Pyspark + Graphframes: “recursive” message aggregation

Question

I\'ve created the following graph:

spark = SparkSession.builder.appName(\'aggregate\').getOrCreate()
vertices = spark.createDataFrame([(\'1\', \'foo\', 99),