log_error

pymc_marketing.mlflow.log_error(func, file_name)

Log arbitrary caught error and traceback to MLflow.

Note

The error will still be raised with the program. It is just logged to MLflow

Parameters:

funcCallable

Arbitrary function

file_namestr

The name of the MLflow artifact

Examples

import mlflow

from pymc_marketing.mlflow import log_error

def raising_function():
    raise NotImplementedError("Sorry. Not implemented")

func = log_error(raising_function, file_name="raising-function")

with mlflow.start_run():
    func()