The set operations missing from SparkR's RDD API when compared to PySpark are
def intersection(self, other): def glom(self): def cartesian(self, other): def subtractByKey(self, other, numPartitions=None): def subtract(self, other, numPartitions=None):
This will not block the merging, we can continue to work on this after merging.
glom() implemented in https://github.com/amplab-extras/SparkR-pkg/pull/225
intersect implemented in PR 238 and subtract, subtractByKey implemented in PR 244