electric-barometer 0.2.0

Umbrella package for the Electric Barometer ecosystem (metrics, evaluation, features, adapters, and related tooling).

Electric Barometer

Electric Barometer is a modular, cost-aware forecasting evaluation framework designed for operational decision-making. It provides a structured way to evaluate, compare, and select forecasts when error costs are asymmetric and operational consequences matter.

Rather than delivering a single monolithic library, Electric Barometer is intentionally organized as a small ecosystem of focused packages, each with a clear responsibility.

This repository serves as the umbrella distribution and conceptual entry point for the Electric Barometer ecosystem.

---

The Electric Barometer Ecosystem

Electric Barometer is composed of several interoperable packages:

• eb-metrics
  Defines individual forecast error and service metrics, including cost-asymmetric measures such as Cost-Weighted Service Loss (CWSL), Forecast Readiness Score (FRS), and related primitives.

• eb-evaluation
  Provides DataFrame-first utilities for applying metrics across entities, groups, hierarchies, and time windows. This layer handles evaluation, comparison, and selection logic while delegating metric math to eb-metrics.

• eb-adapters
  Normalizes interfaces for external forecasting and regression libraries so they can be evaluated consistently. Adapters expose a common .fit / .predict contract for heterogeneous models.

• eb-examples
  Contains worked examples, notebooks, and practical demonstrations showing how the Electric Barometer ecosystem is used end-to-end in real scenarios.

• eb-papers
  The source of truth for conceptual definitions, theoretical foundations, and methodological rationale behind Electric Barometer metrics and frameworks.

Each package is versioned, tested, and documented independently, but designed to work together seamlessly.

---

What This Repository Provides

This electric-barometer repository:

• Acts as the canonical entry point to the ecosystem
• Provides a single install surface for core Electric Barometer functionality
• Establishes the conceptual map of the project
• Ensures compatible dependency resolution across subpackages

It intentionally contains minimal implementation code.

---

Installation

Install the Electric Barometer umbrella package via pip:

pip install electric-barometer

This installs the core dependencies required to work with Electric Barometer metrics. Additional functionality is provided by installing the underlying packages directly (e.g., eb-evaluation, eb-adapters) or via future optional extras.

For development:

pip install -e .

---

Design Philosophy

Electric Barometer is built around a few core principles:

• Separation of concerns
  Metric definitions, evaluation logic, and model interfaces live in separate packages.

• Cost-aware evaluation
  Forecast accuracy is evaluated in terms of operational impact, not symmetric error alone.

• Operational realism
  Metrics and frameworks are designed for environments where underbuild and overbuild have different consequences.

• Composable tooling
  Users can adopt only the layers they need without committing to a monolith.

---

Examples and Tutorials

Examples, notebooks, and applied workflows are maintained in the separate eb-examples repository.

This repository intentionally avoids embedding example code to keep the core packages lean and focused.

---

Documentation

Unified documentation for the Electric Barometer ecosystem is available at:

https://economistician.github.io/eb-docs/

Documentation is generated directly from source code docstrings and kept consistent across packages.

---

Status

Electric Barometer is under active development. Public APIs may evolve prior to the first stable release.

---

Authorship and Stewardship

The Electric Barometer ecosystem is designed and maintained by
Kyle Corrie under the Economistician moniker.

The project reflects applied research and production experience in forecasting, operations research, and cost-asymmetric decision systems within large-scale operational environments.

For questions, collaboration, or research inquiries:

• GitHub: https://github.com/Economistician
• Contact: [e-mail]

Conceptual foundations and formal methodology are documented in the companion research repository eb-papers.

---

License

This project is licensed under the BSD 3-Clause License.