v3.2.0

[3.2.0] - 2018-09-05

This release introduces the long-anticipated ability to programmatically
export figures as high quality static images in both raster and vector
formats.

JupyterLab Versions

For use with JupyterLab, the following versions of the following packages
must be installed:

• Python Packages

  • plotly==3.2.0
  • ipywidgets>=7.2
  • notebook>=5.3
  • jupyterlab==0.34

• JupyterLab Extensions

  • [email protected]
  • @jupyter-widgets/[email protected]
  • @jupyterlab/[email protected]

Added

• plotly.js version 1.40.1, which introduces the following features:
  • Allow contour, contourcarpet and histogram2dcontour to have corresponding legend items using showlegend
    (plotly/plotly.js#2891,
    plotly/plotly.js#2914)
  • Add scatterpolar and scatterpolargl attributes r0, dr, theta0 and dtheta
    (plotly/plotly.js#2895)
  • Add layout attributes piecolorway and extendpiecolors for more control over pie colors
    (plotly/plotly.js#2870)
  • Add splom attribute dimensions[i].axis.type to easily override axis type in splom-generated axes
    (plotly/plotly.js#2899)
  • Add support for on-graph text in scatterpolargl traces
    (plotly/plotly.js#2895)
  • See the plotly.js CHANGELOG
    for bug fixes and more information.
• Support for offline static image export with the to_image and write_image
  functions in the new plotly.io package (#1120).
  • Note: Image export requires the plotly orca
    command line utility and the psutil Python package.
• New documentation sections covering Static Image Export
  and Orca Management
• Support for displaying FigureWidget instances in static contexts
  (e.g. nbviewer) just like the built-in ipywidgets
  (#1117)
• Full integration of the Cividis colorscale (#883)
• conda packaging
  • From here forward, new versions of plotly.py will be published to the plotly anaconda channel
    on the same day they are published to PyPI.
    (72ad0e4)
  • The README now includes conda installation instructions alongside the pip instructions.
  • In addition to the existing installation approaches, orca is now also available as a
    conda package from the plotly anaconda channel.

Updated

• Show traces at the top of the Gantt chart's colorbar (#1110)
• Significantly improved validation performance for numeric pandas Series objects (#1149)
• Specialize auto-generated docstrings for Python syntax
• More robust and specific logic for retrying requests to the plot.ly cloud service (#1146)
• Support basic authentication when using the streaming API behind a proxy server (#1133)

Fixed

• Validators for dash properties (e.g. scatter.line.dash) incorrectly rejected dash length lists (#1136)
• Annotated heatmap error when custom colorscale was specified (#1151)
• Incorrect deprecation warning for deprecated plotly.graph_objs.Annotations class (#1138)
• Harmless JavaScript console error when opening an html file produced by plotly.offline.plot (#1152)
• Incorrect validation errors when writing data to the streaming API (#1145)

v3.1.1

[3.1.1] - 2018-08-10

This release is a minor bug-fix update to version 3.1.0

JupyterLab Versions

For use with JupyterLab, the following versions of the following packages
must be installed:

• Python Packages

  • plotly==3.1.1
  • ipywidgets>=7.2
  • notebook>=5.3
  • jupyterlab==0.33

• JupyterLab Extensions

  • [email protected]
  • @jupyter-widgets/[email protected]
  • @jupyterlab/[email protected]

Updated

• Updated plotly.js to version 1.39.4.
  • This is a bug-fix release of plotly.js
  • See the plotly.js CHANGELOG for more information

Fixed

• Fixed error in validation of configkeys
  plotly/plotly.js#1065
• Fixed error in presentation of named colorscales
  plotly/plotly.js#1089
• Fixed numerical precision error when using plotly.tools.make_subplots
  to create figures with a large number of subplots
  plotly/plotly.js#1091
• Fixed problem that prevented the use of the .update method to initialize
  an array property (e.g. layout.shapes)
  plotly/plotly.js#1091
• Fixed FigureWidget problem causing scroll zoom on 3D plots to stutter
  plotly/plotly.js#1094
• Fixed invalid tickmode property in matplotlylib
  plotly/plotly.js#1101

v3.1.0

[3.1.0] - 2018-07-20

JupyterLab Versions

For use with JupyterLab, the following versions of the following packages
must be installed. See README.md for instructions.

• Python Packages

  • plotly==3.1.0
  • ipywidgets>=7.2
  • notebook>=5.3
  • jupyterlab==0.32.1

• JupyterLab Extensions

  • [email protected]
  • @jupyter-widgets/[email protected]
  • @jupyterlab/[email protected]

Updated

• Updated Plotly.js to version 1.39.2
• See highlights below
• See the plotly.js CHANGELOG for more information.

Added

• Added 3D streamtube traces
  plotly/plotly.js#2658
• Added support for on-graph text in scattergl traces
• Added gridshape attribute to polar subplots with values 'circular' (the default) and 'linear' (to draw polygon grids)
  plotly/plotly.js#2739

v3.0.2

[3.0.2] - 2018-07-17

This is a minor bug-fix release to 3.0.0

JupyterLab plotlywidget version: 0.1.1

Plotly.js version: 1.38.3

Fixed

• Several errors related to numbered subplot labels (e.g. xaxis2, polar3, etc.)
  GH1057
• Error where the v property was ignored in cone traces
  GH1060
• Assorted performance improvements when constructing graph objects
  GH1061

v3.0.0

[3.0.0] - 2018-07-05

This is a major version with many exciting updates. See the Introducing plotly.py 3.0.0 post for more information.

JupyterLab plotlywidget version: 0.1.1

Plotly.js version: 1.38.3

Added

• Full Jupyter ipywidgets integration with the new graph_objs.FigureWidget class
• FigureWidget figures can be updated interactively using property assignment syntax
• The full trace and layout API is generated from the plotly schema to provide a great experience for interactive use in the notebook
• Support for setting array properties as numpy arrays. When numpy arrays are used, ipywidgets binary serialization protocol is used to avoid converting these to JSON strings.
• Context manager API for animation. Run help(go.Figure().batch_animate) for the full doc string.
• Perform automatic retries when communicating with plot.ly services. This introduces a new required dependency on the retrying library.
• Improved data validation covering the full API with clear, informative error messages. This means that incorrect properties and/or values now always raise a ValueError with a description of the error, the invalid property, and the available properties on the level that it was placed in the graph object. Eg. go.Scatter(foo=123) raises a validation error. See https://plot.ly/python/reference/ for a reference to all valid properties and values in the Python API.
• Error message for plotly.figure_factory.create_choropleth is now helpful to Anaconda users who do not have the correct modules installed for the County Choropleth figure factory.

Changed / Deprecated

Please see the migration guid for a full list of the changes and deprecations in version 3.0.0

v2.0.0

Same as the pre-release, but this one corresponds to the 2.0.0 on PyPI and is the first official 2.x release. Please note the deprecation section from the changelog (copied below for convenience).

[2.0.0] - 2017-01-25

Changed

• plotly.exceptions.PlotlyRequestException is always raised for network
  failures. Previously either a PlotlyError, PlotlyRequestException, or a
  requests.exceptions.ReqestException could be raised. In particular, scripts
  which depend on try-except blocks containing network requests should be
  revisited.
• plotly.py:sign_in now validates to the plotly server specified in your
  config. If it cannot make a successful request, it raises a PlotlyError.
• plotly.figure_factory will raise an ImportError if numpy is not
  installed.
• plotly.figure_factory.create_violin() now has a rugplot parameter which
  determines whether or not a rugplot is draw beside each violin plot.

Deprecated

• plotly.tools.FigureFactory. Use plotly.figure_factory.*.
• (optional imports) plotly.tools._*_imported It was private anyhow, but now
  it's gone. (e.g., _numpy_imported)
• (plotly v2 helper) plotly.py._api_v2 It was private anyhow, but now it's
  gone.

v2.0.0-pre

[2.0.0]

Most notably, this changes how exceptions are handled and raised for all network requests. See CHANGELOG.md for more details (copied below for convenience).

Changed

• plotly.exceptions.PlotlyRequestException is always raised for network
  failures. Previously either a PlotlyError, PlotlyRequestException, or a
  requests.exceptions.ReqestException could be raised. In particular, scripts
  which depend on try-except blocks containing network requests should be
  revisited.
• plotly.py:sign_in now validates to the plotly server specified in your
  config. If it cannot make a successful request, it raises a PlotlyError.
• plotly.figure_factory will raise an ImportError if numpy is not
  installed.

Deprecated

• plotly.tools.FigureFactory. Use plotly.figure_factory.*.
• (optional imports) plotly.tools._*_imported It was private anyhow, but now
  it's gone. (e.g., _numpy_imported)
• (plotly v2 helper) plotly.py._api_v2 It was private anyhow, but now it's
  gone.

v1.13.0

[1.13.0] - 2016-01-17

Added

• Python 3.5 has been added as a tested environment for this package.

Updated

• plotly.plotly.create_animations and plotly.plotly.icreate_animations now return appropriate error messages if the response is not successful.
• frames are now integrated into GRAPH_REFERENCE and figure validation.

Changed

• The plot-schema from https://api.plot.ly/plot-schema is no longer updated on import.

1.6.12 - stable strict json encoder, ci fixes

• Previous version (1.6.11) of custom json encoder only supported python 2.7.
• Previous PATH env variable in circle needed adjusting to properly run python versions

1.6.8 - Remove fullData from IPython widget callbacks

#196