Package com.luciad.network.graph.numeric

Contains classes for working with numeric graphs.

See: Description

Interface Summary

Interface	Description
ILcdNumericGraphMappingHandler<N,E>	Callback interface for reporting the mappings between graphs and numeric graphs.

Class Summary

Class	Description
TLcdNumericGraph	A graph implementation whose nodes, edges and partitions are all represented by a numerical identifer.
TLcdNumericGraphDecoder	Decoder for decoding numeric graphs, encoded by the TLcdNumericGraphEncoder.
TLcdNumericGraphEncoder	Encoder for exporting graphs to numeric graphs and saving numeric graphs after they have been modified.

Enum Summary

Enum	Description
TLcdNumericGraphEncoder.EdgeValueEncoding	Enumeration describing the encoding of edge values.

Package com.luciad.network.graph.numeric Description

Contains classes for working with numeric graphs.

Numeric graphs are graphs in which the nodes and edges are all numeric values. Numeric graphs in LuciadLightspeed are built in such a way that the edge or node itself (the numeric value) describes in which (list of) partition(s) it is included. Numeric graphs have several advantages over other graphs:

• They are very compact in memory (they are represented by a single long value).
• They can be stored on disk even more compact.
• They are very fast to query (partition containment, connectivity, edge values).
• They are very well suited for on-the-fly loading.

Due to their superior performance, numeric graphs should be considered whenever graphs become too large to be contained completely in memory, or performance of algorithms applied on them is not sufficient.

Using numeric graphs has the following limitations:

• Only zeroth-order edge value functions are supported.

This package is part of the Graph & Routing Engine component.

Since:

9.1