DataForgeML

DataForgeML is an automated feature engineering and ML pipeline library that provides a structured, phase-based approach to data preparation.

API Reference

• Pipeline
  • SemanticType
  • Modality
  • PipelinePhase
  • PipelineConfig
• Profiling
  • StructuralProfiler
  • ProfileConfig
  • MissingnessProfileConfig
  • NumericProfileConfig
  • NonlinearityProfileConfig
  • TypeDetectionConfig
  • CategoricalProfileConfig
  • CorrelationProfileConfig
  • DatetimeProfileConfig
  • StructuralProfileResult
  • ColumnProfile
  • DatasetStats
  • NumericKind
  • OverrideCoercionError
• Imputation
  • The guarantee boundary
  • Decision layer
  • Execution layer
  • Configuration
  • Results
  • Exceptions and warnings
• Evaluation
  • Reading a report
  • Entry point
  • Configuration
  • Records
  • Outcome vocabulary
• Splitting
  • DataSplitter
  • SplitConfig
  • SplitResult
  • FoldResult
  • HoldoutCVResult
• Persistence
  • serialize()
  • deserialize()
  • inspect()
  • IncompatibleArtifactError
  • ArtifactPythonVersionWarning
• Observability
  • PipelineEvent
  • EventType
  • stderr_observer()
• Utilities
  • DataLoader
  • UnsupportedFormatError