Vaex

Made compatible with Python 3.6

Requires vaex-core 4.7 for uniform settings

* Performance
* IncrementalPredictor uses parallel chunked support (2x speedup possible) [515](https://github.com/vaexio/vaex/pull/515)
* Fix
* IncrementalPredictor: epochs now iterate over the whole DataFrame instead on a batch level [523](https://github.com/vaexio/vaex/pull/523)
* Rename `vaex.ml.sklearn.SKLearnPredictor` to `vaex.ml.sklearn.Predictor` [524](https://github.com/vaexio/vaex/pull/524)
* IncrementalPredictor can be used with `sklearn.linear_model.SGDClassifier` [539](https://github.com/vaexio/vaex/pull/539)
* Features
* CycleTransformer [532](https://github.com/vaexio/vaex/pull/532)
* BayesianTargetEncoder [533](https://github.com/vaexio/vaex/pull/533)
* WeightOfEvidenceEncoder [534](https://github.com/vaexio/vaex/pull/534)
* Improve the consistency of the vaex.ml API for model transformers [552](https://github.com/vaexio/vaex/pull/552)

* Features
* IncrementalPredictor for `scikit-learn` models that support the `.partial_fit` method [497](https://github.com/vaexio/vaex/pull/497)
* Fixes
* Adding unique function names to dataframes to enable adding a predictor twice [492](https://github.com/vaexio/vaex/pull/492)

* Fixes
* Docstrings and minor improvements

vaex-graphql 0.0.1 (2019-10-15)