Interface GraphComputer


  • public interface GraphComputer
    The GraphComputer is responsible for the execution of a VertexProgram and then a set of MapReduce jobs over the vertices in the Graph. It is up to the GraphComputer implementation to determine the appropriate memory structures given the computing substrate.
    Author:
    Marko A. Rodriguez (http://markorodriguez.com), Matthias Broecheler (me@matthiasb.com)
    • Method Detail

      • program

        GraphComputer program​(VertexProgram vertexProgram)
        Set the VertexProgram to be executed by the GraphComputer. There can only be one VertexProgram for the GraphComputer.
        Parameters:
        vertexProgram - the VertexProgram to be executed
        Returns:
        the updated GraphComputer with newly set VertexProgram
      • mapReduce

        GraphComputer mapReduce​(MapReduce mapReduce)
        Add a MapReduce job to the set of MapReduce jobs to be executed by the GraphComputer. There can be any number of MapReduce jobs.
        Parameters:
        mapReduce - the MapReduce job to add to the computation
        Returns:
        the updated GraphComputer with newly added MapReduce job
      • workers

        GraphComputer workers​(int workers)
        Set the desired number of workers to execute the VertexProgram and MapReduce jobs. This is a recommendation to the underlying GraphComputer implementation and is allowed to deviate accordingly by the implementation.
        Parameters:
        workers - the number of workers to execute the submission
        Returns:
        the updated GraphComputer with newly set worker count
      • vertices

        GraphComputer vertices​(Traversal<Vertex,​Vertex> vertexFilter)
                        throws IllegalArgumentException
        Add a filter that will limit which vertices are loaded from the graph source. The provided Traversal can only check the vertex, its vertex properties, and the vertex property properties. The loaded graph will only have those vertices that pass through the provided filter.
        Parameters:
        vertexFilter - the traversal to verify whether or not to load the current vertex
        Returns:
        the updated GraphComputer with newly set vertex filter
        Throws:
        IllegalArgumentException - if the provided traversal attempts to access vertex edges
      • edges

        GraphComputer edges​(Traversal<Vertex,​Edge> edgeFilter)
                     throws IllegalArgumentException
        Add a filter that will limit which edges of the vertices are loaded from the graph source. The provided Traversal can only check the local star graph of the vertex and thus, can not access properties/labels of the adjacent vertices. The vertices of the loaded graph will only have those edges that pass through the provided filter.
        Parameters:
        edgeFilter - the traversal that determines which edges are loaded for each vertex
        Returns:
        the updated GraphComputer with newly set edge filter
        Throws:
        IllegalArgumentException - if the provided traversal attempts to access adjacent vertices
      • vertexProperties

        GraphComputer vertexProperties​(Traversal<Vertex,​? extends Property<?>> vertexPropertyFilter)
        Add a filter that will limit which vertex properties are loaded from the graph source. The loaded vertices will only have those properties that pass through the provided filter. To drop all vertex properties, provide a traversal like __.properties("dummy") where "dummy" is not a valid vertex property.
        Parameters:
        vertexPropertyFilter - the traversal that determines which vertex properties are loaded for each vertex
        Returns:
        the updated GraphComputer with newly set vertex property filter
      • configure

        default GraphComputer configure​(String key,
                                        Object value)
        Set an arbitrary configuration key/value for the underlying Configuration in the GraphComputer. Typically, the other fluent methods in GraphComputer should be used to configure the computation. However, for some custom configuration in the underlying engine, this method should be used. Different GraphComputer implementations will have different key/values and thus, parameters placed here are generally not universal to all GraphComputer implementations. The default implementation simply does nothing and returns the GraphComputer unchanged.
        Parameters:
        key - the key of the configuration
        value - the value of the configuration
        Returns:
        the updated GraphComputer with newly set key/value configuration