Abstract
When we execute Apache Spark code from .NET, we are calling methods and classes in the Java Virtual Machine, and Apache Spark reads, writes, aggregates, and transforms our data, according to our requirements. It is entirely possible and quite common that the .NET application never has any sight of the actual data, and the JVM handles all of the data modifications. This is fine if Apache Spark has all of the classes and methods you need to complete your processing. However, what do we do when we need to do something that isn't supported by Apache Spark? The answer is User-Defined Functions (UDFs) and User-Defined Aggregate Functions (UDAFs). UDFs and UDAFs allow us to bring the data back into .NET and then run any processing that we can dream of in .NET.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
