Under the hood of GraphFrames are two Spark DataFrames: one for the vertices and other one for the edges. GraphFrames might be thought of as the next generation of Spark's GraphX library, with some major improvements over the latter:
- GraphFrames leverages the performance optimizations and simplicity of the DataFrame API.
- By using the DataFrame API, GraphFrames can be interacted with through Python, Java, and Scala APIs. In contrast, GraphX was only available through the Scala interface.
You can find the latest information on GraphFrames within the GraphFrames overview at https://graphframes.github.io/.Â
