deductable 2025.10.1

A Python library for defining and applying column functions to DuckDB tables

Deductable

de-duck-table A way to deduce new columns in your duckdb table

Description

For a client I was doing research on a collection of websites. I wanted an easy way to add a column with findings based on the previous information. For instance, I started with a list of names, for which i didn't have the URL. An agent took in the name, and found the website url.

Installation

uv pip install deductable

Requirements

• Python 3.13+
• DuckDB 1.4.1+
• Loguru 0.7.3+
• Pandas 2.3.3+

Quick Start

from typing import Optional
import duckdb
from deductable import Deductable
from my_agents import find_url

# Connect to a DuckDB database
with duckdb.connect('companies.duckdb') as con:
    # Create a Deductable instance for the table
    # Deductable expects at least an id column
    dt = Deductable(con, table_name="companies")

    @dt.column
    def company_url(company_name: str) -> Optional[float]:
        # company_name should be a column in the duckdb
        return find_url(company_name)

    # Apply the column function to populate the weight column
    dt.materialize()

    # View the result
    print(dt.df())

Features

• Type-safe column functions: Define column functions with proper type annotations
• Automatic column creation: Columns are automatically created based on function names
• Dependency handling: Functions can depend on other columns
• Optional values: Support for nullable columns with Optional type annotations
• Pandas integration: Easy conversion to pandas DataFrames

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.