Group by takes too long with Spark y Pyspark

Question

I need your support I have the following code fragment where the delay occurs and I think the "groupBy" does not work for large dataframes:

\'\'\'va