Skip to content

Automated Feature Engineering

KiteML extracts predictive features from datetimes, text strings, and numeric interactions.

1. Feature Generators

  • Datetime: Year, month, day, dayofweek, hour, is_weekend.
  • Interactions: Numeric ratios and product terms.
  • Text Metrics: String length, word count, digit ratio.
from kiteml.feature_engineering import FeatureEngineeringPlanner

planner = FeatureEngineeringPlanner(enable_datetime=True, enable_interactions=True)
transformed_df = planner.fit_transform(df)