Efficient Way to do global ranking on a large amount of data in Pyspark

Question

I am trying to do global rank on a large amount of data using Pyspark.

I searched online and found many answers used window:

window_rank = Window.orderBy(F