gluonts 0.15.1

Probabilistic time series modeling in Python.

Stars: 4571, Watchers: 4571, Forks: 750, Open Issues: 431

The awslabs/gluonts repo was created 5 years ago and the last code push was 2 months ago.
The project is very popular with an impressive 4571 github stars!

How to Install gluonts

You can install gluonts using pip

pip install gluonts

or add it to a project with poetry

poetry add gluonts

Package Details

Author

Amazon

License

Apache License 2.0

Homepage

https://github.com/awslabs/gluonts/

PyPi:

https://pypi.org/project/gluonts/

Documentation:

https://ts.gluon.ai/stable/

GitHub Repo:

https://github.com/awslabs/gluon-ts

Related Packages

Errors

A list of common gluonts errors.

Code Examples

Here are some gluonts code examples and snippets.

GitHub Issues

The gluonts package has 431 open issues on GitHub

• integrate with Lightning ecosystem CI
• Predictor error with lead_time >0
• Deprecate validated
• RuntimeError: Cannot serialize type lightgbm.sklearn.LGBMRegressor.
• Monte carlo em masking notebook
• Adding torch.MQF2 (multivariate in time series)
• pytorch: use distribution transforms when sampling
• Add Deep NPTS model
• Adding torch.deepvar
• Bug in NPTS.log_weighted_distance_kernel - np.ndarray instead of np.array
• ETS why are only additive models available?
• Recommendations for scaling real features with DeepAR?
• Support for Dynamic Real Features in Multivariate Grouper
• Compute metrics from user forecasts
• NaN losses leads to overfitting

See more issues on GitHub