DSS and Python

DSS includes deep integration with Python. In many parts of DSS, you can write Python code:

• In recipes
• In Jupyter notebooks
• In standard webapp backends
• In scenarios, metrics and checks
• In plugins
• For custom models in visual ML
• In API node, for custom prediction models or custom functions endpoints

Any Python package may be used in DSS.

In addition, DSS features a complete Python API, which has its own complete documentation.

The following highlights how a few specific Python packages can be used in DSS. DSS features advanced integration with most of the packages described below.

• Installing Python packages
  • Installing in a specific code environment (recommended)
  • Installing in the builtin DSS environment (not recommended)
    • Installing a package available through pip
    • Installing without Internet access
    • Installing a “python setup.py install” package
  • Installing custom Python packages
• Reusing Python code
  • Sharing Python code within a project
    • Permissions
  • Sharing Python code globally
    • Permissions
    • Caution
  • Manually edition of code library folders
  • Packaging code as plugins
• Using Matplotlib
• Using Bokeh
  • Installing Bokeh
  • Creating a Bokeh interactive web application
  • Displaying charts in a Jupyter notebook
    • Using interactive controls in the Jupyter notebook
  • Displaying Bokeh charts on a dashboard
    • Refreshing charts on a dashboard
• Using Plot.ly
  • Installing plot.ly
  • Displaying charts in a Jupyter notebook
  • Displaying charts on a dashboard
    • Refreshing charts on a dashboard
• Using Ggplot