Aggregating Kolmogrov Smirnov Test in pyspark

Question

Is there a way to apply the KS test from the spark.mllib library in pyspark using the groupBy clause or some method of aggregation? For example, I have a dataframe