Skip to content

opera-eco

Purpose

opera-eco coordinates the OPERA forecasting ecosystem. It keeps compatible package releases together, distributes shared coding-assistant skills, and checks that the packages work together through common interfaces and workflows testing.

Features

  • Install a compatible set of OPERA forecasting packages.
  • Install bundled skills for the OPERA ecosystem and its individual modules.
  • Run shared contract tests for package interfaces and data exchange.
  • Run end-to-end pipeline tests across models, forecasting, evaluation, combination, and news decomposition.
  • Support package maintainers as they test changes across the ecosystem.

The package keeps model logic separate from real-time workflow orchestration. Model packages provide estimation and forecasting; workflow packages run models across vintages, evaluate results, combine forecasts, and explain revisions.

Quick start

Install the package with its test dependencies:

python -m pip install "opera-eco[test]"

Run the shared test suites from a package repository:

pytest --pyargs opera.tests -m contract
pytest --pyargs opera.tests -m pipeline

Install the bundled skills with the command-line interface:

opera install skills

Repository

Read the implementation and full API reference in the opera-eco repository.