Welcome to clusterxplain’s documentation!

clusterxplain is a Python library meant to offer utilites to calculate feature importance values for clustering results. It is based on the paper (TBD).

Contents

• Installation
• Quickstart
  • Imports
  • Fit a k-means clustering
  • Explaining the Clustering
• Explainers
  • XKM (eXplainable K-Medoids)
  • NEON (neuralization-propagation)
  • Tree-based explanation_results
  • Shapley values
  • Gradient-based explanation_results
• Explanation Results
  • GlobalExplainedClustering
  • ClusterExplainedClustering
  • PointwiseExplainedClustering
  • ExplainedClustering
  • BaseExplainer
• Available Metrics
  • Metric
  • ManhattenMetric
  • EuclideanMetric
  • get_distance_metric()
• Utilities
  • get_resource_string()
• Custom Errors
  • XkmError
  • NotFittedError
  • NonExistingRelevanceError
  • NonExsitingXkmFlavourError
  • InconsistentNamingError
  • MetricError
  • MetricNotImplementedError
• Indices and tables
• License