Skip to content

multidimensional_evaluation_engine

The multidimensional_evaluation_engine provides a structured framework for evaluating alternatives using typed inputs and configurable policy rules.

It includes components to:

  • load candidate data (CSV to typed factor values)
  • load policy configurations (TOML to factor specs, constraints, scoring rules)
  • evaluate candidates using constraint and scoring logic
  • compute aggregate scores and admissibility indicators
  • map results to qualitative interpretations
  • produce structured outputs for inspection and comparison

The engine is domain-neutral and supports scenarios where tradeoffs must be made explicit without embedding domain-specific assumptions in the core.

Installation

pip install multidimensional-evaluation-engine

Or with uv:

uv add multidimensional-evaluation-engine
uv sync

Documentation

  • API Reference
  • Concepts
  • Examples
  • Troubleshooting