geometris overlay display overextents to the wrong slices

[qianyizhang]

hi, as shown in the gif, the cube geometries overextent to the slices beyond its original range.

ps: notice on y+z plane, it's extented to the direction larger than supposed to, while on x plane, it's extented to slices smaller than supposed to.

minimum code to replicate:

# helper functions
import vtk
def pts_to_cube(pos, radius=2):
    """
    covert pos to cube (vtkPolyData) for visulization
    """
    # x = array of 8 3-tuples of float representing the vertices of a cube:
    pts = [(0.0, 0.0, 0.0), (1.0, 0.0, 0.0), (1.0, 1.0, 0.0), (0.0, 1.0, 0.0),
           (0.0, 0.0, 1.0), (1.0, 0.0, 1.0), (1.0, 1.0, 1.0), (0.0, 1.0, 1.0)]

    pts = [np.array(pt)*radius * 2 + pos - radius for pt in pts]
    faces = [(0, 1, 2, 3), (4, 5, 6, 7), (0, 1, 5, 4),
             (1, 2, 6, 5), (2, 3, 7, 6), (3, 0, 4, 7)]
    return np_to_polydata(pts, faces)


def np_to_polydata(pts, faces):
    """
    convert np.points (+ faces) to vtk.polydata
    """
    polyData = vtk.vtkPolyData()
    numberOfPoints = len(pts)
    points = vtk.vtkPoints()
    for x, y, z in pts:
        points.InsertNextPoint(x, y, z)
    polyData.SetPoints(points)
  
    polys = vtk.vtkCellArray()
    for indices in faces:
        polys.InsertNextCell(len(indices))
        for ind in indices:
            polys.InsertCellPoint(ind)
    polyData.SetPolys(polys)
    return polyData

# actual jupyter codes
import numpy as np
from itkwidgets import view

img_size = 32
pts_num = 3
img = np.zeros([img_size]*3)
pts_list = (np.random.random((pts_num, 3))*img_size).astype(int)
img[tuple(pts_list[:, ::-1].T)] = 1
cubes = [pts_to_cube(pts) for pts in pts_list]
view(img, geometries=cubes)

[jourdain]

https://github.com/Kitware/vtk-js/issues/1553

[qianyizhang]

no, this is not what i reported here. The cube becomes infinite height pipe (in the direction of viewing plane) on 2d planes

[thewtex]

@qianyizhang We could add a Cutter in a plane in front of the image plane.