Hub configuration files

Hub configuration files#

Directory structure#

The hub-config directory in a modeling hub is required to contain three JSON1We do not currently support json references or yaml metadata files due to technical issues. files:

  1. admin.json - JSON file containing generic information about the hub and static configuration settings for downstream tools such as validations, visualizations, etc. This file also contains optional cloud settings for hubs that use cloud storage. See the hub administrative configuration (admin.json) interactive schema section below for details on the admin.json file.

  2. tasks.json - JSON file specifying modeling tasks and model output formats, which may be round-specific. See the hub model task configuration (tasks.json file) section below for more details on the tasks.json file.

  3. model-metadata-schema.json - JSON file defining format of model metadata files. The template metadata schema file section has more information on the model-metadata-schema.json file.

Hubs with target data may contain the following optional JSON file:

  • target-data - JSON file specifying target data objects with top-level properties that describe expectations across target datasets, and with the ability to override these defaults for specific dataset types (time-series and oracle-output). See the target data configuration (target-data.json) preview below.

Purpose#

The files within the hub-config directory specify general configurations for a hub and (possibly round-specific) details of what model outputs are requested or required. Hub configuration files are used for:

  • Validating model output submissions

    • tasks.json file specifies the file format, task ID, output type, and value combinations (both required or optional) the submitted model output data must adhere to

    • tasks.json file also specifies the submission window for each round (with the time zone information in the admin.json file)

  • Scoring model outputs

    • the hub configuration files specify the scores that are used

    • the task ID variables specified in the tasks.json can be used to join model output data with truth data to score forecasts

  • Configuring model output visualizations

    • Visualization tools may benefit from programmatically identifying task ID variables so that a separate visualization of model outputs can be generated for each combination of those variables (e.g., via facetting or menu selections). For example, producing separate visualizations for different locations or scenario IDs may be beneficial.

    • The tasks.json file contains metadata about the targets, including a human-readable description and units that can be used for visualization.

NOTE

All objects in admin-schema.json and tasks.json disallow additional properties by default. Custom properties are only permitted within additional_metadata fields, which are allowed in tasks.round objects, tasks.rounds.model_tasks.target_metadata objects, and at the top levels of target-data.json and admin.json files.

Hub administrative configuration (admin.json file)#

The administrative hub configuration file contains global administrative settings expected to remain fixed throughout a hub’s existence. These settings apply to all rounds in a hub.

As of v2.0.1, admin.json contains optional settings for hubs that store their configuration and model-output data in the cloud (at this time, Amazon Web Services is the supported cloud provider). Like other admin settings, the cloud information should not be updated once the hub has been launched (doing so will break the process that syncs hub data to the cloud).

Hub administrative configuration (admin.json) interactive schema#

Schema Version: v6.0.0#

Please note that the preview below does not show the required fields. Please click on the raw schema link below to see all required fields.

See raw schema

Hub model task configuration (tasks.json file)#

The hub model task configuration file specifies the model tasks (task IDs and targets) and model output types. The tasks.json file is flexible enough to accommodate different hub styles. Hubs can vary from a simple forecast hub (see US Forecast Hub example to a more complex round-related scenario hub (see US Scenario Modeling Hub example).

Model tasks (tasks.json) interactive schema#

Schema Version: v6.0.0#

See raw schema

Hub target data configuration (target-data.json file)#

For hubs that use target data, the optional target data configuration file contains multiple top-level properties that describe target data format expectations across target datasets. Top-level properties can be overridden at the dataset-specific (time-series and oracle output) level. The target data configuration file removes the need to inspect dataset contents to infer schemas, thus allowing target data to be loaded faster and ensuring reproducible data validation and versioning behavior.

Hub target data configuration (target-data.json) interactive schema#

Schema Version: v6.0.0#

See raw schema