Plotly

Added
- Offline matplotlib to Plotly figure conversion. Use `offline.plot_mpl` to convert and plot a matplotlib figure as a Plotly figure independently of IPython/Jupyter notebooks or use `offline.iplot_mpl` to convert and plot inside of IPython/Jupyter notebooks. Additionally, use `offline.enable_mpl_offline` to convert and plot all matplotlib figures as plotly figures inside an IPython/Jupyter notebook. See examples below:

An example independent of IPython/Jupyter notebooks:

from plotly.offline import init_notebook_mode, plot_mpl
import matplotlib.pyplot as plt

init_notebook_mode()

fig = plt.figure()
x = [10, 15, 20]
y = [100, 150, 200]
plt.plot(x, y, "o")

plot_mpl(fig)


An example inside of an IPython/Jupyter notebook:

from plotly.offline import init_notebook_mode, iplot_mpl
import matplotlib.pyplot as plt

init_notebook_mode()

fig = plt.figure()
x = [10, 15, 20]
y = [100, 150, 200]
plt.plot(x, y, "o")

iplot_mpl(fig)


An example of enabling all matplotlib figures to be converted to
Plotly figures inside of an IPython/Jupyter notebook:

from plotly.offline import init_notebook_mode, enable_mpl_offline
import matplotlib.pyplot as plt

init_notebook_mode()
enable_mpl_offline()

fig = plt.figure()
x = [10, 15, 20, 25, 30]
y = [100, 250, 200, 150, 300]
plt.plot(x, y, "o")
fig

Added
- Offline plotting now works outside of the IPython/Jupyter notebook. Here's an example:

from plotly.offline import plot
from plotly.graph_objs import Scatter

plot([Scatter(x=[1, 2, 3], y=[3, 1, 6])])


This command works entirely locally. It writes to a local HTML file with the necessary [plotly.js](https://plot.ly/javascript) code to render the graph. Your browser will open the file after you make the call.

The call signature is very similar to `plotly.offline.iplot` and `plotly.plotly.plot` and `plotly.plotly.iplot`, so you can basically use these commands interchangeably.

If you want to publish your graphs to the web, use `plotly.plotly.plot`, as in:


import plotly.plotly as py
from plotly.graph_objs import Scatter

py.plot([Scatter(x=[1, 2, 3], y=[5, 1, 6])])


This will upload the graph to your online plotly account.

Added
- Check for `no_proxy` when determining if the streaming request should pass through a proxy in the chunked_requests submodule. Example: `no_proxy='my_stream_url'` and `http_proxy=my.proxy.ip:1234`, then `my_stream_url` will not get proxied. Previously it would.

**Bug Fix**: Previously, the "Export to plot.ly" link on
offline charts would export your figures to the
public plotly cloud, even if your `config_file`
(set with `plotly.tools.set_config_file` to the file
`~/.plotly/.config`) set `plotly_domain` to a plotly enterprise
URL like `https://plotly.acme.com`.

This is now fixed. Your graphs will be exported to your
`plotly_domain` if it is set.

Added
- The FigureFactory can now create annotated heatmaps with `.create_annotated_heatmap`. Check it out with:

import plotly.tools as tls
help(tls.FigureFactory.create_annotated_heatmap)

- The FigureFactory can now create tables with `.create_table`.

import plotly.tools as tls
help(tls.FigureFactory.create_table)

Added
- Ternary plots with support for scatter traces (trace type `scatterternary`) [390]

Fixed
- Toggling the visibility of `scatter3d` traces with `surfaceaxis` now works [405]
- `scatter3d` traces with `surfaceaxis` turned now feature real 3D opacity [408]
- `plotly_unhover` is now properly triggered over `pie` traces [407]
- Better grammar in `scatter` attribute descriptions [406]