netgraph3d 0.1.12

Convert any NetworkX graph to an interactive 3D HTML visualization

netgraph3d 🌐

Turn any NetworkX graph into a stunning interactive 3D visualization — with one line of code.

from netgraph3d import to_html
to_html(G)

A single HTML file is generated and opens automatically in your browser. No server needed. No dependencies beyond NetworkX.

---

Installation

pip install netgraph3d

---

Quick Start

import networkx as nx
from netgraph3d import to_html

G = nx.karate_club_graph()
to_html(G, title="Karate Club Network")

That's it. Your browser opens with a fully interactive 3D graph.

---

Demo

▶️ Watch video: https://github.com/user-attachments/assets/48905b43-2a79-4510-aa7d-c88efff3f773

---

Features

• 🖱️ Click any node to inspect all its attributes in a side panel
• 🔍 Search bar to highlight nodes by name or label
• 🏷️ Toggle node labels on/off
• 🔗 Toggle edge labels on/off
• 🔄 Auto-rotate the scene (toggle on/off)
• 🖱️ Drag to rotate manually
• 🖱️ Scroll to zoom in/out
• ⌨️ Escape to deselect a node
• 📄 Self-contained HTML — one file, share anywhere

---

Supported Input Formats

NetworkX graph (direct)

import networkx as nx
from netgraph3d import to_html

G = nx.Graph()
G.add_node("Alice", role="admin", score=95)
G.add_node("Bob",   role="user",  score=72)
G.add_edge("Alice", "Bob", weight=0.8, relation="colleague")

to_html(G, title="My Network")

---

CSV file

import networkx as nx
import pandas as pd
from netgraph3d import to_html

# edges.csv columns: source, target, weight, relation
edges_df = pd.read_csv("edges.csv")
G = nx.from_pandas_edgelist(edges_df, source="source", target="target", edge_attr=True)

# Optional: load node attributes from nodes.csv
# nodes_df = pd.read_csv("nodes.csv").set_index("id")
# for node, attrs in nodes_df.to_dict(orient="index").items():
#     if G.has_node(node):
#         G.nodes[node].update(attrs)

to_html(G, title="CSV Network")

edges.csv format:

source,target,weight,relation
Alice,Bob,0.8,friend
Bob,Carol,0.5,colleague

---

Excel file

import networkx as nx
import pandas as pd
from netgraph3d import to_html

# graph.xlsx: sheet "edges" with columns source, target, weight, relation
edges_df = pd.read_excel("graph.xlsx", sheet_name="edges")
G = nx.from_pandas_edgelist(edges_df, source="source", target="target", edge_attr=True)

to_html(G, title="Excel Network")

graph.xlsx sheet "edges" format:

source	target	weight	relation
Alice	Bob	0.8	friend
Bob	Carol	0.5	colleague

---

JSON file

import networkx as nx
import json
from netgraph3d import to_html

with open("graph.json", encoding="utf-8") as f:
    data = json.load(f)

G = nx.Graph()
for n in data["nodes"]:
    nid = n["id"]
    attrs = {k: v for k, v in n.items() if k != "id"}
    G.add_node(nid, **attrs)
for e in data["edges"]:
    G.add_edge(e["source"], e["target"], **e.get("properties", {}))

to_html(G, title="JSON Network")

graph.json format:

{
  "nodes": [
    {"id": "Alice", "label": "Alice", "role": "admin"},
    {"id": "Bob",   "label": "Bob",   "role": "user"}
  ],
  "edges": [
    {"source": "Alice", "target": "Bob", "weight": 0.8, "relation": "friend"}
  ]
}

---

Built-in NetworkX graphs (great for testing)

import networkx as nx
from netgraph3d import to_html

G = nx.karate_club_graph()            # 34 nodes, classic social network
# G = nx.les_miserables_graph()       # Characters from Les Misérables
# G = nx.barabasi_albert_graph(50, 2) # Random scale-free network

to_html(G, title="Test Network")

---

API Reference

to_html(
    G,                                 # NetworkX graph (Graph, DiGraph, etc.)
    output_path="network_graph.html",  # Path to save the HTML file
    title="3D Network Graph",          # Title shown in browser tab and top bar
    pos_3d=None,                       # Custom {node: (x, y, z)} positions (optional)
    seed=42,                           # Random seed for layout reproducibility
    open_browser=True                  # Automatically open in browser
)

Node attributes

Any attribute added to a node is automatically displayed in the info panel when clicked:

G.add_node("Alice", label="Alice Smith", department="Engineering", level=3)

The label attribute is used as the node's display name. All other attributes appear in the side panel.

Edge attributes

G.add_edge("Alice", "Bob", weight=0.9, relation="manager", since=2021)

• weight — controls edge thickness and opacity (0.0 to 1.0)
• relation or type — used as the edge label in the visualization
• All other attributes are stored and accessible

---

Requirements

• Python >= 3.8
• networkx

Optional (for CSV/Excel input):

• pandas
• openpyxl (for Excel files: pip install openpyxl)

---

License

MIT License — free to use, modify, and distribute.