because of the @save decorator implemented into dscim which allows for saving through the config:

global_parameters:
  save_files:
      - global_consumption_no_pulse

When calculating country specific SCGHGs, discounting using only the future socioeconomics of that country would be improper. To obtain the proper SCGHG, calculate discount factors using global consumption no pulse, then discount the marginal damages from that country and sum across years.

[brews]

I'm going to update the CHANGELOG manually after this merges as time is tight on this.