Description
pandas bogs down on a few gigabytes, and a data warehouse is overkill. DuckDB is an in-process, high-speed analytical database, and this is its official Python package for querying local files and DataFrames with SQL.
It reads Parquet, CSV and pandas DataFrames directly, blazing fast with no server.
In-process:No server.
SQL:Query files.
Fast:Columnar engine.
pandas:Seamless.
It reads Parquet, CSV and pandas DataFrames directly, blazing fast with no server.
Features
In-process:No server.
SQL:Query files.
Fast:Columnar engine.
pandas:Seamless.
