You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Reproduced the problem in a new virtualenv with only neuralprophet installed, directly from GitHub.
Checked the Answered Questions on the Github Discussion board and found no solution to my issue.
Verified that my issue isn't already filed under existing issues.
Considered posting a question but determined my issue is a bug that needs addressing.
Describe the bug
An AttributeError occurs when config_regressors is None, specifically in df_utils.py at line 991, when attempting to access regressors attribute of a NoneType object in configurations using future regressors.
To Reproduce
Install NeuralProphet from GitHub in a new virtual environment.
Setting the model to include a future regressor with add_future_regressor('temperature').
After running m.fit(df) followed by m.predict(df).
Encountered AttributeError: 'NoneType' object has no attribute 'regressors'.
Expected behavior
Expected the model to handle future regressors gracefully without throwing an error, even when config_regressors is not explicitly defined for some paths through the data.
What actually happens
The error occurs consistently when predicting with future regressors without prior validation of config_regressors.
Environment
Python environment: Python 3.11.5, in a standalone venv with only neuralprophet installed.
NeuralProphet version: 1.0.0rc7.
Additional context
The issue can be mitigated by checking if config_regressors is not None before attempting to access its regressors attribute, ensuring robust error handling and preventing runtime errors.
The text was updated successfully, but these errors were encountered:
Prerequisites
Describe the bug
An
AttributeError
occurs whenconfig_regressors
isNone
, specifically indf_utils.py
at line 991, when attempting to accessregressors
attribute of aNoneType
object in configurations using future regressors.To Reproduce
add_future_regressor('temperature')
.m.fit(df)
followed bym.predict(df)
.AttributeError: 'NoneType' object has no attribute 'regressors'
.Expected behavior
Expected the model to handle future regressors gracefully without throwing an error, even when
config_regressors
is not explicitly defined for some paths through the data.What actually happens
The error occurs consistently when predicting with future regressors without prior validation of
config_regressors
.Environment
Additional context
The issue can be mitigated by checking if
config_regressors
is notNone
before attempting to access itsregressors
attribute, ensuring robust error handling and preventing runtime errors.The text was updated successfully, but these errors were encountered: