Can't get axes to be fixed in 3D scatter plot

I’m trying to visualize pointclouds in the 3D scatter plot. Everything plots fine, except my axes keep changing with each update. The annoying consequence is that the visualization is distorted (as the axes are not all scaled similarly), making it quite difficult to tell if anything is working. It seems from the examples that xtickmax and the like should do the trick, but it doesn’t seem to be. What am I missing?

I am using the most recent visdom available via pip.

My code snippet:

vis.scatter(
	X=pts,
	Y=labels,
	win='points',
	opts=dict(
		title='Points',
		legend=['Predicted', 'Ground Truth'],
		markersize=3,
		markercolor=torch.tensor([[0,0,255], [255,0,0]]).int().numpy(),
		xtickmin=-10,
		xtickmax=10,
		xtickstep=0.5,
		ytickmin=-10,
		ytickmax=10,
		ytickstep=0.5,
		ztickmin=-10,
		ztickmax=10,
		ztickstep=0.5))

Output snapshots:

[JackUrb]

It’s possible that something isn’t working properly with 3dplots and plotly - I can investigate sometime this week.

[JackUrb]

Hi @meder411 - Try installing from source, this should work for you now. As an aside, tickstep is ignored on 3d plots