Earth-2 Model Intercomparison Project (MIP) is a python framework that enables climate researchers and scientists to inter-compare AI models for weather and climate.
Improve handling of history. Previously models with history (e.g. DLWP) could not be initialized with data sources without history. A work-around was to manually building a new data source, and made the notebooks/examples messy. Also, this was preventing using dlwp for ensemble inference with CDS inits for example.
I added support for history to the function get_initial_condition_for_model, which is used by all the inference routines.
Checklist
[x] I am familiar with the Contributing Guidelines.
[x] New or existing tests cover these changes.
[x] The documentation is up to date with these changes.
[x] The CHANGELOG.md is up to date with these changes.
Earth-2 MIP Pull Request
Description
Improve handling of history. Previously models with history (e.g. DLWP) could not be initialized with data sources without history. A work-around was to manually building a new data source, and made the notebooks/examples messy. Also, this was preventing using dlwp for ensemble inference with CDS inits for example.
I added support for history to the function
get_initial_condition_for_model
, which is used by all the inference routines.Checklist
Dependencies