Pyspark GroupBy and agg by multiple columns performance

Question

I have a large DataFrame with approximately 30,000 rows and 6 columns "A","B","C","D","E","F". I would like to gr