Dataiku DSSInstall

Overview

This node integrates with the Dataiku DSS API to manage and interact with various resources within a Dataiku project. Specifically, for the Machine Learning - Saved Model resource and the Get Partial Dependencies of Version operation, it retrieves all partial dependencies computed for a specific version of a saved machine learning model.

This functionality is useful when you want to analyze how different features influence the predictions of a particular model version by examining its partial dependency plots or data. It helps in interpreting model behavior and understanding feature impact.

Practical example:
You have a saved ML model version deployed in your Dataiku project, and you want to fetch the precomputed partial dependencies to visualize or further analyze them in your workflow. This node will call the appropriate API endpoint to retrieve that information.

Properties

Output

The output is a JSON object representing the partial dependencies data computed for the specified saved model version. This typically includes feature names and their corresponding partial dependency values, allowing users to understand the effect of each feature on the model's predictions.

If the API returns binary data (not typical for this operation), it would be provided as binary output, but for this operation, the output is JSON.

Dependencies

• Requires an active connection to a Dataiku DSS instance.
• Requires valid API credentials (an API key) for authentication with the Dataiku DSS API.
• The node expects the Dataiku DSS server URL and user API key to be configured in the credentials.
• The project must contain the saved model and version specified.

Troubleshooting

• Missing Credentials Error: If the API key or server URL is not set, the node will throw an error indicating missing credentials. Ensure credentials are properly configured.
• Missing Required Parameters: The node validates required parameters such as Project Key, Save Model ID, and Version ID. Omitting any of these will cause an error. Provide all required inputs.
• API Errors: If the Dataiku DSS API returns an error (e.g., model/version not found, permission denied), the node will throw an error with the message from the API. Verify the existence of the model/version and your access rights.
• Network Issues: Connectivity problems to the Dataiku DSS server will result in request failures. Check network connectivity and server availability.

Links and References

• Dataiku DSS API Documentation (general reference for API endpoints)
• Partial Dependence Explanation (conceptual background)

This summary focuses on the "Machine Learning - Saved Model" resource and the "Get Partial Dependencies of Version" operation as requested.