DuckDB Kernel

DuckDB Kernel is a powerful SQL execution environment designed for Jupyter notebooks, implemented in Go. This kernel enhances data analysis workflows by providing seamless integration with DuckDB, known for its efficient analytical capabilities. Key features include:

• SQL Execution: Execute SQL queries using an embedded DuckDB instance, enabling fast access to data.
• Interactive Result Visualizations: Leverage the Perspective viewer to transform query results into dynamic, interactive visualizations.
• Geo Map Plugin: Visualize complex datasets containing geometries (points, lines, and polygons) on interactive maps using deck.gl, accommodating large volumes of data without compromising performance.
• Optimized Data Handling: Arrow IPC streaming with LZ4 compression allows users to work with extensive datasets smoothly, without blocking the user interface.
• User-Friendly Database Explorer: Navigate through catalogs, tables, schemas, functions, and extensions with an integrated sidebar for efficient database exploration.
• Session Recovery: Automatically reload query results after a page refresh, ensuring a smooth user experience.
• Meta Commands: Utilize convenient commands for quick database exploration and management.
• Compatibility: Fully compatible with JupyterLab, JupyterHub, and Visual Studio Code environments.

Example Usage

The DuckDB Kernel enables straightforward SQL operations directly within Jupyter notebooks. For instance, a simple table creation and data insertion could be executed as follows:

CREATE TABLE users (id INTEGER, name VARCHAR);
INSERT INTO users VALUES (1, 'Alice'), (2, 'Bob');
SELECT * FROM users;

Session Sharing

Share a DuckDB session across multiple notebooks by setting the DUCKDB_SHARED_SESSION environment variable:

eexport DUCKDB_SHARED_SESSION=my-analysis
jupyter lab

Interactive Visualizations

Enable enhanced data insights by using the Perspective Viewer extension, which provides:

• Interactive Table: Sort, filter, and scroll through results effortlessly.
• Pivot Tables: Group and aggregate data dynamically.
• Charts and Geo Maps: Visualize data distributions and spatial information effectively.

Overall, the DuckDB Kernel serves as a robust tool for data analysts and developers looking for a flexible, interactive solution for SQL workflows in Jupyter environments.