Поделиться через

R Function Library Reference

  • Статья

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.

Library details Description
Supported platforms Machine Learning Server 9.x
Microsoft R Client (Windows and Linux)
Microsoft R Server 9.1 and earlier
SQL Server 2016 and later (Windows only)
Azure HDInsight
Azure Data Science Virtual Machines
Built on: R 3.5.2 (included when you install a product that provides this package).

R function libraries

Package Version Description
MicrosoftML 1.5.0 A collection of functions in Microsoft R used for machine learning operations, including training and transformations, scoring, text and image analysis, and feature extraction for deriving values from existing data.
mrsdeploy 1.1.2 Deployment functions for interactive remote execution at the command line, plus web service functions for bundling R code blocks as discrete web services that can be deployed and managed on an R Server instance. Formally known as DeployR.
olapR 1.0.0 A collection of functions for constructing MDX queries against an OLAP cube. Runs only on the Windows platform.
RevoIOQ 8.0.7 Installation and Operational Qualification test functions, used in conjunction with the RUnit package to run a set of unit tests. It has only one user-facing function, also called RevoIOQ.
RevoMods 11.0.0 Microsoft modifications and extensions to standard R functions. Reference documentation is online only.
RevoPemaR 10.0.0 Developer functions for coding custom parallel external memory algorithms.
RevoScaleR 9.4 Data acquisition, manipulation and transformations, visualization, and analysis. RevoScaleR provides functions for the full range of statistical and analytical tasks. It's the backbone of R Server functionality.
RevoTreeView 10.0.0 Decision tree functions, including the rxDTree function. Reference documentation is online only.
RevoUtils 10.0.4 Utility functions useful when programming and developing R packages.
RevoUtilsMath 10.0.0 Microsoft's distribution of the Intel Math Kernel Library (MKL). Reference documentation is online only.
sqlrutils 1.0.0 A collection of functions for executing stored procedures against SQL Server.

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:

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:

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:

See also

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