v5.2.1

See the full release announcement here: https://community.plotly.com/t/announcing-plotly-py-5-2-1-trendlines-ecdf-plots-markers-on-lines-sharper-webgl-legend-group-titles/55681

Updated

• Updated Plotly.js to from version 2.2.0 to version 2.3.1. See the plotly.js CHANGELOG for more information. These changes are reflected in the auto-generated plotly.graph_objects module. Notable changes include:
  • new projections in geo subplots from a more recent version of d3-geo
  • new number-formatting directives from a more recent version of d3-format
  • Improve rendering of scattergl, splom and parcoords by implementing plotGlPixelRatio for those traces
  • performance improvements and bug fixes

Added

• Extra flags were added to the gapminder and stocks dataset to facilitate testing, documentation and demos #3305
• All line-like Plotly Express functions now accept markers argument to display markers, and all but line_mapbox accept symbol to map a field to the symbol attribute, similar to scatter-like functions #3326
• px.scatter and px.density_contours now support new trendline types 'rolling', 'expanding' and 'ewm' #2997
• px.scatter and px.density_contours now support new trendline_options argument to parameterize trendlines, with support for constant control and log-scaling in 'ols' and specification of the fraction used for 'lowess', as well as pass-through to Pandas for 'rolling', 'expanding' and 'ewm' #2997
• px.scatter and px.density_contours now support new trendline_scope argument that accepts the value 'overall' to request a single trendline for all traces, including across facets and animation frames #2997
• A new px.ecdf() function for Empirical Cumulative Distribution Functions #3330

Fixed

• Fixed regression introduced in version 5.0.0 where pandas/numpy arrays with dtype of Object were being converted to list values when added to a Figure (#3292, #3293)
• Better detection of Chrome and Chromium browsers in the Renderers framework, especially on Linux (#3278) with thanks to @c-chaitanya for the contribution

v5.1.0

• Updated Plotly.js from version 2.1.0 to version 2.2.0. See the plotly.js CHANGELOG for more information. These changes are reflected in the auto-generated plotly.graph_objects module. Notable changes include:
  • new <trace>.legendgrouptitle attribute for legend group titles
  • new %h text formatting directive for half-years
  • performance improvements and bug fixes

v5.0.0 - a new federated Jupyter extension, Icicle charts, and Bar chart patterns

See the full release announcement here: https://community.plotly.com/t/introducing-plotly-py-5-0-0-a-new-federated-jupyter-extension-icicle-charts-and-bar-chart-patterns/54039

Updated/Changed

Items in this section may be considered backwards-incompatible changes for the purposes of Semantic Versioning but we expect the vast majority of users to be able to upgrade to version 5.0 without encountering any issues.

• Dropped support for Python older than 3.6 #3160
• Updated Plotly.js to from version 1.58.4 to version 2.1.0. See the plotly.js CHANGELOG for more information. These changes are reflected in the auto-generated plotly.graph_objects module. Notable changes include:
  • dropped support for IE9 and IE10
  • dropped support for long-deprecated graph_objects like area traces and scatter.(t|r) and layout.(radial|angular)axis attributes
  • modebar no longer has hovermode or spikeline buttons by default (can be added back with fig.update_layout(modebar_add=["v1hovermode", "toggleSpikeLines"]))
  • "Aa" text no longer appears on legend items unless mode="text"
  • In bar traces, textposition now defaults to "auto"
  • Font size for legend and colorbar titles now matches axis title font size (slightly bigger)
  • deprecated heatmapgl, pointcloud traces as well as all transform attributes
• Combined plotlywidget into jupyterlab-plotly and packaged them as a federated extension #3142 with massive thanks to @fcollonval for the contribution
  • In addition to this change, large Plotly.js bundles are now lazily loaded on-demand by JupyterLab
• Plotly.js CDN url will now be versioned by default for HTML exports using include_plotlyjs='cdn' and for "connected" renderers. #2961 with thanks to @adehad for the contribution
• Recommending Kaleido by default over Orca #3094
• Replaced retrying dependency with tenacity #2911 with thanks to @jmsmdy for the contribution
• Plotly Express now always takes into account every value in category_orders when computing discrete mappings (color, symbol, line-dash, pattern-shapes) as well as facets, even those values which are absent in the data #3247

Added

• Additions due to bumping Plotly.js from 1.58.4 to 2.1.0 (see changelog):
  • New icicle trace type, with thanks to @Kully and @mtwichan of Zyphr for their contribution!
  • New marker.pattern options for bar-like trace types with thanks to @s417-lama for the contribution!
  • New legendrank attribute to control rank of traces within legends
• Plotly Express' px.bar(), px.histogram() and px.bar_polar() now support the pattern_shape argument #3252
• New Plotly Express px.icicle() function, with thanks to @Kully and @mtwichan of Zyphr for their contribution! #3256
• New functions in plotly.colors: get_colorscale() and sample_colorscale() #3136 and #3186 with thanks to @CarlAndersson for the contributions
• Faster JSON encoding when orjson is present #2955

Fixed

• Pandas and Numpy datetime serialization fixes #3022
• Fixed selected points of histograms in FigureWidget #2771 with thanks to @meffmadd for the contribution
• Static image export now honors layout.(width|height)#3240
• Improvements to "matplotlylib" conversion utility in plotly.tools.mpl_to_plotly() with thanks to @fdion #3143

v5.0.0rc2

Find out more about the RC process on our Forum: https://community.plotly.com/t/plotly-py-5-0-is-coming-soon-try-the-release-candidate-today/53470

Updated

• Updated Plotly.js to version 2.0.0. See the plotly.js 2.0.0 CHANGELOG for more information. These changes are reflected in the auto-generated plotly.graph_objects module. Notable changes include:
  • new marker.pattern options for bar-like trace types
  • dropped support for IE9 and IE10
  • dropped support for long-deprecated graph_objects like area traces and scatter.(t|r) and layout.(radial|angular)axis attributes
  • deprecated heatmapgl, pointcloud traces as well as all transform attributes
• Combined plotlywidget into jupyterlab-plotly and packaged them as a federated extension #3142 with massive thanks to @fcollonval for the contribution
• Plotly.js CDN url will now be versioned by default for HTML exports using include_plotlyjs='cdn' and for "connected" renderers. #2961 with thanks to @adehad for the contribution
• Dropped support for Python older than 3.6 #3160
• Recommending Kaleido by default over Orca #3094
• Replaced retrying dependency with tenacity #2911 with thanks to @jmsmdy for the contribution

Added

• New functions in plotly.colors: get_colorscale() and sample_colorscale() #3136 and #3186 with thanks to @CarlAndersson for the contributions
• Faster JSON encoding when orjson is present #2955

Fixed

• Pandas and Numpy datetime serialization fixes #3022
• Fixed selected points of histograms in FigureWidget #2771 with thanks to @meffmadd for the contribution

v5.0.0rc1

Find out more about the RC process on our Forum: https://community.plotly.com/t/plotly-py-5-0-is-coming-soon-try-the-release-candidate-today/53470

Updated

• Updated Plotly.js to version 2.0.0-rc.2. See the plotly.js 2.0.0 CHANGELOG for more information. These changes are reflected in the auto-generated plotly.graph_objects module. Notable changes include:
  • new marker.pattern options for bar-like trace types
  • dropped support for IE9 and IE10
  • dropped support for long-deprecated graph_objects like area traces and scatter.(t|r) and layout.(radial|angular)axis attributes
  • deprecated heatmapgl, pointcloud traces as well as all transform attributes
• Combined plotlywidget into jupyterlab-plotly and packaged them as a federated extension #3142 with massive thanks to @fcollonval for the contribution
• Plotly.js CDN url will now be versioned by default for HTML exports using include_plotlyjs='cdn' and for "connected" renderers. #2961 with thanks to @adehad for the contribution
• Dropped support for Python older than 3.6 #3160
• Recommending Kaleido by default over Orca #3094
• Replaced retrying dependency with tenacity #2911 with thanks to @jmsmdy for the contribution

Added

• New functions in plotly.colors: get_colorscale() and sample_colorscale() #3136 and #3186 with thanks to @CarlAndersson for the contributions
• Faster JSON encoding when orjson is present #2955

Fixed

• Pandas and Numpy datetime serialization fixes #3022
• Fixed selected points of histograms in FigureWidget #2771 with thanks to @meffmadd for the contribution

v4.14.3

• px.timeline() now allows hover_data formatting of start and end times 3018
• Small change to packaging of plotlywidget extension for JupyterLab 3 3021

v4.14.2

Updated

• JupyterLab extensions now compatible with JupyterLab 3.x 3016
• Updated Plotly.js to version 1.58.4. See the plotly.js CHANGELOG for more information. These changes are reflected in the auto-generated plotly.graph_objects module. Notable changes include:
  • fixes for rendering 3d plots on recent Safari versions
  • fixes to inside ticklabels
  • regression fixes

Fixed

• px.histogram() Y-axis labels now take into account histnorm and barnorm 2989
• px.histogram() marginal and facet_* now work correctly together 3014

v4.14.1

See the full announcement at https://community.plotly.com/t/announcing-plotly-py-4-14-faceted-and-animated-images-and-heatmaps-inside-tick-labels-better-axis-type-detection/48079

Updated

• Updated Plotly.js to version 1.58.2. See the plotly.js CHANGELOG for more information. These changes are reflected in the auto-generated plotly.graph_objects module. Notable changes include:
  • fixes for new ticklabelposition attribute
  • fixes for a regression related to treemaps in the previous version

v4.14.0 - Faceted and Animated Images and Heatmaps, Inside Tick Labels, Better Axis Type Detection

See the full announcement at https://community.plotly.com/t/announcing-plotly-py-4-14-faceted-and-animated-images-and-heatmaps-inside-tick-labels-better-axis-type-detection/48079

Added

• px.imshow now supports facet_col and animation_frame arguments for visualizing 3-d and 4-d images 2746
• px.defaults now supports color_discrete_map, symbol_map, line_dash_map, labels and category_orders as well as a .reset() method 2957

Fixed

• axes will now auto-type numeric strings as categorical data rather than linear in the default templates 2951

Updated

• Updated Plotly.js to version 1.58.1. See the plotly.js CHANGELOG for more information. These changes are reflected in the auto-generated plotly.graph_objects module. Notable changes include:
  • a new ticklabelposition attribute to enable positioning tick labels inside the plotting area
  • better support for scaleanchor and matches on cartesian axes for matched square subplots
  • a new autotypenumbers attribute which is now set to strict in the default templates
  • various fixes relating to automargins for small figures

v4.13.0 - Magical Error Messages and Documentation, plus Speedups and Faceted Maps

See the fully release announcement at https://community.plotly.com/t/announcing-plotly-py-4-13-magical-error-messages-and-documentation-plus-speedups-and-faceted-maps/47497

Added

• px.choropleth, px.scatter_geo and px.line_geo now support faceting as well as fitbounds and basemap_visible 2923
• px.scatter_geo and px.line_geo now support geojson/featureidkey input 2923
• px.scatter_geo now supports symbol 2923
• go.Figure now has a set_subplots method to set subplots on an already
  existing figure. 2866
• Added Turbo colorscale and fancier swatch display functions
  2882
• A utility function image_array_to_data_uri has been added in
  plotly.utils, in order to transform NumPy arrays to data b64 URIs (which
  can be passed to the source parameter of go.Image, or to layout images).
  2879
• the selector argument to updater/selector functions now accepts ints and strs
  2894

Updated

• the JSON serialization of plotly figures has been accelerated thanks to a
  different handling of Infinity and NaN values. For example, a figure with a
  1000x1000 Heatmap should now serialize 2x faster. 2880
• Coding mistakes with "magic underscores" now return significantly more ergonomic error
  messages 2843
• Error messages related to impossible subplot geometries are now much more helpful
  2897

Fixed

• px.scatter_geo support for text is fixed 2923
• the x and y parameters of px.imshow are now used also in the case where
  an Image trace is used (for RGB data or with binary_string=True). However,
  only numerical values are accepted (while the Heatmap trace allows date or
  string values for x and y). 2761