API Reference

This section provides comprehensive API documentation for all Seeknal modules. The documentation is auto-generated from Python docstrings using mkdocstrings.

Module Overview

Seeknal is organized into several key modules, each serving a specific purpose in the data and ML engineering workflow.

Core Modules

Module	Description
Core	Core classes including Project, Entity, Flow, and Context
FeatureStore	Feature store management, feature groups, and materialization
Tasks	Task definitions for different data processing engines

Workflow Modules

Module	Description
Intervals	Interval tracking and cron-based scheduling for incremental processing
Change Detection	SQL-aware change detection for efficient incremental rebuilds
State Backends	Pluggable state storage for distributed execution
Environments	Environment management for plan/apply workflow
Prefect Integration	Prefect flows for distributed execution

Module Hierarchy

seeknal/
├── project.py          # Project management
├── entity.py           # Entity definitions
├── flow.py             # Data pipeline flows
├── context.py          # Execution context
├── configuration.py    # Configuration management
├── featurestore/       # Feature store module
│   ├── featurestore.py # Feature store core
│   └── feature_group.py# Feature group definitions
├── tasks/              # Task definitions
│   ├── base.py         # Base task class
│   ├── sparkengine/    # Spark engine tasks
│   └── duckdb/         # DuckDB tasks
└── exceptions/         # Exception classes

Quick Navigation

Core Classes

• Project - Manage Seeknal projects and their lifecycle
• Entity - Define entities with join keys for feature stores
• Flow - Create and manage data transformation pipelines
• Context - Execution context and session management

FeatureStore Classes

• FeatureStore - Feature store management
• FeatureGroup - Define and materialize feature groups
• Materialization - Configure feature materialization
• HistoricalFeatures - Retrieve historical feature values

Task Classes

• Task - Base class for all tasks
• SparkEngineTask - Spark-based data processing
• DuckDBTask - DuckDB-based data processing

Docstring Format

All API documentation follows the Google Python Style Guide for docstrings. Each documented item includes:

• Description: What the class/function does
• Args/Parameters: Input parameters with types and descriptions
• Returns: Return value type and description
• Raises: Exceptions that may be raised
• Example: Usage examples where applicable

Filtering

Private Members

Private members (those starting with _) are excluded from the documentation by default. Only public APIs are documented.