Skip to content

Run gcm_run.j directly -- part2, rename suites#703

Draft
Dooruk wants to merge 9 commits intofeature/exec_geos_direct_part1from
feature/exec_geos_direct_part2
Draft

Run gcm_run.j directly -- part2, rename suites#703
Dooruk wants to merge 9 commits intofeature/exec_geos_direct_part1from
feature/exec_geos_direct_part2

Conversation

@Dooruk
Copy link
Collaborator

@Dooruk Dooruk commented Feb 12, 2026

More low-level changes to make this change mergeable with develop, still work in progress

See #677 for more details

Partially addresses #700

Remainin things:

  • Get rid of obsolete entries/tasks
  • Documentation update

@Dooruk
Copy link
Collaborator Author

Dooruk commented Feb 18, 2026

@mranst a question for you regarding suite_config.py. For this PR, I would like to test the forecast_coupled_geos (originally forecast_geos) suite, which is not in tier1 tests. I'm trying to change the cycle_times but is key only used if it's within model level? When I create with the following my experiment.yaml cycle_times is still T00:

# --------------------------------------------------------------------------------------------------

class SuiteConfig(QuestionContainer, Enum):

    # --------------------------------------------------------------------------------------------------

    forecast_coupled_geos = QuestionList(
        list_name="forecast_coupled_geos",
        questions=[
            sq.all_suites,
            qd.start_cycle_point("2021-07-02T12:00:00Z"),
            qd.final_cycle_point("2021-07-03T12:00:00Z"),
            qd.cycle_times([
                "T12",
            ]),
            qd.geos_build_method("use_existing"),
            qd.forecast_duration("P1D"),
        ],
    )

    # --------------------------------------------------------------------------------------------------

    forecast_coupled_geos_tier1 = QuestionList(
        list_name="forecast_coupled_geos_tier1",
        questions=[
            forecast_coupled_geos
        ]
    )

    # --------------------------------------------------------------------------------------------------

@mranst
Copy link
Collaborator

mranst commented Feb 18, 2026

@mranst a question for you regarding suite_config.py. For this PR, I would like to test the forecast_coupled_geos (originally forecast_geos) suite, which is not in tier1 tests. I'm trying to change the cycle_times but is key only used if it's within model level? When I create with the following my experiment.yaml cycle_times is still T00:

Yes, cycle_times is usually a model-dependent key, it's also handled weirdly by swell.

# If there are no models and the cycle_times is in the keys then remove the models key from
# the cycle_times question dictionary
if 'cycle_times' in self.question_dictionary_model_ind.keys():
if not self.suite_needs_model_components:
self.question_dictionary_model_ind['cycle_times'].pop('models')
self.question_dictionary_model_ind['cycle_times']['default_value'] = 'T00'

You can change this to ->

        # If there are no models and the cycle_times is in the keys then remove the models key from
        # the cycle_times question dictionary
        if 'cycle_times' in self.question_dictionary_model_ind.keys():
            if not self.suite_needs_model_components:
                self.question_dictionary_model_ind['cycle_times'].pop('models')
                if self.question_dictionary_model_ind['cycle_times']['default_value'] == 'defer_to_model':
                    self.question_dictionary_model_ind['cycle_times']['default_value'] = 'T00'

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants

Comments