ExperimentalExperimentalCreates an instance of the RPT client.
OptionalmodelDeployment: ModelDeployment<SapRptModel>This configuration is used to retrieve a deployment. Depending on the configuration use either the given deployment ID or the model name to retrieve matching deployments. If model and deployment ID are given, the model is verified against the deployment.
Optionaldestination: The destination to use for the request.
OptionalserviceBindingTransformFn?: ServiceBindingTransformFunctionCustom transformation function to control how a Destination is built from the given Service.
The name of the destination to retrieve from service bindings.
Optionalservice?: undefinedExperimentalPredict based on Parquet file data. Parquet is a binary tabular data format with typed columns.
Parquet data and prediction configuration to base prediction on.
OptionalrequestConfig: Omit<RptRequestOptions, "compress">Custom request configuration.
Prediction response.
ExperimentalPredict based on prediction data with data schema inferred.
Prefer using predictWithSchema when the data schema is known.
Data to base prediction on.
OptionalrequestConfig: RptRequestOptionsCustom request configuration.
Prediction response.
ExperimentalPredict based on data schema and prediction data. Prefer using this method when the data schema is known.
Prediction data follows this schema. When using TypeScript, the data schema type is used to infer the types of the prediction data. In that case, the data schema must be provided as a constant (as const).
Data to base prediction on.
OptionalrequestConfig: RptRequestOptionsCustom request configuration.
Prediction response.
Copyright Ⓒ 2026 SAP SE or an SAP affiliate company. All rights reserved.
Representation of an RPT client to make predictions. This class is experimental and may change at any time without prior notice.