update scalars does not applies correct alpha values

[ZZaphodd]

Describe the bug, what's wrong, and what you expected.

Whole mesh goes partially transparent
if i give non 1.0 as alpha value only one of cell_data

Steps to reproduce the bug.

import numpy as np
import pyvista
import random

def cube_3_3_3():
    _coords: set[tuple[float, float, float]] = set()
    cubes: list[tuple[int, int, int, int]] = list()
    coords: list[tuple[float, float, float]] = None
    coords_to_index: dict[tuple[float, float, float], int] = dict()
    cells: list[list[int]] = list()

    dim0 = 3
    dim1 = 3
    dim2 = 3

    for i0 in range(dim0):
        for i1 in range(dim1):
            for i2 in range(dim2):
                cubes.append((float(i0), float(i1), float(i2), 1))

    for cube in cubes:
        i0, i1, i2, cs = cube
        _coords.add((i0, i1, i2))
        _coords.add((i0, i1, i2+cs))
        _coords.add((i0, i1+cs, i2))
        _coords.add((i0, i1+cs, i2+cs))
        _coords.add((i0+cs, i1, i2))
        _coords.add((i0+cs, i1, i2+1))
        _coords.add((i0+cs, i1+cs, i2))
        _coords.add((i0+cs, i1+cs, i2+cs))

    coords = list(_coords)
    for idx, key in enumerate(coords):
        coords_to_index[key] = idx

    for cube in cubes:
        i0, i1, i2, cs = cube
        cells.append([8,
                      coords_to_index[(i0, i1, i2)],
                      coords_to_index[(i0+cs, i1, i2)],
                      coords_to_index[(i0+cs, i1+cs, i2)],
                      coords_to_index[(i0, i1+cs, i2)],
                      coords_to_index[(i0, i1, i2+cs)],
                      coords_to_index[(i0+cs, i1, i2+cs)],
                      coords_to_index[(i0+cs, i1+cs, i2+cs)],
                      coords_to_index[(i0, i1+cs, i2+cs)]
                      ])

    cell_types = np.array([pyvista.CellType.HEXAHEDRON] * len(cells), np.int8)

    return pyvista.UnstructuredGrid(cells, cell_types, coords)

def main():
    random_generator = random.Random(42)    
    
    plotter = pyvista.Plotter()

    grid1 = cube_3_3_3()
    colors = [[random_generator.uniform(0.0, 1.0),
               random_generator.uniform(0.0, 1.0),  
                random_generator.uniform(0.0, 1.0,), 
                    1.0
                    ] for _ in range(grid1.n_cells)]
    grid1.cell_data['colors_rgba'] = colors

    #plotter.add_mesh(grid1, scalars="colors_rgba", rgb=True)
    plotter.add_mesh(grid1, scalars="colors_rgba")

    def flip_opacity_only_one_cell():
        nonlocal colors, grid1, plotter
        # random cell
        colors[8][3] = 1.0 - colors[8][3]

        print("colors where 0<a<1: ", end="") # no entries will shows
        for rgba in colors:
            if 0<rgba[3]<1:
                print(rgba, end="")
        print()

        grid1.cell_data['colors_rgba'] = colors
        plotter.render()

    plotter.add_key_event("7",flip_opacity_only_one_cell)

    plotter.show()



if __name__ == '__main__':
    main()

System Information

--------------------------------------------------------------------------------
  Date: Sun Apr 28 08:49:41 2024 대한민국 표준시

                OS : Windows
            CPU(s) : 24
           Machine : AMD64
      Architecture : 64bit
               RAM : 32.0 GiB
       Environment : Python
        GPU Vendor : NVIDIA Corporation
      GPU Renderer : NVIDIA GeForce GTX 1060 3GB/PCIe/SSE2
       GPU Version : 4.5.0 NVIDIA 546.12
  MathText Support : False

  Python 3.11.5 | packaged by Anaconda, Inc. | (main, Sep 11 2023, 13:26:23)
  [MSC v.1916 64 bit (AMD64)]

           pyvista : 0.43.5
               vtk : 9.3.0
             numpy : 1.26.2
        matplotlib : 3.8.4
            scooby : 0.9.2
             pooch : 1.8.1
            pillow : 10.3.0
             PyQt5 : 5.15.10
           IPython : 8.15.0
              tqdm : 4.66.1
      nest_asyncio : 1.5.6
--------------------------------------------------------------------------------

Screenshots

screenshot of initial state

after '7' key pressed

[ZZaphodd]

tying to figure out when mesh really updated, #5158
not sure same problem, #5204