plyrmr version 0.6.0

• Completed Spark support. Passes all tests! Speed needs more work.
• Added spark I/O, a few I/O formats and compatibility with SparkR
  • e.g use input(path, "csv") to read a csv file, output(expression, path, "json") to write in json
  • use as.RDD to get an RDD out of a plyrmr pipe and work in sparkR from there onwards, conversely input(rdd) to go from sparkR to plyrmr
• tests ported to latest Quickcheck 3.0.0
• added support for specifying groups by character vector, for programming