Releases: dataiku/dss-plugin-timeseries-forecast
Releases · dataiku/dss-plugin-timeseries-forecast
Version 1.2.1 - Deprecation
The plugin is now deprecated because the forecasting features are now native to DSS since v11.
Version 1.2.0 - New features
Merge pull request #66 from dataiku/feature/1.2-forecast-new-data Release 1.2
Version 1.1.0 - New feature and bugfix release
- ✨ Added Python 3.7 support (introduced in DSS 9.0)
- 🪲 Fixed wrong default value for the end of week frequency
Version 1.0.1 - Bugfix release
- 🪲 Bugfix with the FeedForward model when using external features of different lengths
- 💅 Small naming and logging enhancements
Version 1.0.0 - Initial release
- 🔥 Multivariate forecasting, scaling to 1000s of time series
- 😎 Deep Learning models from GluonTS (DeepAR, Transformer, ...)
- 🗓 Frequencies from year to minute, inc. business day
- 🐍 Easy-to-install code environment in Python
- 👾 GPU version to train your models even faster
Version 1.0.0 - Release Candidate 2
Merge pull request #53 from dataiku/feature/seasonal-models add stl model and season length parameter
Version 1.0.0 - Release Candidate 1
v1.0.0-rc1 Merge branch 'release/1.0'
Version 1.0.0 - Beta 3
Merge pull request #43 from dataiku/feature/default-key-value-param Feature/default key value param
Version 1.0.0 - Beta 2
Apply suggestions from code review Co-authored-by: StanislasGuinel <[email protected]>
Version 1.0.0 - Beta 1
Merge pull request #36 from dataiku/feature/data-preparation Feature/data preparation