R Function Library Reference

This section contains the R reference documentation for proprietary packages from Microsoft used for data science and machine learning on premises and at scale.

You can use these libraries and functions in combination with other open-source or third-party packages, but to use the revo packages, your R code must run against a service or on a computer that provides the interpreters.

R function libraries

How to get packages

The packages documented in this section are found only on installations of the Microsoft products or Azure services that provide them. Setup programs or scripts install the proprietary R packages from Microsoft and any package dependencies. Unless noted otherwise, all of the packages listed in the preceding table are installed with the product or service.

By default, packages are installed in the \Program Files\Microsoft\ML Server\R_SERVER\library folder on Windows, and in the /opt/microsoft/mlserver/9.2.1 folder on the Linux native file system.

Ships in:

• Machine Learning Server
• SQL Server 2017 Machine Learning Services (Windows only)
• SQL Server Machine Learning Server (Standalone)
• Microsoft R Client (Windows and Linux)
• Microsoft R Server 9.1 and earlier
• Azure HDInsight
• Azure Data Science Virtual Machines

How to list packages and versions

To get the version of an R package installed on your computer, open an R console application and execute the following command: installed.packages()

To determine the version of the package, open its help page to view the version number just under the title: library(help="<package-name>")

How to list functions in a package

To list all of the functions in a package, execute the following command: ls("package:RevoScaleR")

To search for a function by full or partial string matching, add a pattern. For example, to return all functions that include the term "Spark":

> ls("package:RevoScaleR", pattern = "Spark")
[1] "rxGetSparklyrConnection" "RxSpark"                 "rxSparkCacheData"        "rxSparkConnect"          "rxSparkDisconnect"      
[6] "rxSparkListData"         "rxSparkRemoveData"      

How to view built-in help pages

Most R packages come with built-in help pages that open in separate window.

1. Open an R console tool, such as Rgui.exe or another R IDE.
2. List the packages using installed.packages()
3. Open help for a specific package using: library(help="<package-name>")
4. Open help for a specific function using: ?<function_name>

R is case-sensitive. Be sure to type the package name using the correct case: for example, library(help = "RevoScaleR").

Deprecated & discontinued packages

The following packages exist for backward compatibility but are no longer under active development:

• RevoRpeConnector
• RevoRsrConnector
• revolpe

For a list of deprecated or discontinued functions within an existing package, see Deprecated, discontinued, or changed features.

Next steps

First, read the introduction to each package to learn about common use case scenarios:

• MicrosoftML
• mrsdeploy
• olapR
• RevoPemaR
• RevoScaleR
• RevoUtils
• sqlrutils

Next, add these packages by installing R Client or Machine Learning Server. R Client is free. Machine Learning Server developer edition is also free.

Lastly, follow these tutorials for hands-on experience:

• Explore R and RevoScaleR in 25 functions
• Quickstart: Run R Code in Machine Learning Server

See also

How-to guides in Machine Learning Server
Machine Learning Server
Additional learning resources and sample datasets