v0.13.1

v0.13.1

Bug Fixes

• Fixed a bug where a legend with blank title or text could result in a wrong justification.

• Fixed a bug where setting legend justification to a string i.e one of left, right, top, bottom, center had no effect.

v0.13.0

API Changes

• Requires python >= 3.9

• Using the print or repr functions to draw and show the plot has been deprecated. Use ggplot.show().

• The name of the calculated aesthetic of stat_function changed from y to fx.

• stat_ecdf has gained the pad parameter. The default is set to True, which pads the domain with -inf and inf so that the ECDF does not have discontinuities at the extremes. To get the behaviour, set pad to False. (#725)

• Removed the environment parameter from ggplot.

• When a ggplot object is the last in a jupyter cell, the output image will not be followed by string meta information about the figure/image.

  This will happen even if the backend is set to an interactive one.

  If you set the backend to an interactive one, use show to draw the plot.

• The default horizontal alignment for the plot title is center if it there is no subtitle. When there is a subtitle, the default is to have both aligned to the left.

• Some parameters that control the look and feel of guide_colorbar and guide_legend have been removed. For their place, the theme parameter has been introduced and it gives better control of the look and feel.

• Themeables legend_entry_spacing, legend_entry_spacing_x and legend_entry_spacing_y have been renamed to legend_key_spacing, legend_key_spacing_x and legend_key_spacing_y respectively

• facet_grid now accepts two parameters, rows and cols, to specify the variables along the two dimensions of the panels.

  The previous way of using a single parameter will still work if it is a string. For cases where the value was a list, e.g.

  facet_grid(facets=["col1", "col2"])

  can be rewritten as any one of;

  facet_grid("col1", "col2")
  facet_grid(["col1"], ["col2"])
  facet_grid(rows="col1", cols="col2")
  facet_grid(rows=["col1"], cols=["col2"])

• The facets parameter in facet_wrap has also changed to a more straight forward specification for the column variables. It expects a single string or a list/tuple of strings. (#545)

  However, the R-style formula strings are still silently accepted.

New

• Added symmetric logarithm transformation scales scale_x_symlog and scale_y_symlog

• Gained themeables

  • plot_margin_left
  • plot_margin_right
  • plot_margin_top
  • plot_margin_bottom

  to set the plot margin on each side independently.

• Gained themeables

  • axis_ticks_length_major_x
  • axis_ticks_length_major_y
  • axis_ticks_length_minor_x
  • axis_ticks_length_minor_y

  to control the x & y axis ticks length.

• Gained themeables

  • axis_ticks_pad_major_x
  • axis_ticks_pad_minor_y

  to control the x & y tick padding.

• Some parameters in element_text can now accept lists/tuples to set the values on individual text objects. (#724)

• Gained the option figure_format to set the format of the inline figures in an interactive session.
  e.g.

  from plotnine.options import set_option
  
  set_option("figure_format", "svg")

  will output all subsequent figures in svg format.

• Improved support for customizing guides/legends.

  • You can now add a frame to the colobar

  • You can now apply themes to individual guides, e.g.

    + guides(color=guide_colorbar(theme=theme_xkcd()))

    or

    + guides(color=guide_legend(theme=theme_minimal())

  • You can now place legends at more than one position around the panels. e.g.

    + guides(
        color=guide_colorbar(position="left"),
        fill=guide_legend(position="bottom"),
        size=guide_legend(position="bottom")
    )

    Puts the fill and size guides at the bottom, and the color guide on the left.

  • You can easily justify the legend along the four edges of the space around the panel area.

  • You can now place the legend inside the panels with easily control its location.

• The space parameter of facet_grid now responds to the values free, free_x and free_y to have panels whose relative width and/or height depends on the data range. (#545)

Bug Fixes

• Fixed handling of minor breaks in scale_continuous to accept numpy arrays and when the scale has a transform, that the minor breaks are supplied in user space and not transform space. Just like the major breaks. (#685)

• Fixed theming of axis_ticks with the size parameter. (#703)

• Fixed space handling around axis_label, axis_text and axis_ticks when the ticks are turned off.

• Fixed bug in geom_path where the lineend parameter was ignored. (#727)

• Fixed bug where theme(legend_background=element_blank()) messed up the position of the legend, instead of only removing the background.

• Fixed using facet_grid with a column named key. (#734)

• Fixed using legend when using an identity scale and reordering the breaks. (#735)

• Fixed drawing the upper outline of geom_ribbon. (#728)

• Fixed issue where the gridlines overlap the panel border. (#638)

Enhancements

• All __all__ variables are explicitly assigned to help static typecheckers infer module attributes. (#685)

• You can now pickle the drawn matplotlib figures. (#729)

• The facet titles in the strip_text are now better aligned are appear more centered withing the strip_background.

v0.12.4

This is a maintenance release with a few changes, see the changelog.

v0.12.3

A maintenance release. See changelog.

v0.12.2

This release is exactly the same as the last one (v0.12.1), except that it requires 0.9.0 <= mizani < 0.10.0, whereas the last one requires 0.9.0 <= mizani.

v0.12.1

(2023-05-09)

New Features

• A layout manager. Now you do not have to adjust spacing parameters to prevent objects around the panels from overlapping. Specifically, you can:

  1. Set the legend position to "top", "left" or "bottom"
  2. Use a large or multiline plot title
  3. Use a large or multiline plot caption
  4. Use facet_wrap with scales="free" scales="free_x" or `scales="free_y"

  You can now also align the plot_title, axis_title_x, axis_title_y and plot_caption with respect to the panels. Set these to "left", "right" & "center" for the horizontal flowing text. And "top", "right" & "center" for the vertical flowing text.

  Also, the size of the figure is exactly determined by the theme setting. For example, this:

  theme(figure_size=(8, 6), dpi=100)

  will create an 800px x 600px image.

• You can create a subtitle using labs and style it using the plot_subtitle parameter to theme.

theme(plot_subtitle=element_text(size=8))

Enhancements

• ggplot object gained a new method save_helper(). It gives you access to the matplotlib figure that will be saved to file.

• When plotting with an ipython interactive backend (e.g. in a jupyter notebook). The default image output is retina. You do not need to run this command.

  %config InlineBackend.figure_format = "retina"

  Plotnine still respects any values set the user.

• In an interactive setting, after drawing an image. The size of the figure (in pixels) is printed e.g. <Figure Size: (640 x 480)>. Previously, something like <ggplot: (336175301)> was printed.

API Changes

• __add__() and __iadd__() now accept the same types of objects. In this change __iadd__() has gained the ability to accept a list of objects.
• draw() no longer accepts the argument return_ggplot and the return value is always a matplolib figure.
• Themeables strip_margin, strip_margin_x and strip_margin_y have been renamed to strip_align(), strip_align_x() strip_align_y() repectively.
• subplots_adjust has been deprecated. You no longer need to use theme(subplots_adjust={"right": 0.85}) and the like to make space for the legend or text around the panels. In the future, this will through an error.
• Changed default font-family (san-serif) from DejaVu Sans to Helvetica.

Bug Fixes

• Fixed bug where a discrete position scale failed when mapping an empty variable. (#647)
• Fixed bug where facet_grid with a datetime column run into an exception. (#629)
• Fixed bug where geom_violin with facetting and "scales = free" did not work. (#655)
• Fixed bug in position_dodge2 to work with preserve="single" for geom_rect and any other geoms that accept a min-max range along the x-axis.
• Fixed legend for geom_point to accurately display transparent fill colors that have been manually set. (#665)
• Fix issue where infinite limits for the x or y scales lead to an exception. (#664)
• Fixed geom_text and geom_label when used with string formatting so that missing values are removed. (#651)

v0.10.1

See the changelog.

v0.10.0

See the changelog.

v0.9.0

See the changelog.

v0.8.0

API Changes

• How you map to calculated aesthetics has changed. Use the
  ~plotnine.aes.after_stat function. The old methods 'stat(name)'
  and '..name..' have been deprecated.

New Features

• You can now map to aesthetics at three different stages. See
  ~plotnine.aes.aes, ~plotnine.aes.after_stat,
  ~plotnine.aes.after_scale and ~plotnine.aes.stage.
• ~plotnine.geoms.geom_violin gained the a new parameter style
  with which you can draw half violin (density curve on one side and
  flat on the other).
• Added ~plotnine.geoms.geom_raster.
• geoms gained new parameter raster for the
  ~plotnine.layer.Layer. You can use it to rasterize any layer when
  the resulting plot is of vector format e.g. pdf.

Bug Fixes

• Fixed issue where some plots with a colorbar would fail for specific
  themes. (424)
• Fixed ~plotnine.geoms.geom_map to plot MultiLineString geom
  types.
• Fixed ~plotnine.geoms.geom_text to allow any order of mapping
  and data positional arguments.
• Fixed bug were the plotted image may have ignored theming that
  relied on some Matplotlib rcParams. (451)
• Fixed the weight aesthetic in ~plotnine.geoms.geom_boxplot,
  previously ignored it is now recognised. (438)
• Fixed ~plotnine.geoms.annotation_logticks and
  ~plotnine.geoms.annotation_stripes to work without global data and
  aesthetics. (469)
• Fix ~plotnine.scales.scale_shape_discrete when print many unfilled
  shapes not to assign the same shapes to more than one group. (473)
• Fixed bug in ~plotnine.stats.stat_ellipse where the center of the
  ellipse assuming a multivariate t-distribution was incorrectly
  calculated. (493)
• Fixed calculation of ndensity in ~plotnine.stats.stat_bin.
  (494)

Enhancements

• Manual scales now match the values of the breaks if the breaks are
  given. (445)

• Using print to show a ggplot object will not show the hash
  (<ggplot: ...>) anymore. There is now a difference between
  repr(p) and str(p). (453)

• Added option to for the base_family of a theme, now you can set it
  once with and have it be applied to all themes. (436)

  from plotnine.options import set_option
  set_option('base_family', 'Comic Sans MS')

• You can now add None to a ggplot, doing so returns a copy of the
  the ggplot object. (474)

• Better handling of multiline facet labels. (484)