DuckDB is an in-process OLAP database designed for fast analytical queries on local data. Supports CSV, Parquet, JSON, and many other file formats.
Installation
Required packages: duckdb-engine
Drivers
duckdb-engineRecommended
PyPI Package:
duckdb-engineduckdb:////path/to/duck.dbSupported Features
JOINsSubqueriesDynamic SchemaCatalog SupportDynamic CatalogSSH TunnelingQuery CancellationFile UploadUser ImpersonationCost EstimationSQL Validation
Feature Score: 38/201
Time Grains
Common Time Grains:
SECONDMINUTEHOURDAYWEEKMONTHQUARTERYEAR
Extended Time Grains:
FIVE_SECONDSTHIRTY_SECONDSFIVE_MINUTESTEN_MINUTESFIFTEEN_MINUTESTHIRTY_MINUTESHALF_HOURSIX_HOURSWEEK_STARTING_SUNDAYWEEK_STARTING_MONDAYWEEK_ENDING_SATURDAYWEEK_ENDING_SUNDAYQUARTER_YEAR
Compatible Databases
The following databases are compatible with the DuckDB driver:
MotherDuckMotherDuck is a serverless cloud analytics platform built on DuckDB, offering collaborative data sharing and cloud-native scalability.
duckdb:///md:{database}?motherduck_token={token}Parameters:
| Parameter | Description |
|---|---|
| database | MotherDuck database name |
| motherduck_token | Service token from MotherDuck dashboard |
Cloud-hosted DuckDB with collaboration features.
Notes
DuckDB supports both local file and in-memory databases. Use `:memory:` for in-memory database.