PREDICT FROM SAVED MATHMODEL

Loads a previously saved math model (ML model persisted in the d3VIEW platform) and runs inference against a supplied dataset, returning predictions alongside LUCY ML diagnostics. Use this worker whenever a trained model needs to be applied to new input data within a workflow.

When to use

Tagged: inference, lucy, mathmodel, ml, predict, scoring.

Inputs

Label ID Type Default Required Description
Mathmodel mathmodel_id remote_lookup   Reference to the saved math model to use for inference; selected via the remote lookup widget keyed to ‘remote_lookup_mathmodels’ — leave null to resolve dynamically from an upstream worker.
Dataset dataset dataset Input dataset whose rows will be scored; column schema must match the feature schema the math model was trained on (validated at runtime against mathmodel_id).

Outputs

Label ID Type Description
Mathmodel mathmodel_id integer Integer identifier of the math model that was used for prediction, passed through for downstream traceability.
LUCY JSON lucy_json json Raw LUCY ML JSON response containing stdout, diagnostics, and the full inference payload; useful for debugging or extracting additional model metadata.
Meta Info meta_info dataset Dataset containing auxiliary metadata returned by the LUCY ML engine (e.g., model version, run statistics, feature importances if available).
Predicted Dataset dataset dataset Predicted dataset where each row from the input is annotated with its model-predicted value(s); row IDs are preserved from the input dataset.

Disciplines

  • ai_ml.supervised.classification
  • ai_ml.supervised.regression
  • ai_ml.surrogate

Auto-generated from platform schema. Worker id: mathmodel_predict. Schema hash: 256726b51dac. Hand-curated docs in workerexamples/ override this page when present.