如何使用NumPy绘制指定三维矢量场？箭头长度控制问题求助

Hey there! Let's tackle your 3D vector field problem head-on. You've got the right start with meshgrids and normalization, but let's iron out the kinks in your code and clarify the confusing bits.

Core Issues Addressed #

First, let's recap and solve your two main questions, plus the error you ran into:

1. Correctly Calculating Vector Components with NumPy #

Your initial component definitions (F_x = y*e**x) won't work because:

• You need to use NumPy's exponential function np.exp() instead of a raw e (which isn't defined in this context)
• You have to use the meshgrid arrays (X_grid, Y_grid, Z_grid) you created, not undefined variables x, y, z

2. Controlling Arrow Length #

The length parameter in plt.quiver sets the fixed length of each arrow, but it works alongside other parameters like scale and normalization to adjust how vectors are displayed:

• If you want all arrows to have the same fixed length (useful for visualizing direction only), normalize your vectors first
• If you want arrow length to represent vector magnitude, skip normalization and use scale to adjust the overall scaling so arrows don't overlap

3. Fixing the TypeError: Shape should contain integers only #

This error pops up when the input arrays to quiver don't match in shape, or if you're passing non-integer values where they're expected. In your case, it's likely because you used undefined variables instead of your meshgrid arrays.

4. What Does quiverkey Do? #

quiverkey adds a scale bar to your plot. It tells viewers how much a specific arrow length corresponds to in your vector units (e.g., the example you had uses $2 \frac{m}{s}$ to say "this arrow length represents a vector magnitude of 2 m/s").

Full Working Code #

Here's the corrected, commented code that fixes all these issues:

import numpy as np
import matplotlib.pyplot as plt

# 1. Define the grid points
xf = np.linspace(-0.15, 2.25, 8)
yf = np.linspace(-0.15, 2.25, 8)
zf = np.linspace(-0.75, 2.50, 8)
X_grid, Y_grid, Z_grid = np.meshgrid(xf, yf, zf)

# 2. Calculate vector components using NumPy operations
F_x = Y_grid * np.exp(X_grid)
F_y = X_grid**2 + np.exp(X_grid)
F_z = Z_grid**2 * np.exp(Z_grid)

# 3. Optional: Normalize vectors to uniform length (for direction visualization)
# Remove this block if you want length to represent magnitude
magnitude = np.sqrt(F_x**2 + F_y**2 + F_z**2)
F_x_normalized = F_x / magnitude
F_y_normalized = F_y / magnitude
F_z_normalized = F_z / magnitude

# 4. Set up 3D plot
fig = plt.figure(figsize=(10, 8))
ax = fig.add_subplot(projection='3d')

# 5. Plot the vector field
# Option A: Normalized arrows (fixed length for direction)
Q = ax.quiver(
    X_grid, Y_grid, Z_grid,
    F_x_normalized, F_y_normalized, F_z_normalized,
    length=0.2,  # Fixed length for each arrow
    units='xy',
    color='blue'
)

# Option B: Magnitude-scaled arrows (length represents vector size)
# Q = ax.quiver(
#     X_grid, Y_grid, Z_grid,
#     F_x, F_y, F_z,
#     scale=50,  # Adjust this to scale arrow lengths (higher = shorter arrows)
#     color='red'
# )

# 6. Add a quiver key (scale bar)
# For normalized arrows, adjust the value to match your length/magnitude ratio
qk = ax.quiverkey(Q, 0.9, 0.9, 0.2, r'$\text{Normalized Vector}$', labelpos='E', coordinates='figure')
# For magnitude-scaled arrows, use a value that makes sense for your data:
# qk = ax.quiverkey(Q, 0.9, 0.9, 10, r'$10$', labelpos='E', coordinates='figure')

# 7. Add labels and title
ax.set_xlabel('X')
ax.set_ylabel('Y')
ax.set_zlabel('Z')
ax.set_title('3D Vector Field: $F_x = y e^x, F_y = x^2 + e^x, F_z = z^2 e^z$')

plt.show()

Key Notes #

• Normalization: Use the normalized option if you care more about seeing the direction of vectors clearly, especially when some vectors are much larger than others.
• Arrow Length Control:
  • For normalized vectors, tweak the length parameter to make arrows shorter/longer without changing their direction.
  • For magnitude-scaled vectors, adjust scale to prevent arrows from overlapping (higher scale = shorter arrows).
• 3D Plotting: Remember to use ax = fig.add_subplot(projection='3d') to enable 3D plotting in Matplotlib—your original code missed this, which would have caused issues too!

内容的提问来源于stack exchange，提问作者JD_PM