A benchmark between an implementation of a hashing algorithm written using a Spark UserDefinedFunction and its equivalent using not but functionality provided through org.apache.spark.sql.functions.
The results indicate that after a warm-up time, the performance is similar.