Reference#

Technical reference documentation for TanaT. This section is intended for looking up specific information about classes, methods, parameters, and return types.

What’s in this section?#

Glossary

Definitions of key terms and concepts used throughout TanaT documentation.

Builder & Storage

Builder lifecycle, all source methods (add_dataframe, add_parquet, add_csv, add_sql), builder options, trajectory composition, and workspace management.

Data Manipulation

One-stop lookup for all operations on pools and sequences: navigation, feature engineering, persistence, composition, type conversion, and temporal alignment.

Zeroing & Alignment

Setting a reference date (T0) with the four built-in strategies (position, direct, feature, query), pool-level inspection, and null handling.

Criteria

Filter and select sequences using declarative criteria on entity attributes, static features, time bounds, value patterns, sequence length, or rank. Covers level compatibility, parameters, and chaining.

Metadata

Metadata objects (SequenceMetadata, TrajectoryMetadata, feature types), time index inspection, and all cast_* methods available on sequence and trajectory pools.

Metrics

Hierarchical composition model (EntityMetric → SequenceMetric → TrajectoryMetric), all available metrics by category.

Clustering

Available clustering algorithms (Hierarchical, PAM, CLARA), complexity classes, characteristics, and practical usage notes.

API Documentation

Complete API reference automatically generated from source code. Browse by module to find all available classes, functions, and their signatures.