Introducing plyrmr

[Antonio Piccolboni]

Hi,

this is to announce a preview-level availability of a new package, plyrmr. It builds on top of rmr2 to provide an easier, data frame focused approach to writing mapreduce programs. An expression in plyrmr may look like this:

```

output(transform(input("big/data/set/"), ratio = col1/col2), "some/directory")

```

As the name suggests, this work is inspired by `plyr` (many thanks to Hadley for fruitful discussions) but combines other types of data frame processing from other packages and some of its own. Head over to our wiki for details. We are releasing this early to get feedback to help us improve the package. That means that not everything is in place, in particular you need to build plyrmr from source and rmr2 from the dev branch to get things going. It's all explained in the wiki, but just so that you don't get disappointed.  There is a tutorial and also the help function will return useful info, but still a little unpolished. A first release is in the works. Please kick the tires and let us know what you think. Post your examples, questions and what not here. For issues that require a resolution you can also use the plyrmr issue tracker. Thanks

Antonio