Chapter 5. Creating Custom Graph Aggregation Operators
In the previous chapter, we have seen various operations for transforming the elements of a graph and for modifying its structure. Here, we will learn to use a generic and powerful operator named
aggregateMessages that is useful for aggregating the neighborhood information of all nodes in the graph. In fact, many graph-processing algorithms rely on iteratively accessing the properties of neighboring nodes and adjacent edges. One such example is the PageRank algorithm.
aggregateMessages to the NCAA College Basketball datasets, you will be able to:
- Understand the basic mechanisms and patterns of
- Apply it to create custom graph aggregation operations
- Optimize the performance ...