Python API

The preferred Python entry points are the public facade modules. Native binding classes may appear in signatures because the facade delegates to the compiled extension, but user code should import through mtlearn.morphology and mtlearn.layers.

• Morphology Facade
  • TreeType
  • create_max_tree()
  • create_min_tree()
  • create_tree_of_shapes()
  • build_tree()
  • normalize_tree_type()
  • normalize_tos_interpolation()
  • compute_attributes()
  • compute_single_attribute()
  • describe_attribute()
  • describe_all_attributes()
  • expand_attribute_group()
  • create_attribute_filter()
  • is_tree()
• Layers
  • Primary CFP Layer
  • Checkpoint Helpers
• CFP Preparation and Storage
  • Preparation
  • Storage policies
  • Normalization snapshot
• Native Binding Classes
  • Morphology Trees and Filters
  • CFP Native Helpers
• Data Helpers
  • DatasetSpec
  • repo_root()
  • default_data_dir()
  • dataset_path()
  • require_local_dataset()
  • dropbox_download_url()
  • format_size()
  • download_file()
  • extracted_content_root()
  • replace_directory()
  • has_existing_files()
  • ensure_dataset()
  • parse_args()
  • list_datasets()
  • main()
• Datasets
  • GeneratedTargetImageDataset
  • PairedImageDataset
  • SplitValidationReport
  • validate_split_ids()
  • subsets_from_ids()
  • split_ids_by_group()
  • PairAudit
  • audit_image_pairs()
  • SplitManifest