Languages supported on the Data Science Virtual Machine

To build artificial intelligence (AI) applications, the Data Science Virtual Machine (DSVM) includes with several prebuilt languages and development tools:

Python

Category Value
Language versions supported Python 3.8
Supported DSVM editions Windows Server 2019, Linux
How is it configured and installed on the DSVM? Multiple conda environments include different, preinstalled Python packages. Run conda env list to list all available environments in your machine.

How to use and run it

  • At a command prompt:

    Use one of these methods, depending on the version of Python you want to run:

    conda activate <conda_environment_name>
    python --version
    
  • Use in an IDE:

    The DSVM images have several IDEs installed - for example, VS.Code or PyCharm. You can use them to edit, run, and debug your Python scripts.

  • Use in Jupyter Lab:

    Open a Launcher tab in Jupyter Lab, and select the type and kernel of your new document. To place your document in a specific folder, first navigate to that folder in the File Browser on the left side.

  • Install Python packages:

    To install a new package, you must first activate the proper environment. The proper environment is the location where your new package will be installed. The package will then only become available in that environment.

    To activate an environment, run conda activate <environment_name>. Once the environment is activated, you can use a package manager -for example, conda or pip - to install or update a package.

    As an alternative, if you use Jupyter, you can also run !pip install --upgrade <package_name> in a cell to install packages directly.

R

Category Value
Language versions supported CRAN R 4.0.5
Supported DSVM editions Linux, Windows

How to use and run it

  • Run at a command prompt:

    Open a command prompt and type R.

  • Use in Jupyter Lab

    Open a Launcher tab in Jupyter Lab, and select the type and kernel of your new document. To place your document in a specific folder, first navigate to that folder in the File Browser on the left side.

  • Install R packages:

    You can install new R packages with the install.packages() function.

Julia

Category Value
Language versions supported 1.0.5
Supported DSVM editions Linux, Windows

How to use and run it

  • Run at a command prompt

    Open a command prompt and run julia.

  • Use in Jupyter:

    Open a Launcher tab in Jupyter Lab, and select the type and kernel of your new document. To place your document in a specific folder, first navigate to that folder in the File Browser on the left side.

  • Install Julia packages:

    You can install or update packages with Julia package manager commands like Pkg.add().

Other languages

C#: Available on Windows and accessible through the Visual Studio Community edition. You can also run the csc command at the Developer Command Prompt for Visual Studio.

Java: OpenJDK is available on both the Linux and Windows DSVM editions. It's set on the path. To use Java, type the javac or java command at a command prompt in Windows, or on the bash shell in Linux.

Node.js: Node.js is available on both the Linux and Windows editions of the DSVM. It's set on the path. To access Node.js, type the node or npm command at a Windows command prompt or in a Linux Bash shell. On Windows, the Visual Studio extension for the Node.js tools is installed. It provides a graphical IDE for Node.js application development.

F#: Available on Windows and accessible through the Visual Studio Community edition or at a Developer Command Prompt for Visual Studio, where you can run the fsc command.