Class MultiLabelFeatureFunction

    • Constructor Summary

      Constructors 
      Constructor Description
      MultiLabelFeatureFunction​(java.util.Map<org.neo4j.gds.NodeLabel,​org.neo4j.gds.ml.core.functions.Weights<org.neo4j.gds.ml.core.tensor.Matrix>> weightsByLabel, int projectedFeatureDimension)  
    • Method Summary

      All Methods Instance Methods Concrete Methods 
      Modifier and Type Method Description
      org.neo4j.gds.ml.core.Variable<org.neo4j.gds.ml.core.tensor.Matrix> apply​(org.neo4j.gds.api.Graph graph, long[] nodeIds, org.neo4j.gds.core.utils.paged.HugeObjectArray<double[]> features)
      This method expects the graph to be validated beforehand, such that each node has exactly one label See feature initialization in GraphSageHelper.
      int projectedFeatureDimension()  
      java.util.Map<org.neo4j.gds.NodeLabel,​org.neo4j.gds.ml.core.functions.Weights<org.neo4j.gds.ml.core.tensor.Matrix>> weightsByLabel()  
      • Methods inherited from class java.lang.Object

        clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
    • Constructor Detail

      • MultiLabelFeatureFunction

        public MultiLabelFeatureFunction​(java.util.Map<org.neo4j.gds.NodeLabel,​org.neo4j.gds.ml.core.functions.Weights<org.neo4j.gds.ml.core.tensor.Matrix>> weightsByLabel,
                                         int projectedFeatureDimension)
    • Method Detail

      • weightsByLabel

        public java.util.Map<org.neo4j.gds.NodeLabel,​org.neo4j.gds.ml.core.functions.Weights<org.neo4j.gds.ml.core.tensor.Matrix>> weightsByLabel()
      • apply

        public org.neo4j.gds.ml.core.Variable<org.neo4j.gds.ml.core.tensor.Matrix> apply​(org.neo4j.gds.api.Graph graph,
                                                                                         long[] nodeIds,
                                                                                         org.neo4j.gds.core.utils.paged.HugeObjectArray<double[]> features)
        This method expects the graph to be validated beforehand, such that each node has exactly one label See feature initialization in GraphSageHelper.
        Specified by:
        apply in interface FeatureFunction
        Parameters:
        nodeIds - batch of node IDs
        features - the global property array
        Returns:
        Create a matrix variable around a batch of nodes.
      • projectedFeatureDimension

        public int projectedFeatureDimension()