如何使用igraph绘制椭圆节点？求自定义节点形状完整示例

Absolutely, let's break down how to create custom node shapes (especially ellipses) in igraph—since the official docs can be a bit sparse on this. I'll walk you through a full, working example with both basic and advanced tweaks, using Python's python-igraph and matplotlib (the most common setup for this kind of customization).

Make sure you have the required packages installed:

pip install python-igraph matplotlib

The key here is defining a custom drawing function that returns a matplotlib Ellipse patch, then passing that function to igraph's plotting engine via the vertex_shape parameter.

Here's a complete runnable example:

import igraph as ig
import matplotlib.pyplot as plt
from matplotlib.patches import Ellipse

# Step 1: Create a sample graph
g = ig.Graph()
g.add_vertices(5)
g.add_edges([(0,1), (1,2), (2,3), (3,4), (4,0)])

# Step 2: Define the custom ellipse drawing function
def draw_ellipse(vertex, coords, size, **kwargs):
    # coords = (x, y) position of the node
    # size = base size set in igraph's plot params
    x, y = coords
    
    # Adjust width/height to get an ellipse (tweak these multipliers as needed)
    ellipse_width = size * 1.5
    ellipse_height = size * 0.8
    
    # Create and return the Ellipse patch
    return Ellipse(
        (x, y),
        width=ellipse_width,
        height=ellipse_height,
        angle=0,  # Add a value here (e.g., 30) to rotate the ellipse
        **kwargs  # Passes color, edge width, etc., from igraph's plot settings
    )

# Step 3: Plot the graph with custom ellipses
fig, ax = plt.subplots(figsize=(8, 6))

ig.plot(
    g,
    target=ax,
    vertex_shape=draw_ellipse,  # Critical: Use our custom function
    vertex_size=20,  # Base size that our function scales
    vertex_color="#6a994e",  # Fill color for ellipses
    vertex_edge_color="#386641",  # Border color
    vertex_edge_width=2,  # Border thickness
    edge_color="#bc4749",  # Edge color
    edge_width=2,
    vertex_label=["A", "B", "C", "D", "E"],  # Node labels
    vertex_label_size=12,
    vertex_label_color="#ffffff"  # Label text color
)

# Adjust axes to fit the graph neatly
ax.set_xlim(ax.get_xlim()[0] - 5, ax.get_xlim()[1] + 5)
ax.set_ylim(ax.get_ylim()[0] - 5, ax.get_ylim()[1] + 5)
ax.axis('off')  # Hide default axes

plt.show()

Want different ellipses for different nodes? You can attach custom attributes to vertices and use them in your drawing function. Here's how:

# Add custom attributes to each node (width, height, rotation angle)
for idx, vertex in enumerate(g.vs):
    vertex['ellipse_width'] = 25 + idx * 5  # Increase width per node
    vertex['ellipse_height'] = 15 + idx * 3  # Increase height per node
    vertex['ellipse_angle'] = idx * 15  # Rotate each node incrementally

# Updated drawing function that uses vertex attributes
def draw_custom_ellipse(vertex, coords, size, **kwargs):
    x, y = coords
    # Fall back to default values if the attribute doesn't exist
    width = vertex.get('ellipse_width', size * 1.5)
    height = vertex.get('ellipse_height', size * 0.8)
    angle = vertex.get('ellipse_angle', 0)
    
    return Ellipse(
        (x, y),
        width=width,
        height=height,
        angle=angle,
        **kwargs
    )

# Plot with personalized ellipses
fig, ax = plt.subplots(figsize=(8, 6))

ig.plot(
    g,
    target=ax,
    vertex_shape=draw_custom_ellipse,
    vertex_color=["#6a994e", "#f2e8cf", "#bc4749", "#386641", "#a7c957"],
    vertex_edge_color="#000000",
    edge_color="#555555",
    vertex_label=["A", "B", "C", "D", "E"],
    vertex_label_size=12
)

ax.set_xlim(ax.get_xlim()[0] - 5, ax.get_xlim()[1] + 5)
ax.set_ylim(ax.get_ylim()[0] - 5, ax.get_ylim()[1] + 5)
ax.axis('off')

plt.show()

• The vertex_shape parameter accepts any function that returns a matplotlib Patch object—so you're not limited to ellipses; you could draw rectangles, polygons, or even custom paths!
• Use vertex attributes to make shapes dynamic per node (size, rotation, color, etc.)
• Always use the target=ax parameter to plot onto a matplotlib axis, which gives you full control over the final output.

内容的提问来源于stack exchange，提问作者baidao