README.md

Built-In Algorithms

PGX includes a wide selection of optimized graph algorithms that can be invoked through the Analyst. The following list provides the links to the detail pages of the available algorithms, grouped by category.

• Classic graph algorithms
  • Prim's Algorithm
  • Prim's Algorithm (undirected)
• Community detection
  • Label Propagation
  • Conductance Minimization (Soman and Narang Algorithm)
  • Infomap
  • Louvain
  • Louvain (directed)
  • Speaker Listener Label Propagation
  • Filtered Speaker Listener Label Propagation
  • Filtered Speaker Listener Label Propagation
  • Weighted Speaker Listener Label Propagation
• Connected components
  • Strongly Connected Components (SCC) via Kosaraju's algorithm
  • Strongly Connected Components (SCC) via Tarjan's algorithm
  • Weakly Connected Components (WCC)
• Link prediction
  • WTF (Whom To Follow) Algorithm
• Matrix factorization
  • Matrix Factorization (Gradient Descent)
  • Estimate Rating
• Path finding
  • All Vertices and Edges on Filtered Path
  • All Vertices and Edges on Filtered Path
  • Enumerate Simple Paths
  • Compute High-Degree Vertices
  • Compute Distance Index
  • Fast Path Finding
  • Filtered Fast Path Finding
  • Classic Dijkstra Algorithm
  • Classic Dijkstra Algorithm (ignoring edge directions)
  • Filtered Dijkstra Algorithm
  • Filtered Dijkstra Algorithm (ignoring edge directions)
  • Bidirectional Dijkstra Algorithm
  • Bidirectional Filtered Dijkstra Algorithm
  • Bidirectional Filtered Dijkstra Algorithm
  • Bidirectional Dijkstra Algorithm (ignoring edge directions)
  • Classic Bellman-Ford
  • (Backwards) Bellman-Ford Algorithm
  • Bellman-Ford (Ignoring edge directions)
  • Hop Distance
  • Hop Distance (Backwards)
  • Hop Distance (undirected)
  • Fattest Path
  • Fattest Path (ignoring edge directions)
• Ranking and walking
  • Classic PageRank
  • Approximate PageRank
  • Weighted PageRank
  • Personalized PageRank
  • Personalized PageRank (for a set of vertices)
  • Personalized Weighted PageRank
  • Personalized Weighted PageRank (for a set of vertices)
  • Vertex Betweenness Centrality
  • Approximate Vertex Betweenness Centrality with Random Seeds
  • Approximate Vertex Betweenness Centrality From seeds
  • Closeness Centrality (Unit Length)
  • Closeness Centrality (with weights)
  • Hyperlink-Induced Topic Search (HITS)
  • Eigenvector Centrality
  • Out-Degree Centrality
  • In-Degree Centrality
  • Degree Centrality
  • Random Walk with Restart
  • Harmonic Centrality
  • Article Rank
  • Classic SALSA
  • Personalized SALSA
  • Personalized SALSA (for a set of vertices)
• Structure evaluation
  • Bipartite Check
  • Diameter / Radius
  • Periphery / Center
  • Local Clustering Coefficient (LCC)
  • Local Clustering Coefficient for Directed Graphs (LCC)
  • Local Clustering Coefficient for ignoring edge directions (LCC)
  • Find Cycle
  • Find Cycle from Node
  • Reachability
  • Reachability (undirected)
  • Topological Sort
  • Topological Schedule
  • Triangle Counting
  • Triangle Counting (undirected)
  • Adamic-Adar index
  • Adamic-Adar index (Ignoring edge direction)
  • Conductance
  • Modularity
  • Partition Conductance
  • Out-Degree Distribution
  • In-Degree Distribution
  • K-Core
• Other
  • Filtered BFS (Breadth-First Search)
  • Filtered DFS (Depth-First Search)