Predict

fyt.commands.predict

Batch prediction command: load a saved pipeline and score new data.

read_table(path)

Read a csv or parquet file by extension.

Parameters:

Name	Type	Description	Default
path	Path	Input file path.	required

Returns:

Type	Description
DataFrame	The loaded DataFrame.

Raises:

Type	Description
ValueError	For unsupported file extensions.

run_prediction(config)

Score a data file with a saved pipeline and write the results.

Parameters:

Name	Type	Description	Default
config	PredictCommandConfig	Prediction configuration.	required

Returns:

Type	Description
DataFrame	The DataFrame that was written to config.output_path.

Raises:

Type	Description
ValueError	If config.id_column is not present in the input data.

write_table(df, path)

Write a DataFrame as csv or parquet by extension.

Parameters:

Name	Type	Description	Default
df	DataFrame	DataFrame to write.	required
path	Path	Output file path.	required

Raises:

Type	Description
ValueError	For unsupported file extensions.