v4.11.0

[4.11.0] - 2026-06-30

Added

• theme(): new tooltip_merge and tooltip_max_count parameters to combine the general tooltips of multiple targets into a single tooltip.

  See: example notebook.

• Text halos improve readability on varied backgrounds. New halo_width and halo_color parameters are supported in geom_text()/geom_text_repel() and by labels in corr_plot().

  See: example notebook.

Fixed

• Incorrect hjust/vjust/angle justification of vertical-axis tick labels (theme(axis_text_y=element_text(...)))
• Rotated vertical-axis tick labels could be wrongly dropped or kept by overlap-based break filtering
• Misaligned or mismatched axis tick labels when some breaks are dropped after layout

v4.11.0rc1

Updated version v4.11.0rc1

v4.10.1

[4.10.1] - 2026-05-20

Fixed

• facet_wrap(), error when ncol or nrow parameter not specified: 'ncol' must be positive, was null

v4.10.0

[4.10.0] - 2026-05-20

Added

• ggdeck() — a new function for overlaying multiple independent plots in a single unified view.

  In a typical scenario, one axis is shared by all plots in the deck, enabling dual-axis effects and multivariate comparisons.

  See examples:

  • dual-axis effect
  • multivariate comparison

• Datetime: %f formatting pattern — support for milliseconds when formatting datetime values, zero-padded to 3 digits [#1482].

  See: example notebook.

• In legends:

  • override_aes in guide_legend() can now customize filled 2D legend keys:

    • size controls the key border width;
    • width and height control the relative key size.

    See: example notebook.

• Indication of removed records [LPK-81], [#686].

  When records in data are dropped by active sampling or because they contain missing or out-of-bounds values, the user is now informed of the number of dropped records and the reason they were dropped.

  The new na_rm parameter in geom_xxx() and stat_xxx() functions controls the display of such messages:

  • na_rm=false (default) — records are removed and messages are shown;
  • na_rm=true — records are removed silently.

  See: example notebook.

• Alpha (transparency) component:

  • Hex colors accept #RRGGBBAA or #RGBA notation.
  • Colors accept an opacity suffix in the form "named color / opacity", for example "steelblue / 0.35".

  See: example notebook.

• Facet strip labels now honor angle in element_text() for strip_text_* [#1383].

  Thanks to a contribution by tentrillion.

  See: example notebook.

• In geom_imshow(), new parameters for controlling the colorbar breaks and labels [#1486]:

  • breaks, labels, lablim, format

• geom_livemap(): the new interactive parameter controls map interactivity. Set interactive=False to disable panning and zooming and hide map controls.

  See: example notebook.

• Memo for Kotlin API users:

  • Added WasmJS support [LPK-296], [LPC-52].

Changed

• [BREAKING] Dropped support for Python 3.9 as it is in the "end-of-life" of its release cycle. The minimum supported Python version is now 3.10.

• Swing-Batik components are deprecated. Use the Swing/AWT components instead (doesn't affect Python users).

Fixed

• Add 'synchronized tooltips' feature [#1415].
• Alpha is not supported in element_text() [#1462].
• scale_alpha: conflict of constant and mapped values of alpha aesthetic [#706].
• geom_imshow(): should render transparency for NaNs when all other pixel values are identical [#1485].
• scale_color_gradient(): guide='legend' is rendered as a colorbar [#1489].

v4.10.0rc1

Updated version v4.10.0rc1

v4.9.0

[4.9.0] - 2026-03-11

Added

• Python 3.14 support.

• Python 3.14 free-threading support [#1454].

• Plot Annotations:

  • New labels parameter in geom_smooth() designed to display statistical summaries of the fitted model directly on the plot.
    This parameter accepts a smooth_labels() object, which provides access to model-specific variables like $R^2$, the regression equation and others.

    See: example notebook.

  • Plot tags. A tag can be specified via labs(tag=...) and styled using theme parameters [#1407].

    See: example notebook and updated plot layout scheme.

  • Plot tags customization parameters in theme():

    • plot_tag - sets the tag style via element_text().
    • plot_tag_location - specifies the area used for positioning the tag.
    • plot_tag_position - specifies the position of the tag within the selected area.
    • plot_tag_prefix - text added before the tag value.
    • plot_tag_suffix - text added after the tag value.

    See: example notebook.

• Geometries:

  • New geom_bracket(), geom_bracket_dodge() [#1114].

    See: example notebook.

  • geom_imshow():

    • Support for custom colormaps [#780].

    • New cguide parameter: use to customize the colorbar for grayscale images.

      See: example notebook.

• Color Scales:

  • New palette() method for color scales: generates a list of hex color codes that can be used with scale_color_manual() to maintain consistent colors across multiple plots [#1444].

    See: example notebook.

  • New overflow parameter in scale_color_brewer() / scale_fill_brewer(): controls how colors are generated when more colors are needed than the palette provides.
    Options: 'interpolate' ('i'), 'cycle' ('c'), 'generate' ('g').

    See: example notebook.

• Positional Scales:

  • New break_width parameter that specifies a fixed distance between axis breaks.

    See examples:

    • datetime scale
    • time (duration) scale
    • log10 scale

  • Support for axis minor ticks via axis_minor_ticks and axis_minor_ticks_length parameters in theme() [#1379].

    See: example notebook.

• gggrid(): interactive pan/zoom now propagates across subplots with shared axes (sharex/sharey) [#1413].

  See: example notebook.

Changed

• [BREAKING]: ColorBrewer palettes: changed default behavior when the requested number of colors exceeds the palette's maximum size.
  Now defaults to 'interpolate' for sequential/diverging palettes and 'generate' for qualitative palettes.
  Previously, depending on the palette type, this either resulted in duplicate colors or random additional colors.
  Use the new overflow parameter to explicitly control this behavior.

• Reduced import overhead by nearly two orders of magnitude (from ~1.2s down to ~0.02s) [#1469].

• Missing values in geom_area_ridges() create gaps in geometries instead of being interpolated over.

• Discrete color scales (Brewer, Manual) now produce a colorbar guide when used with continuous data.
  Previously they produced a legend guide regardless of the data type.

• Changes affecting users on the JVM platform:

  • Upgraded the Kotlin version to 2.2.20 (was 1.9.25).
  • New artifact for JVM Swing applications: org.jetbrains.lets-plot:lets-plot-swing.
    This artifact provides the SwingPlotPanel class, which can be used to display plots in Swing applications instead of the now-obsolete DefaultPlotPanelBatik.
    For details, see the jvm-swing-app example in the "lets-plot-mini-apps" repository.
  • [BREAKING]: Removed JavaFX artifacts.
    The org.jetbrains.lets-plot:lets-plot-jfx artifact is no longer available.
    Replace it with new org.jetbrains.lets-plot:lets-plot-swing dependency and use SwingPlotPanel instead of DefaultPlotPanelJfx.
    For details, see the jvm-javafx-app example in the "lets-plot-mini-apps" repository.
  • [BREAKING]: Removed plot-image-export module.
    The org.jetbrains.lets-plot:lets-plot-image-export artifact is no longer available.
    The PlotImageExport utility has been moved to the platf-awt module.
    The required org.jetbrains.lets-plot:platf-awt dependency is likely already present in your project.

Fixed

• Drop commons-io dependency [#1421].
• Unexpected replacement of double curly brackets with a single curly bracket [#1433].
• Upgrade to a newer version of ws [#1150].
• geom_imshow: unclear error message when mixing transparencies [#1088].
• geom_imshow and scale_y_reverse [#1210].
• Nice to be able to get a list of colors from a color scale object [#1444].
• Allow tooltips param to accept list [#1455].
• Allow grouped tooltips for plots with multiple univariate geoms [#1460].
• Fixed a regression in geom_text_repel() / geom_label_repel(): points with empty labels were incorrectly skipped
  before building the repel obstacle set, so they were not included in collision avoidance and labels could overlap
  dense point clusters.

v4.9.0rc1

Updated version v4.9.0rc1

v4.8.2

[4.8.2] - 2026-12-19

Fixed

• Hyperlinks support for lets-plot-compose.
• Colab no longer rendering lets-plot with empty setup HTML [#1436]

v4.8.2rc1

Updated version v4.8.2rc1

v4.8.1

[4.8.1] - 2025-12-01

Fixed

• saving plots as png takes a very long time [#1423].
• stat_ecdf() takes a very long time for even moderately sized datasets [#1424].
• lets-plot plots are not displayed in plots pane [PST-8170].
• inconsistencies in theme/flavor inheritance in gggrid() subplots.