Skip to main content

Machine Learning Components

Run ML Notebook Component

Note

The "Run ML Notebook" component allows you to trigger a Jupyter Notebook from inside the Analysis UI. This allows you to extend your Analysis with custom calculations, e.g. running an automated Root Cause Analysis, where the calculations happen in Python and the results are shown to the Business User in the Analysis UI.

35555710.png

Settings

After dragging the component into the canvas, select the name of the Workbench and the Notebook inside this Workbench that you want to run. After that, hit the run button on the component and the Notebook is executed.

Passing Parameters

35555713.png

When you press the run button, the component is passing a couple of parameters to the Notebook which can be used in variables in Python. The parameters are:

  • variables: Passed as JSON string. Note: When passing PQL as a variable, you need to evaluate it on the Python side with pycelonis. Refer to the pycelonis documentation here.

  • sheetId: This is the identifier of the Sheet where the component was executed from.

  • dataModel_id: This is the identifier of the Datamodel that is used in the Analysis.

  • queries: These are the user selections (filters) that are applied at the time the execution was started.

  • analysis_id: The identifier of the Analysis.

Using Parameters in a Jupyter Notebook

35555711.png

When developing your Python code, you can define default values for the parameters. To do this, tag the first cell of your Notebook with the Tag "parameters" and set default values in this cell. While developing and running the code from Jupyter, these values will be used. When you are calling the Notebook from the ML component, the default values will be overwritten by the ones passed by the component.