Changelogs » Potpourri

Potpourri

0.12.0

* seasalt.MultiBinaryToMultiClass added

0.11.0

* explicit mechansim to load verto transformers with default settings

0.10.1

* Fixed the globals() issue in transformer_custom_settings

0.10.0

* verto.raw1 to copy the raw features with mlxtend's CopyTransformer

0.9.2

* hstack replaced by concatenate in transformer_custom_settings b/c it throws error in production

0.9.1

* loading error fixed

0.9.0

* read "trans*" objects from globals to merge multiple transformers with seasalt.transformer_custom_settings

0.8.1

* verto.gnb1 renamed as verto.nb1

0.8.0

* MultiOutputClassifier2 that can handle fit_params
* verto.rnd1.RandomFeature for Feature Importance analysis
* verto.gnb1.SimpleGNB to fit simple NB models and use predictions as features

0.7.0

* OneHot encoded KMeans as feature

0.6.0

* QuantileTransformer added

0.5.0

* Isolation Forest added as verto.out1. Use anomaly score as feature

0.4.0

* Binning transformer verto.pdf1.HistDensity to estimate
the PDF of a certain value.

0.3.1

* allow to set L1 alpha in verto.dim4.PcaModelSelect

0.3.0

* verto.PcaModelSelect, verto.dim4, dim5, dim5 added

0.2.0

* specified how to generate feature names in verto
* `verto.dim1` added
* `seasalt.create_feature_names` and `verto` code adjusted to it

0.1.4

* Fix 2 -- pretty summary

0.1.3

* Fix 3 -- verto.dim2 Pipeline wrapped into PcaMinka class

0.1.2

* Fix 1 -- "__repr__" and "__str__" method of verto.dim3.PcaEv aligned

0.1.1

* modules added in setup.py

0.1.0

* Initial upload

0.0.0

* Pre-release development phase started