Unable to change the color of points3D

[GrowlingM1ke]

Hello there,

I'm developing an interactive GUI, where I've plotted a set of 3D points onto a plane. I managed to register clicks from the user side to recognize, which point has been selected, however I'm struggling to change the color of any selected point. All I want is for my white points to turn red when they are selected.

Instantiation of the points:

self.points_3d = mlab.points3d(*points_coords, color=(1, 1, 1), mode="cube", scale_factor=2.0)

Me attempting to update the color of a given point:

self.points_3d.mlab_source.set(x=points_coords[0], y=points_coords[1], z=points_coords[2])
sc=tvtk.UnsignedCharArray()
sc.from_array(points_coords[3])
self.points_3d.mlab_source.dataset.point_data.scalars = sc
self.points_3d.mlab_source.dataset.modified()

As it stands now whatever colors I try to input, either the points blow up in size or disappear but they don't actually change their color.

[GrowlingM1ke]

I've figured it out. For some reason if you attempt to overwrite the scalars with a new array it makes the points disappear. The solution? Overwrite the scalar information by indexing as needed:

self.points_3d.mlab_source.set(x=points_coords[0], y=points_coords[1], z=points_coords[2])
# Get indices where equal to 1
indices_selected = np.where(points_coords[3] == 1)[0]
indices_not_selected = np.where(points_coords[3] == 0)[0]
for idx in indices_selected:
  self.points_3d.mlab_source.dataset.point_data.scalars[int(idx)] = 2
for idx in indices_not_selected:
  self.points_3d.mlab_source.dataset.point_data.scalars[int(idx)] = 1
self.points_3d.mlab_source.dataset.modified()

[GrowlingM1ke]

FYI you will also need to declare the scalars in the creation of the points:

self.points_3d = mlab.points3d(*points_coords, np.ones(len(points_coords[0])), colormap="prism", mode="cube", scale_factor=2.0)