The big goal of this project seems to be generating Pig Code for Hadoop from Idiomatic Clojure.
This seems quite similar to Parkour (Hadoop MapReduce in idiomatic Clojure)
It is interesting to note that the argument here “this makes writing Pig code so much easier” was also used by Nathan Marz when he was announcing Cascalog three years ago:
The main distinguishing feature of this library compared to the other two seems to be that this one actually generates Pig code (although not for human consumption) as opposed to calling Hadoop or Cascading libraries so the benefit is ‘runs where Pig runs’).
The big goal of this project seems to be generating Pig Code for Hadoop from Idiomatic Clojure. This seems quite similar to Parkour (Hadoop MapReduce in idiomatic Clojure)
https://github.com/damballa/parkour/
The other Elephant in the room is Cascalog (generating Cascading Mapreduce queries for Hadoop using a Datalog language)
http://cascalog.org/
It is interesting to note that the argument here “this makes writing Pig code so much easier” was also used by Nathan Marz when he was announcing Cascalog three years ago:
http://www.slideshare.net/nathanmarz/clojure-at-backtype (at slide 42)
The main distinguishing feature of this library compared to the other two seems to be that this one actually generates Pig code (although not for human consumption) as opposed to calling Hadoop or Cascading libraries so the benefit is ‘runs where Pig runs’).
Form your own opinion of course.