for testing and deploying your application
for finding and fixing issues
for empowering human code reviews
import pandas as pd
from data.encoding import EncoderFactory, AbstractEncoder
class NominalEncoder(AbstractEncoder):
def encode(self, *args, **kwargs):
dataset, feature = args[0], args[1]
return pd.get_dummies(dataset.datapoints.observations.df, f'{feature.label}-{feature.state.current}')
@EncoderFactory.register_as_subclass('pandas')
class DFEncoderFactory(EncoderFactory):
def nominal_encoder(self, feature, **kwargs):
return NominalEncoder()
boromir674 /
so-magic
