You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
If you have installed multiple Python environments, or use multiple R tools, it is easy to install a package to the wrong library and then not be able to find it later, or get confused as to which environment you are using..
24
+
If you have installed multiple Python environments, or use multiple R tools, it is easy to install a package to the wrong library or environment and then not be able to find it later.
25
25
26
26
This article provides some queries you can use to determine your current version, and to list the packages that are installed in the current SQL Server environment.
For more information about the variable `sys.path`and how it is used to set the interpreter’s search path for modules, see the Python documenttion](https://docs.python.org/2/tutorial/modules.html#the-module-search-path)
241
+
For more information about the variable `sys.path`and how it is used to set the interpreter’s search path for modules, see the [Python documentation](https://docs.python.org/2/tutorial/modules.html#the-module-search-path)
Copy file name to clipboardExpand all lines: docs/advanced-analytics/r/install-additional-r-packages-on-sql-server.md
+3-3Lines changed: 3 additions & 3 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -32,7 +32,7 @@ There are multiple methods for installing new R packages, depending on which ver
32
32
33
33
+[Install new R packages on a server with **no** internet access](#bkmk_offlineInstall)
34
34
35
-
If the server does not have internet access, some additional steps are required to prepare the packages. This topic describes how to prepare files required for installation of the package and its dependencies.
35
+
If the server does not have internet access, some additional steps are required to prepare the packages. This section describes how to prepare files required for installation of the package and its dependencies.
36
36
37
37
+[Install packages using the CREATE EXTERNAL LIBRARY statement](#bkmk_createlibrary)
38
38
@@ -122,7 +122,7 @@ This procedure assumes that you have prepared all the packages that you need, in
This article describes the default libraries for R and Python that are installed with SQL Server.
25
+
This article describes the default libraries for R and Python that are installed with SQL Server.
26
26
27
27
## What is the default instance library
28
28
@@ -68,7 +68,7 @@ This section provides a summary of the R or Python features that are installed b
68
68
69
69
By default the R **base** packages are installed. Base packages include core functionality provided by packages such as `stats` and `utils`.
70
70
71
-
A base installation of R also includes numerous sample datasets, and standard R tools such as RGui (a lightweight interactive editor) and RTerm (a command line tool).
71
+
A base installation of R also includes numerous sample datasets, and standard R tools such as RGui (a lightweight interactive editor) and RTerm (an R command prompt).
72
72
73
73
Installation of R in SQL Server 2016 or SQL Server 2017 also includes the **RevoScaleR** package, and related enhanced packages and providers, which supports remote compute contexts, streaming, parallel execution of rx function, and many other features.
74
74
@@ -106,14 +106,14 @@ The permissions required for package installation have changed between SQL Serve
106
106
107
107
+ In SQL Server 2017, you can continue to install packages as an administrator for both R and Python, and this is probably the easiest method.
108
108
109
-
However, if you enable package management on the instance, you can also use database roles and DDL statements to install R packages at the database level. This feature will be extended to support Python in later releases.
109
+
The DDL statement, CREATE EXTERNAL LIBRARY, allows the database administrator to install packages without using R tools.
110
110
111
-
After this feature has been enabled, a database administrator can grant users the ability to install their own packages on a per database basis. You can also use R commands from a remote client to install packages into the instance library. This feature requires some additional configuration on the instance. for more information, see [Enable package management using DDLs](r-package-how-to-enable-or-disable.md).
111
+
If you use the package management feature for Machine Learning Server, you can use RevoScaleR to install R packages at the database level. The database administrator must enable the feature and then grant users the ability to install their own packages on a per database basis. For more information, see [Enable package management using DDLs](r-package-how-to-enable-or-disable.md).
112
112
113
113
### User libraries are not supported
114
114
115
-
Often users who cannot install a package to a secured location will install a package to a user library. However, this is not possible in the SQL Server environment.
115
+
Users who cannot install a package to a secured location often resort to installing a package to a user library. However, this is not possible in the SQL Server environment.Even file system access is often restricted on the server.
116
116
117
-
Often file system access is restricted on the server, but even if you have admin rights and access to a user document folder on the server, the external script runtime that executes in SQL Server cannot access any packages installed outside the default instance library.
117
+
Even if you have admin rights and access to a user document folder on the server, the external script runtime that executes in SQL Server cannot access any packages installed outside the default instance library.
118
118
119
119
For tips on how to resolve problems related to user libraries, see [Package installed in user libraries](packages-installed-in-user-libraries.md).
Copy file name to clipboardExpand all lines: docs/advanced-analytics/r/packages-installed-in-user-libraries.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -23,7 +23,7 @@ ms.workload: "Inactive"
23
23
24
24
Experienced R users are accustomed to installing R packages in a user library, whenever the default library is blocked or not available. However, this approach is not supported in SQL Server, and installation to a user library usually ends in a “package not found” error.
25
25
26
-
This topic provides workarounds to help you avoid this error. It explains how you can modify your R code, and suggests the correct R package installation process for using R packages from a SQL Server instance.
26
+
This article describes workarounds to help you avoid this error. It explains how you can modify your R code, and suggests the correct R package installation process for using R packages from a SQL Server instance.
27
27
28
28
## Why R user libraries cannot be accessed from SQL Server
Copy file name to clipboardExpand all lines: docs/advanced-analytics/r/r-package-management-for-sql-server-r-services.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -90,4 +90,4 @@ In general, servers that host production databases do not allow connection to th
90
90
91
91
Identifying all dependencies can be complicated. For R, we recommend that you use [miniCRAN](create-a-local-package-repository-using-minicran.md) to prepare an offline package repository.
92
92
93
-
For Python, you must similarly prepare all dependencies and save them locally. Be sure to use a Windows-compatible binaries and use the .WHL format.
93
+
For Python, you must similarly prepare all dependencies and save them locally. Be sure to use a Windows-compatible binaries and use the WHL format.
Copy file name to clipboardExpand all lines: docs/advanced-analytics/r/use-revoscaler-to-manage-r-packages.md
+3-3Lines changed: 3 additions & 3 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -29,7 +29,7 @@ If you have R Server 9.0.1 or later, you can use the [rxInstallPackages](https:/
29
29
30
30
The new version of RevoScaleR also includes these functions:
31
31
32
-
+ The [rxFindPackage](https://docs.microsoft.com/machine-learning-server/r-reference/revoscaler/rxfindpackage)functions gets the path for one or more packages in the specified compute context.
32
+
+ The [rxFindPackage](https://docs.microsoft.com/machine-learning-server/r-reference/revoscaler/rxfindpackage)function gets the path for one or more packages in the specified compute context.
33
33
34
34
You can use a combination of users and scope to find packages or add packages to a particular database:
35
35
@@ -39,7 +39,7 @@ The new version of RevoScaleR also includes these functions:
39
39
40
40
This function checks for dependencies and ensures that any related packages can be installed to SQL Server, just like R package installation in the local compute context.
41
41
42
-
+ The [rxRemovePackages](https://docs.microsoft.com/machine-learning-server/r-reference/revoscaler/rxremovepackages)functions removes packages from a specified compute context.
42
+
+ The [rxRemovePackages](https://docs.microsoft.com/machine-learning-server/r-reference/revoscaler/rxremovepackages)function removes packages from a specified compute context.
43
43
44
44
It also computes dependencies and ensures that packages that are no longer used by other packages on SQL Server are removed, to free up resources.
> If you have enabled the option to see SQL console output, you might also get a lot of status messages from the function that precedes the `print` statement. After you have finished testing your code, set `consoleOutput` to FALSE in the compute context constructor to eliminate messages.
134
+
> If you have enabled the option to see SQL console output, you might get status messages from the function that precedes the `print` statement. After you have finished testing your code, set `consoleOutput` to FALSE in the compute context constructor to eliminate messages.
0 commit comments