User Guide

These guides explain how to use nx-backbone effectively.

• Workflow
  • The score-then-filter pattern
  • Score attributes by method
  • Filtering functions
  • Matrices and edge lists
  • Visual comparison
  • Methods with boolean keep flags
  • Method gallery
• Choosing a Method
  • Decision guide
  • Method comparison table
  • Choosing a statistical method
• Directed Graphs
  • Supported methods
  • Undirected-only methods
  • Behavior differences
  • Example
• Weights, Units, and Transforms
  • Transforming weights
  • Which methods a transform changes
  • Count units
  • The disparity filter’s null model
  • Not offered
• Temporal networks: formats and conversions
  • The three forms
  • Snapshots to events, and back
  • Snapshots to spans, and back
  • Events to snapshots: binning
  • Spans to snapshots
  • Events to spans
  • Spans to events
  • One function for all directions
  • Collapsing to a static weighted graph
  • Datetimes and pandas
  • What each conversion keeps
• Evaluating Backbones
  • Basic fraction metrics
  • Connectivity
  • Distribution preservation
  • Comparing multiple backbones
  • Reporting a backbone
• Complexity Summary
  • Statistical
  • Structural
  • Proximity
  • Hybrid
  • Bipartite
  • Unweighted
  • Filters
  • Measures
  • Visualization
  • Convert
  • Temporal
• Backends
  • What is dispatchable
  • Using a backend
  • Measured timings
  • Testing with a backend