Published March 25, 2021
| Version v0.1.0
Software
Open
KIT-IAI/pyWATTS: pyWATTS 0.1.0
Description
Added
- Added integration tests to github actions by executing the examples in root directory (#47)
- Implementation of the profile neural network (#71)
- Imports from the init files for exception, modules, and wrapper. (#27)
- Add rolling_variance, rolling_kurtosis, rolling_skewness. (#28)
- Select the kind of groupby in the rolling function by an enum (#28)
- Remove plot, to_csv, summary. Instead we add a callback functionality. Additional, we provide some basic callbacks. (#16)
- Change of the API (#16)
- Use keyword arguments instead of list for determining the input of modules. keyword that start with target are only fed into the fit function.
- Adapt the modules, so that they are compatible with the new API
- Use xarray DataArray instead of xr.DataSets for exchaning the data
- Remove collect step, since through the new API they are not necessary any more
- Add ResultStep, for selecting the desired result if a moudule provides multiple ones.
- Remove whitelister, instead columns can be selected via square brackets.
- CalendarExtraction: Remove encoding, for each encoding and suitable feature a new feature is created. E.g. month_sine. Additionally further calendar features are added. E.g. monday, tuesday, .. and cos encodings. For the different features, a enum type is defined.
Files
KIT-IAI/pyWATTS-v0.1.0.zip
Files
(269.4 kB)
Name | Size | Download all |
---|---|---|
md5:2b8b2b415b2ff2f08ab91aa27d94b298
|
269.4 kB | Preview Download |
Additional details
Related works
- Is supplement to
- https://github.com/KIT-IAI/pyWATTS/tree/v0.1.0 (URL)