garphield for Python

Interactive graph visualization for Python notebooks. Pass a NetworkX graph or pandas DataFrame to show() and get a WebGL-rendered view inline. Select nodes, bind visual properties to algorithms, apply force-directed layouts, annotate structure, then round-trip the result back to Python or export standalone HTML.

import garphield as gph
import networkx as nx

G = nx.karate_club_graph()
view = gph.show(
    G,
    node_color="club",
    node_size=G.degree,
    node_label=str,
)

Once the view is displayed, drive it from Python:

view.select([0, 1, 2]).fit()
view.bind("node_color", gph.algorithm("louvain", resolution=1.1))
edited = view.to_project()

Install

pip install garphield

With NetworkX and pandas adapters:

pip install 'garphield[tables,networkx]'

From pandas

import pandas as pd
import garphield as gph

edges = pd.DataFrame({"source": ["a"], "target": ["b"], "weight": [1.5]})
project = gph.Project.from_pandas(edges)
tables = project.to_pandas()

Project files

Read, validate, and write .gph project files:

import garphield as gph

project = gph.Project.load("project.gph")
print(project.semantic_fingerprint())
project.save("project-copy.gph")

Project.save() writes deterministic, human-readable JSON. Semantic and project fingerprints use RFC 8785/JCS canonical bytes.

Notebook round-trip

The wheel carries a version-matched notebook renderer. Use Open in Garphield to continue in the full workbench at garphield.com, then Return to notebook to bring the result back to Python.

Use chrome=["minimap"] or chrome=["toolbar"] to keep embeds lightweight. project.save_html("graph.html", chrome=["minimap"]) writes an interactive HTML shell that loads Garphield's remote /embed renderer. It requires network access and a browser/host context that permits framing.

Create a complete project URL without opening a browser or uploading data:

url = gph.share_url(project)

share_string() and share_url() validate and encode locally. Links above the 262,144-character inline limit raise ProjectShareTooLargeError; use a .gph file for larger projects.

Local Jupyter, VS Code notebooks, JupyterHub, Google Colab, and Snowflake Snowpark notebooks are supported.

Links

• garphield.com
• GitHub