A simple interface containing easy tensorflow model building blocks and feature encoding pipelines. This module contains functionality similar to what sklearn does with its Pipeline, FeatureUnion and ColumnTransformer. Install package:
pip install easy-tensorflow
Encoder: Preprocess feature based on specified preprocessing layer
SequentialEncoder: Preprocessing pipeline to apply multiple encoders in serie
Pipeline: Main interface for transforming features. Apply feature encoder list which can contain both Encoder and SequentialEncoder object types
FeatureUnion: Similar to Pipeline, but concatenates the Layers from Pipeline
Import modules and necessary imports from easyflow
For illustration purposes the example UCI heart dataset.