能否在Matplotlib中使用Soft Light混合模式绘制重叠点？

Absolutely! Starting from Matplotlib 3.4, you can use blend modes like Soft Light for your plot elements—this lets you avoid the washed-out look of alpha blending while keeping overlapping points visible. Here's how to adapt your code to make it work:

import pandas as pd
import geopandas as gpd
import matplotlib.pyplot as plt

# Load your spatial data
roads = gpd.read_file('roads.shp')
points = gpd.read_file('points.shp')

# Sort points by size in descending order (smaller points draw on top)
points_sorted = points.sort_values(by='size', ascending=False)

f, ax = plt.subplots()

# Draw roads as the base layer
roads.plot(ax=ax, linewidth=.5, color='#2D1B1B', zorder=1)

# Draw points and capture the returned PathCollection object
points_plot = points_sorted.plot(ax=ax, column='color', markersize='size', zorder=2)

# Apply the Soft Light blend mode (no alpha needed!)
points_plot.set_blend_mode('soft_light')

plt.axis('equal')
plt.show()

Key Details & Notes: #

• Point sorting: By sorting your points GeoDataFrame by size descending, you ensure smaller points sit on top of larger ones—this is critical for the blend mode to behave as you want, since blending depends on layer order.
• Blend mode explanation: The plot() method for GeoPandas points returns a PathCollection object. Calling set_blend_mode('soft_light') tells Matplotlib to use the Soft Light mixing algorithm instead of default opaque overlay. This mode preserves the depth of darker underlying large points while keeping lighter small points crisp against a white background, no alpha required.
• Version check: Make sure you're running Matplotlib 3.4 or newer. If not, upgrade with pip install --upgrade matplotlib.
• Data prep: Ensure your size column is numeric (use points['size'] = pd.to_numeric(points['size']) if needed) to avoid errors with the markersize parameter.

内容的提问来源于stack exchange，提问作者Ryan