snorkelflow.client.workflows.execute_prediction_api_async

snorkelflow.client.workflows.execute_prediction_api_async(client, df, return_outputs=False)

Execute an exported workflow running via Prediction API on a new Pandas DataFrame.

It is possible to configure your Prediction API service to provide support for asynchronous inference requests in addition to the standard synchronous requests. If your service is configured to provide this functionality, you can use this method to make a request to the asynchronous endpoint.

Parameters

Parameters

Name	Type	Default	Info
client	HTTPClient		SnorkelFlowPredictionAPIClient object.
df	DataFrame		Data over which to execute the workflow.
return_outputs	bool	False	Whether to return workflow output as a dictionary of output names to dataframes. If False, assumes there is only one output in the workflow and directly returns the corresponding the dataframe.

Returns

Returns

Dictionary mapping outputs in Workflow DAG to corresponding Pandas DataFrame. If return_outputs is False, assumes there is only 1 output, and returns the corresponding dataframe.

Return type

Return type

Dict(str, pandas.DataFrame) or pandas.DataFrame