Pyplot技术问询：如何制作带非线性刻度的离散色条？

Alright, let’s solve this problem properly. You need a discrete 5-color colorbar with non-linear scaling without relying on contourf—this approach works universally with most Matplotlib plot types (like scatter, pcolormesh, imshow) so you can adapt it to your use case.

Step-by-Step Approach #

The core of the solution uses BoundaryNorm to map your data to discrete color bins, paired with a custom colormap. We’ll also manually configure non-linear ticks for the colorbar to match your bin boundaries.

Full Example Code #

import matplotlib.pyplot as plt
import numpy as np
from matplotlib.colors import BoundaryNorm, ListedColormap

# ----------------------
# 1. Generate test data (replace with your actual data)
# ----------------------
x = np.linspace(0, 10, 100)
y = np.linspace(0, 10, 100)
X, Y = np.meshgrid(x, y)
# Create data with non-linear distribution (simulating real-world data)
Z = np.exp(X/2) * np.sin(Y)
Z = np.abs(Z)  # Use positive values for easier binning

# ----------------------
# 2. Define NON-LINEAR bin boundaries (5 bins = 6 boundaries)
# Adjust these to match your desired non-linear scale!
# ----------------------
bin_boundaries = [0, 1, 5, 20, 50, 100]
num_bins = len(bin_boundaries) - 1  # Should be 5

# ----------------------
# 3. Create discrete colormap (5 custom colors)
# ----------------------
custom_colors = [
    "#ffcccc",  # Light red
    "#ff9999",  # Red
    "#ff6666",  # Darker red
    "#cc0000",  # Deep red
    "#990000"   # Almost black red
]
cmap = ListedColormap(custom_colors)

# ----------------------
# 4. Create BoundaryNorm to map data to discrete bins
# ----------------------
norm = BoundaryNorm(bin_boundaries, cmap.N, clip=True)

# ----------------------
# 5. Plot your data (using pcolormesh here; swap for scatter/imshow as needed)
# ----------------------
fig, ax = plt.subplots(figsize=(8, 6))
im = ax.pcolormesh(X, Y, Z, cmap=cmap, norm=norm)

# ----------------------
# 6. Add colorbar with NON-LINEAR ticks
# ----------------------
cbar = plt.colorbar(im, ax=ax)
# Set ticks to the MIDPOINT of each bin (or use bin_boundaries if you prefer)
cbar_ticks = [(bin_boundaries[i] + bin_boundaries[i+1])/2 for i in range(num_bins)]
# Set tick labels to show the bin ranges (customize this as needed)
cbar_ticklabels = [f"{bin_boundaries[i]}–{bin_boundaries[i+1]}" for i in range(num_bins)]
cbar.set_ticks(cbar_ticks)
cbar.set_ticklabels(cbar_ticklabels)
cbar.set_label("Non-Linear Data Range")

# Add plot labels
ax.set_xlabel("X Axis")
ax.set_ylabel("Y Axis")
ax.set_title("Discrete Colorbar with Non-Linear Scaling (No contourf)")

plt.tight_layout()
plt.show()

Key Explanations #

• BoundaryNorm: This is the critical tool that turns continuous data into discrete color bins. It uses your non-linear boundaries to map each data point to the correct color—no contourf required.
• Custom Colormap: ListedColormap lets you define exactly 5 distinct colors, ensuring you get a discrete colorbar instead of a gradient.
• Non-Linear Ticks: By manually setting the colorbar ticks and labels, you can clearly communicate the non-linear bin ranges. You can swap midpoint ticks for the boundary values themselves if that fits your visualization better.
• Universal Compatibility: This method works with any plot function that accepts cmap and norm parameters. For example, if you’re using scatter, just pass cmap=cmap, norm=norm to plt.scatter() and configure the colorbar the same way.

内容的提问来源于stack exchange，提问作者brzepkowski