Isolate Partitions Based on a Column in Spark

Question

I am looking to perform isolated work within a single partition of a spark DataFrame. More or less, I want to iterate directly over rows within a single partition, without a