The scripting architecture allows for custom R or Python scripts to be wrapped as XML files describing the inputs and outputs.  These scripts can be loaded into / registered with Oracle Analytics and are then callable like any other standard analytic function through a generic EVALUATE_SCRIPT ("Scriptname", "parameter name and parameter value list") function.

Oracle Analytics Desktop allows you to format custom Python and R Scripts for use. 

To use Python in Oracle Analytics Desktop, you need the script to be embedded in a simple pre-defined XML format.

Once formatted in XML, upload the custom script into Oracle Analytics Desktop by clicking Create --> Script

• Oracle Analytics Desktop - Formatting Custom Python Scripts: https://www.youtube.com/watch?v=9ci3h54SuN0
• Oracle Analytics Desktop - Running Custom Python Scripts in Data Flows: https://www.youtube.com/watch?v=KSZWsDBog_k
• OAC: how to enable Loading of Custom Scripts on your OAC environment: https://www.youtube.com/watch?v=5rIYSDiC4IY

Invoke the script from a Data Flow. Add the "Apply Custom Script" node to the dataflow and select your script. 

Uploaded custom scripts can be found in the Machine Learning tab and can be invoked via Data Flows or using calculations within the project.