Scatter.selected always selects items 0,1,2,... instead of using the indices passed in.

[mnjensen]

I'm brand new to ipyvolume, and I think it's great. But when I set scatter.selected(), it always ends up selecting the first n scatter points, not the n points I have given indices for. For example, scatter.selected = [23, 31, 12, 15] ends up selecting points [0, 1, 2, 3].

import ipyvolume as ipv
import numpy as np
num_arrows = 50
x = np.arange(0,1, 1/num_arrows)
y = np.full((num_arrows,), 0.5)
z = np.full((num_arrows,), 0.5)
scatter = ipv.scatter(x, y, z, size=5, size_selected=8 )

# Try first
scatter.selected = [0, 1, 2, 3]

# Try second
# scatter.selected = [23, 31, 12, 15]

# Try third
# scatter.selected = [33, 16, 19, 4]

ipv.show()

All three examples above lead to 0, 1, 2, and 3 being selected.

My setup: ipydatawidgets==4.2.0 ipykernel @ file:///C:/ci/ipykernel_1596190155316/work/dist/ipykernel-5.3.4-py3-none-any.whl ipython @ file:///C:/ci/ipython_1604083276484/work ipyvolume==0.6.0a6 ipywidgets @ file:///tmp/build/80754af9/ipywidgets_1601490159889/work jupyter==1.0.0 jupyter-client @ file:///tmp/build/80754af9/jupyter_client_1601311786391/work jupyter-console @ file:///tmp/build/80754af9/jupyter_console_1598884538475/work jupyter-core==4.6.3 jupyterlab==2.2.6 pythreejs==2.3.0 traitlets @ file:///tmp/build/80754af9/traitlets_1602787416690/work traittypes==0.2.1

[mnjensen]

It looks like scatter.selected works for me in notebook, but not in jupyter lab. Does someone have advice? Thanks