google-colabsqlviz

Interactive DataFrame visualization widget powered by the Google Data Studio charting engine.

Built on anywidget, google-colabsqlviz renders an interactive visualization and data exploration interface inside any notebook environment that supports IPyWidgets, including Google Colab, VS Code, JupyterLab, and more.

Embeds the Data Studio Charting Engine UI hosted on Google's CDN; however, none of your data is ever sent to Google. All aggregation logic is run in your Python kernel, and all visualizations are rendered strictly in the browser.

Features

• Interactive no-code charting: Configure chart types (bar, line, scatter, histogram, pie, pivot tables, geo charts, etc.), dimensions, metrics, aggregations, sorting, and filters directly in the notebook cell output.
• Multi-engine DataFrame support:
  • pandas.DataFrame
  • bigframes.pandas.DataFrame
  • pyspark.sql.DataFrame, including Spark Connect
• Lightweight footprint: Optional distributed engines (bigframes, pyspark) are detected at runtime and never force-installed or upgraded over your existing cluster/runtime environment.

Installation

Install on the command line in an active virtual environment:

pip install google-colabsqlviz

Or directly in a notebook cell:

%pip install google-colabsqlviz

Quick Start

Pass any supported DataFrame to display (also available as explore_dataframe):

import pandas as pd
from google.colabsqlviz.explore_dataframe import display

df = pd.DataFrame({
    "category": ["A", "B", "A", "C", "B", "A"],
    "value": [10, 25, 15, 40, 30, 20],
})

display(df)

If bigframes (>= 2.35.0) or pyspark is installed in your kernel, you can pass a bigframes.pandas.DataFrame or pyspark.sql.DataFrame directly to explore_dataframe.explore_dataframe(...) and aggregations will be executed by that engine.

Documentation & Resources

• Data Studio Chart References — Documentation on the chart types and aggregations supported by the embedded Data Studio engine.