Skip to content

Metadata Checks

Matthew Cornell edited this page Jul 5, 2022 · 9 revisions

The metadata checks are implemented to conform to the specifications specified in this file. In addition to them, the following checks are run:

  • Metadata file must have proper yaml format.
  • team_name gives the name of the team and is not validated.
  • model_name gives the name of the model and is not validated.
  • model_abbr must be distinct from any already existing model_abbr.
  • model_contributors must include at least one email address and written in the format specified here.
  • team_model_designation should be one of primary, secondary, proposed, or other. There should be only one primary model for every team_abbr, where (in Python) team_abbr = model_abbr.split('-')[0], e.g., 'COVIDhub-baseline' -> 'COVIDhub'.
  • license should be one of the licenses specified in the license column of this CSV.
  • website_url field must be specified. This should be a valid URL.
  • methods is under 200 characters.
  • The name of the directory under which all the forecasts and metadata files are added must be the model_abbr specified in the metadata file.