R is a popular programming language among statisticians, specially in computational sciences. Of course, a primary reason for the popularity and wide adoption is that it is open-source and freely available --- supported by a great community of developers and consumers.
However, there exist multiple channels that distribute an implementation of R (the language specification) besides the freely available GNU GPL version.
Microsoft R Open is an enhanced distribution of R provided by the Microsoft Corporation. The distribution maintains 100% compatibility with existing base-R distribution, and thus, should run all scripts, packages, and applications that work with base R.
In the enhancements are:
Plus these key enhancements:
- Multi-threaded math libraries that brings multi-threaded computations to R.
- A high-performance default CRAN repository that provide a consistent and static set of packages to all Microsoft R Open users.
- The checkpoint package that make it easy to share R code and replicate results using specific R package versions.
When working on a server, installing two versions of R side-by-side can be
tricky. Often it is desirable to maintain an older/legacy version that you know
works well with certain packages (for reproducibility or old time's sake).
I have found that conda works well for this problem.
For example,
conda create -n mro_env mro-base=3.5.1 r-essentials
will install MRO-3.5.1 along with "essential" R-packages (think tidyverse!). For
using this version of R, simply run source activate mro_env and source deactivate to deactivate.
Happy hacking!
Edit (May 18, 2019)
I have been using MRO for a while now. Unless you are in urgent need of specific features offered by the MRO distribution, I would stick with the base R distributions. The primary reasons are that a) MRO does not catch up with latest R release as fast as I would expect, and b) it is difficult to get support in case of bugs.