Streamlit Voronoi Treemap Component (V2)

A custom Streamlit component built on the Streamlit V2 Component Architecture that renders beautiful, dynamic Voronoi Treemaps from pandas DataFrames. It uses D3.js and d3-voronoi-treemap to lay out the data natively in the DOM, avoiding iframe overhead.

Features

• Native Rendering: Built with Streamlit V2 architecture, ensuring native DOM components without iframes.
• Grouped Hierarchies: Supports grouped data to visualize hierarchical relationships.
• Customization Options: Includes controls for color schemes, label scaling, borders, and legends.
• Responsive Sizing: Adapts dynamically to your Streamlit page width.

Installation

You can install this component in editable mode locally:

pip install -e .

To build the frontend, make sure you have npm installed:

cd st_vortree/frontend
npm install
npm run build

Quick Start

Import the custom component and pass a pandas DataFrame with your data. The DataFrame requires at a minimum a name column and a numeric value column.

import streamlit as st
import pandas as pd
from st_vortree import st_vortree

# Sample dataset
data = [
    {"Company": "Apple", "Sector": "Tech", "MarketCap": 2800},
    {"Company": "Microsoft", "Sector": "Tech", "MarketCap": 2900},
    {"Company": "Johnson & Johnson", "Sector": "Healthcare", "MarketCap": 400},
    {"Company": "Pfizer", "Sector": "Healthcare", "MarketCap": 200},
]
df = pd.DataFrame(data)

st.title("Voronoi Treemap Example")

# Render grouped Treemap
st_vortree(
    df,
    name_col="Company",
    value_col="MarketCap",
    group_col="Sector",           # Groups polygons by 'Sector'
    color_scheme="tableau10",     # Available: tableau10, category10, pastel1, dark, cool, warm
    show_values=True,
    show_legend=True,
    label_scale=1.0,
    border_color="#ffffff",
    border_width=2,
    key="my_vortree"
)

API Reference

st_vortree(df, ...)

Renders a given pandas DataFrame as a Voronoi treemap visualization.

Arguments:

• df (pandas.DataFrame): The DataFrame containing the input data.
• name_col (str): The name of the column containing the labels for the polygons. Defaults to "name".
• value_col (str): The name of the column containing the numeric sizes of the polygons. Defaults to "value".
• group_col (str, optional): The name of the column used for grouping features together into a hierarchy. Defaults to None.
• color_scheme (str): D3 categorical color scale to apply. Options include "tableau10", "category10", "pastel1", "dark", "cool", "warm", and "forest" (muted green–brown earth tones). Defaults to "tableau10".
• show_values (bool): Whether to show the metric value and relative percentage inside the polygon labels. Defaults to False.
• label_scale (float): Scale multiplier for the dynamically sized text labels. Defaults to 1.0.
• border_color (str): Hex color of the polygon outlines. Defaults to "#ffffff".
• border_width (int): Thickness of the polygon outlines. Defaults to 1.
• show_legend (bool): Whether to draw a legend beside the nodes. Defaults to True.
• key (str, optional): An optional Streamlit key to uniquely identify the component instance.

Architecture Highlights

This component does not use the deprecated V1 architecture's <iframe> mounting framework, making it significantly more performant and natively styled. The D3 logic parses and modifies standard DOM nodes wrapped by a generic Web Component interface (@streamlit/component-v2-lib).