What's new in `plyrmr` 0.4.0

Features

system of backends a là rmr2: use "local" for learning and debugging, no Hadoop needed; use "hadoop" for real work on a cluster; don't use "spark" yet as it's not ready for prime time, but we are working on it.
system of options also inspired by rmr2: with plyrmr.options you can set the backend or any other option that's relevant to the current backend. Right now that just means that arguments other than backend will be forwarded to rmr.options, but in the future that will change (as the spark backend is independent of rmr2)
Support for fast aggregation using vectorized operations and the dplyr extension.
Special variable to use in connection with %|% operator renamed .. to avoid clashes with . in formulas.