Class DedupedEdgesGraph<T>

Type parameters

T: DataType = any

Hierarchy

Graph<T>
- DedupedEdgesGraph

Index

Accessors

Methods

Accessors

edgeIds

get edgeIds(): DataFrame<{ dst: Int32; id: Int32; src: Int32 }>

Inherited from Graph.edgeIds
- Defined in modules/cugraph/src/graph.ts:92
Returns DataFrame<{ dst: Int32; id: Int32; src: Int32 }>

edges

get edges(): DataFrame<{ dst: T; id: Int32; src: T; weight: Float32 }>

Inherited from Graph.edges
- Defined in modules/cugraph/src/graph.ts:73
Returns DataFrame<{ dst: T; id: Int32; src: T; weight: Float32 }>

nodeIds

get nodeIds(): DataFrame<{ id: Int32 }>

Inherited from Graph.nodeIds
- Defined in modules/cugraph/src/graph.ts:90
Returns DataFrame<{ id: Int32 }>

nodes

get nodes(): DataFrame<{ node: T }>

Inherited from Graph.nodes
- Defined in modules/cugraph/src/graph.ts:71
Returns DataFrame<{ node: T }>

numEdges

get numEdges(): number

Inherited from Graph.numEdges
- Defined in modules/cugraph/src/graph.ts:64
summary
The number of edges in this Graph

Returns number

numNodes

get numNodes(): number

Inherited from Graph.numNodes
- Defined in modules/cugraph/src/graph.ts:69
summary
The number of nodes in this Graph

Returns number

Methods

analyzeClustering

analyzeClustering(options: AnalyzeClusteringOptions): number

- Defined in modules/cugraph/src/graph.ts:217
summary
Compute a score for a given partitioning/clustering. The assumption is that options.clustering is the results from a call to computeClusters and contains columns named vertex and cluster.

Parameters
- options: AnalyzeClusteringOptions
Returns number
The computed clustering score

computeClusters

computeClusters(options: ClusteringOptions): DataFrame<Object>

- Defined in modules/cugraph/src/graph.ts:189
summary
Compute a clustering/partitioning of this graph using either the spectral balanced cut method, or the spectral modularity maximization method.

see
https://en.wikipedia.org/wiki/Cluster_analysis

see
https://en.wikipedia.org/wiki/Spectral_clustering

Parameters
- options: ClusteringOptions
  
  Options for the clustering method
Returns DataFrame<Object>

dedupeEdges

dedupeEdges(): DedupedEdgesGraph<Series<T>["type"]>

Inherited from Graph.dedupeEdges
- Defined in modules/cugraph/src/graph.ts:94
Returns DedupedEdgesGraph<Series<T>["type"]>

degree

degree(): DataFrame<Object>

Inherited from Graph.degree
- Defined in modules/cugraph/src/graph.ts:104
summary
Compute the total number of edges incident to a vertex (both in and out edges).

Returns DataFrame<Object>

forceAtlas2

forceAtlas2(options?: ForceAtlas2Options<any>): Float32Buffer

Inherited from Graph.forceAtlas2
- Defined in modules/cugraph/src/graph.ts:117
summary
ForceAtlas2 is a continuous graph layout algorithm for handy network visualization.

note
Peak memory allocation occurs at 30*V.

Parameters
- options: ForceAtlas2Options<any> = ...
Returns Float32Buffer
The new positions.

Static fromEdgeList

fromEdgeList<T>(src: T, dst: T, weights?: Float32Series, options?: GraphOptions): DedupedEdgesGraph<T["type"]>

Overrides Graph.fromEdgeList
- Defined in modules/cugraph/src/graph.ts:154
Type parameters
- T: Series<DataType>
Parameters
- src: T
- dst: T
- weights: Float32Series = ...
- options: GraphOptions = ...
Returns DedupedEdgesGraph<T["type"]>