API Reference

The Liberata Scientometrics library provides a comprehensive set of modules for analyzing and computing metrics on academic knowledge graphs and portfolios. This section documents the complete API.

Core Modules

The library is organized into four main components:

Metrics (liberata_metrics.metrics package)

Compute various scientometric metrics on your data:

  • Portfolio Metrics: Analyze manuscript portfolios, including returns, risk, correlations, and spectral properties

  • Market Metrics: Evaluate market dynamics and portfolio performance in the academic capital system

  • Distribution Metrics: Analyze distributions of contributors and capital allocation

  • Graph Metrics: Compute network-based metrics on citation and collaboration graphs

  • System Health Metrics: Monitor the overall health of the academic capital system

Generators (liberata_metrics.generators package)

Create synthetic data for testing and analysis:

  • Generate reference matrices from scratch

  • Build contributor-manuscript relationships

  • Create topic assignments based on OpenAlex topics

  • Produce COO-format sparse matrices for efficient computation

Utilities (liberata_metrics.utils package)

Helper functions for common tasks:

  • Data loading from Supabase and local sources

  • Data wrangling and transformation

  • Matrix operations and utilities

  • Sparse matrix utilities

Visualizations (liberata_metrics.visualizations package)

Create visual representations of your data:

  • Matrix visualizations (heatmaps, sparsity plots)

  • Time series visualizations

  • Network and graph visualizations

  • Portfolio analysis plots

Integrations (liberata_metrics.integrations package)

Connect to external systems:

  • Supabase integration for production data access

  • Logging and configuration

Quick Reference

Common workflows:

  1. Generate test data:

    from liberata_metrics.generators import generate_references_matrix

matrices = generate_references_matrix(
    num_manuscripts=100,
    citation_density=0.05
)

  2. Compute portfolio metrics:

    from liberata_metrics.metrics import PortfolioMetrics

pm = PortfolioMetrics(capital_matrix)
returns = pm.compute_returns()
risk = pm.compute_risk()

  3. Analyze system dynamics:

    from liberata_metrics.metrics import MarketMetrics

mm = MarketMetrics(references, capital)
efficiency = mm.compute_efficiency()

Full API Documentation

Complete Reference