How to configure Azure Machine Learning Workbench to work with an IDE

Azure Machine Learning Workbench can be configured to work with popular Python IDEs (Integrated Development Environment). It enables a smooth data science development experience moving between data preparation, code authoring, run tracking and operationalization. Currently the supported IDEs are:

  • Microsoft Visual Studio Code
  • JetBrain PyCharm

Configure workbench

  1. Click on the File menu in the top left corner of the app.
  2. Select the Configure Project IDE option from the flyout
  3. Type in VS Code or PyCharm in the Name field (the name is arbitrary)
  4. Enter the location to the IDE executable (complete with the executable name and extension) in Execution Path

Default install path for Visual Studio Code

  • Windows 32-bit - C:\Program Files (x86)\Microsoft VS Code\Code.exe
  • Windows 64-bit - C:\Program Files\Microsoft VS Code\Code.exe
  • macOS - select the .app path, for example /Applications/Visual Studio Code.app, and the app appends the rest of the path for you. The full path to the executable by default is /Applications/Visual Studio Code.app/Contents/Resources/app/bin/code. If you have executed the Shell Command: Install 'code' command in PATH command in VS Code, then you can also reference the VS Code script at /usr/local/bin/code

Default install path for PyCharm

  • Windows 32-bit - C:\Program Files (x86)\JetBrains\PyCharm Community Edition 2017.2.1\bin\pycharm.exe.
  • Windows 64-bit - C:\Program Files\JetBrains\PyCharm Community Edition 2017.2.1\bin\pycharm64.exe.
  • macOS - select the .app path, for example “/Applications/PyCharm CE.app”, and the app appends the rest of the path for you. The full path to the executable by default is /Applications/PyCharm CE.app/Contents/MacOS/pycharm. You may also find PyCharm at the bin folder, /usr/local/bin/charm

Open project in IDE

Once the configuration is complete, you can open an Azure Machine Learning project by opening the File menu in Azure Machine Learning Workbench, then click Open Project (<IDE_Name>). This action opens the current active project in the configured IDE. Note: If you are not in a project, the Open Project (<IDE_Name>) will be disabled.

Configuring the integrated terminal in Visual Studio Code

Windows

We have overridden the default shell to be cmd instead of PowerShell. On clicking on Open Project (<IDE_Name>), you see a prompt:

Do you allow shell: C:\windows\System32\cmd.exe (defined as a workspace setting) to be launched in the terminal?

Answer yes to allow configuring the shell to work seamlessly with Azure ML Workbench command-line interface.

Mac

To run an az command using Visual Studio Code's integrated terminal on Mac, you need to manually set the PATH to be the same value as PATH in the project's .vscode/settings.json file under the key terminal.integrated.env.osx. You can do so by running the following command in the terminal: PATH=<PATH in .vscode/settings>