PyVista设置nan_opacity=0无法实现含NaN网格区域透明的问题求助

我最近在测试PyVista处理带NaN值的多层网格渲染，按照预期，设置nan_opacity=0.0应该让NaN对应的网格区域完全透明，这样三层非重叠的网格应该能组合成一个红、蓝、绿三段式的球体，但实际运行后只得到了一个全绿色的球体（如下方图片所示）。

以下是我的测试代码：

import pyvista as pv
import numpy as np

def test_pyvista_layers():
    """Test how PyVista handles multiple mesh layers with NaN values and opacity settings"""
    
    # Create a simple sphere
    sphere = pv.Sphere(radius=1.0)
    total_points = sphere.n_points
    
    # Create three meshes with known patterns
    mesh1 = sphere.copy()
    mesh2 = sphere.copy()
    mesh3 = sphere.copy()
    
    # Test 1: Basic overlap with different values and NaN
    values1 = np.full(total_points, np.nan)  # Initialize with NaN
    values1[:total_points//3] = 1  # Set first third to 1
    
    values2 = np.full(total_points, np.nan)  # Initialize with NaN
    values2[total_points//3:2*total_points//3] = 2  # Set middle third to 2
    
    values3 = np.full(total_points, np.nan)  # Initialize with NaN
    values3[2*total_points//3:] = 3  # Set final third to 3
    
    # Validate NaN counts and non-overlap
    nan_count1 = np.sum(np.isnan(values1))
    nan_count2 = np.sum(np.isnan(values2))
    nan_count3 = np.sum(np.isnan(values3))
    
    print(f"\nNaN Counts:")
    print(f"Layer 1: {nan_count1} NaN values ({(nan_count1/total_points)*100:.1f}% of points)")
    print(f"Layer 2: {nan_count2} NaN values ({(nan_count2/total_points)*100:.1f}% of points)")
    print(f"Layer 3: {nan_count3} NaN values ({(nan_count3/total_points)*100:.1f}% of points)")
    
    # Check for overlap between non-NaN regions
    overlap_12 = np.sum(~np.isnan(values1) & ~np.isnan(values2))
    overlap_23 = np.sum(~np.isnan(values2) & ~np.isnan(values3))
    overlap_13 = np.sum(~np.isnan(values1) & ~np.isnan(values3))
    
    print("\nOverlap Check:")
    print(f"Overlap between layers 1-2: {overlap_12} points")
    print(f"Overlap between layers 2-3: {overlap_23} points")
    print(f"Overlap between layers 1-3: {overlap_13} points")
    
    # Validate complete coverage
    total_coverage = np.sum(~np.isnan(values1) | ~np.isnan(values2) | ~np.isnan(values3))
    print(f"\nTotal Coverage: {total_coverage} points ({(total_coverage/total_points)*100:.1f}% of surface)")
    
    plotter = pv.Plotter()
    
    # Add meshes with point data and NaN handling
    mesh1.point_data['values'] = values1
    mesh2.point_data['values'] = values2
    mesh3.point_data['values'] = values3
    
    # Add meshes with nan_opacity=0 to make NaN values completely transparent
    plotter.add_mesh(mesh1, color='red', opacity=1.0, label='Layer 1', nan_opacity=0.0)
    plotter.add_mesh(mesh2, color='blue', opacity=1.0, label='Layer 2', nan_opacity=0.0)
    plotter.add_mesh(mesh3, color='green', opacity=1.0, label='Layer 3', nan_opacity=0.0)
    
    return plotter

test_plotter = test_pyvista_layers()
test_plotter.show()

运行代码后，控制台输出验证了各层的NaN值分布是正确的，且三层的非NaN区域完全没有重叠，覆盖了整个球体的所有点：

NaN Counts:
Layer 1: 562 NaN values (66.7% of points)
Layer 2: 561 NaN values (66.6% of points)
Layer 3: 561 NaN values (66.6% of points)

Overlap Check:
Overlap between layers 1-2: 0 points
Overlap between layers 2-3: 0 points
Overlap between layers 1-3: 0 points

Total Coverage: 842 points (100.0% of surface)

但实际渲染出的效果却是这样的全绿色球体：

我尝试了各种不同的渲染方式，也仔细查阅了PyVista的文档，但都没能解决这个问题。有没有人能告诉我哪里出问题了，或者该怎么正确实现预期的三段式球体渲染效果？

备注：内容来源于stack exchange，提问作者Jonathan