leanframe

(This is not an official Google product.)

LeanFrame is a pandas-like API built on top of Ibis. This is a fork of an early version of bigframes (https://github.com/googleapis/python-bigquery-dataframes) and is intended to be a minimal wrapper on top of Ibis without any of the pandas features that make bigframes more complex.

Installation

LeanFrame is available on PyPI.

pip install leanframe

Also, install the dependencies for an Ibis backend, such as

pip install 'ibis-framework[bigquery]'

Usage

To start using LeanFrame, first create a Session from an Ibis backend, such as DuckDB (local) or BigQuery (Google Cloud).

import ibis
import leanframe

connection = ibis.bigquery.connect(
    project_id="your-project-id",
)
session = leanframe.Session(connection)

Create a DataFrame from a table.

df = session.read_sql_table("bigquery-public-data.usa_names.usa_1910_2013")

Perform pandas operations.

TODO

Get your results as a pandas DataFrame.

pddf = df.to_pandas()

Disclaimer

This is a personal project. Does not reflect my employer.