Cone not working properly?

[jzthree]

Thanks for implementing the feature. I have been having some difficulties in reproducing the notebook https://plot.ly/python/3d-cone/ or tried to use ploly.graph_objs.Cone.

With iplot it shows a plot with 2D layout like scatter while 3d is expected.
With plot to save to a file and open it with chrome. It shows an empty grid.

Also "Edit in Chart Studio" button does not work for me in either case.

I have verified that I am using 3.0.0 version, and the html files says plotly.js 1.38.3. I have been mostly following the notebook except for using plotly.offline.plot and iplot instead.

[jonmmease]

Thanks for the report @jzthree.

What I'm seeing is that I can reproduce the plot in the documentation as long as the validate=False param is set. But, if I don't set validate to False I see the empty grid. I think I've tracked down the empty grid problem to an issue with our code generation logic (Note to self: we codegen a temporary variable named v in the constructor of each graph obj that is conflicting with the v property of cone).

Could you add a screenshot of the 2D layout you're seeing?

Cone is a fairly new trace, so my guess is that Chart studio hasn't been updated yet.

I'll fix the code generation bug next week, and at that point I'll check on the status of chart studio support for the cone trace.

Thanks!

[jzthree]

Thanks for looking into this! This is what i got

from plotly.offline import  init_notebook_mode, plot, iplot
import plotly.graph_objs as go
init_notebook_mode()

data = [{
    'type': 'cone',
    'x': [1], 'y': [1], 'z': [1],
    'u': [1], 'v': [1], 'w': [0]
}]

layout = {
    'scene': {
      'camera': {
        'eye': {'x': -0.76, 'y': 1.8, 'z': 0.92}
      }
    }
}

fig = {"data": data, "layout": layout}
iplot(fig, filename='cone-basic', validate=False)

[jonmmease]

@jzthree Can you try this out with the new version 3.0.1?

[jzthree]

Thanks. Now using plotly.offline.plot to save to a html file works perfectly, but plotly.offline.iplot still behaves like the above image I had.

[jonmmease]

If plotly.offline.plot works but plotly.offline.iplot doesn't, it's probably a Plotly.js version difference. You can try out master (which was updated with the latest Plotly.js version this morning), or we can check in on this again when 3.1 is released.

[jonmmease]

Hi @jzthree , could you try the plotly.offline.iplot approach again with version 3.1.0?

[jzthree]

Thanks for the quick update! I still have the same problem with 3.1.0 though...

[jonmmease]

Oh wait, what interface are you using? That doesn't look like Jupyter notebook/lab.

[jzthree]

Oh I just tried jupyter notebook and that works (see below)

The plot I showed before was nteract (see below).

I also used Atom+Hydrogen and it had the same problem, so I never thought it would be interface-dependent- sorry. I don't have enough expertise to tell whether this shoud be an issue for nteract @rgbkrk. Shoud we close this issue?

[jonmmease]

It might be a matter of what version of plotly.js is bundled with nteract.

@rgbkrk is the plotly.js version that nteract includes based on the version in https://github.com/nteract/minimal-plotly? If so, that would be version 1.25.0, which would explain the issue since cone traces were added in 1.39.0.

[jonmmease]

I'm going to close this for now since I don't think it's something we can fix in plotly.py. I'd recommend filing an issue on the nteract side. Feel free to reopen if it turns out there is something we can do.
Thanks