Version 2.8

PlotPy Version 2.8.4 (2026-02-14)

💥 New features:

• Added official support for Python 3.14

🛠️ Bug fixes:

• Fixed PySide6 compatibility issues causing segfaults in the test suite:

  • Use object instead of C++ type strings (e.g., "QMouseEvent", "QEvent", "QPointF") in Signal declarations — PySide6 segfaults with C++ type name strings

  • Check QObject validity via is_qobject_valid() before accessing widgets in __del__, closeEvent, and panel close operations — PySide6 segfaults on deleted C++ objects instead of raising RuntimeError

  • Restructure BaseSyncPlot.__init__ to defer widget operations until after Qt __init__ completes — PySide6 requires __init__ to have fully completed before the widget can be used as a parent

  • Replace deprecated exec_() calls with exec()

• Fixed test_multiline_tool failing with PyQt6 due to smaller canvas size:

  • The spiral test data started at t=0, placing the first two points so close together that they mapped to the same pixel on PyQt6’s smaller default canvas

  • Start the spiral at t=2π to ensure sufficient pixel spacing between consecutive points

• Fixed pytest running all tests when selecting a single test from VS Code:

  • Move plotpy from addopts to testpaths in [tool.pytest.ini_options]

🔧 Other changes:

• Add missing setuptools to requirements.txt (dev)

• Update GitHub Actions to use latest artifact upload and download versions

• Update cibuildwheel version to v3.3.1 for improved wheel building

• Add .venv* to .gitignore to exclude virtual environment files

PlotPy Version 2.8.3 (2025-12-18)

💥 New features:

• AnnotatedSegment new angle display:

  • Added angle display in annotation label, showing the segment’s angle relative to the horizontal direction (0° to 180°)

  • This implements the equivalent feature submitted in Pull Request #51 by user @deuterium33

🛠️ Bug fixes:

• Fixed IndexError when displaying images with a single row or column (e.g., SIF files with shape (1, N)):

  • The to_bins() function now handles single-element coordinate arrays by assuming a bin width of 1.0 centered on the point

  • Previously, loading such images in DataLab caused an IndexError: index 1 is out of bounds for axis 0 with size 1

• Fixed circle/ellipse shape drawing with non-uniform aspect ratios:

  • Axes were not perpendicular and did not connect to the ellipse edge when plot aspect ratio differed from 1.0

  • Now uses parametric ellipse drawing that correctly handles non-perpendicular axes in pixel space

  • The ellipse properly passes through all four handle points regardless of aspect ratio or rotation

• Fixed angle display range for AnnotatedObliqueRectangle and AnnotatedEllipse:

  • Angle is now displayed in the 0° to 180° range instead of -90° to 90° (the original implementation was also displaying 0° at vertical orientation -additionally to horizontal orientation- which was counter-intuitive)

  • This provides a more intuitive and consistent angle representation

PlotPy Version 2.8.2 (2025-11-10)

🛠️ Bug fixes:

• Fixed RuntimeWarning when changing masked image data type from float to integer:

  • MaskedImageItem.update_mask() now handles NaN and None filling_value gracefully

  • When converting to integer dtypes, NaN/None values are replaced with 0 instead of triggering numpy cast warnings

  • When converting to float dtypes, NaN is preserved as the fill value

  • Added comprehensive tests to validate dtype conversion scenarios

PlotPy Version 2.8.1 (2025-11-06)

🛠️ Bug fixes:

• Issue #50 - Fixed ImageStatsTool displaying “No available data” for XYImageItem and MaskedXYImageItem:

  • Added IExportROIImageItemType to XYImageItem.types() so that get_items_in_rectangle() can properly identify XY image items

  • Updated __implements__ tuples for consistency across XYImageItem, MaskedXYImageItem, and MaskedImageItem

  • The tool now correctly displays statistics for images with non-uniform coordinates

• Fixed snapshot tool failing with SystemError on XYImageItem and MaskedXYImageItem:

  • Fixed assemble_imageitems passing list instead of tuple to C extension function _scale_rect

  • Added missing export_roi method to XYImageItem to properly handle non-uniform coordinate transformations

  • Snapshots of XY images now render correctly instead of producing black images

PlotPy Version 2.8.0 (2025-10-24)

💥 New features / Enhancements:

• Curve fitting: added support for locked parameters:

  • New locked parameter in FitParam class to lock parameter values during automatic optimization

  • New locked field in FitParamDataSet to configure parameter locking via the settings dialog

  • When locked, parameters retain their manually-adjusted values during auto-fit

  • Visual indicators: locked parameters show a 🔒 emoji and are grayed out with disabled controls

  • All optimization algorithms (simplex, Powell, BFGS, L-BFGS-B, conjugate gradient, least squares) fully support locked parameters

  • Enables partial optimization workflows: fix well-determined parameters, optimize uncertain ones

  • Improves fit convergence by reducing problem dimensionality

• Configurable autoscale margin:

  • Added autoscale_margin_percent parameter to BasePlotOptions for intuitive percentage-based margin control

  • Users can now specify autoscale margins as percentages (e.g., 0.2 for 0.2%, 5.0 for 5%)

  • Replaces the previous decimal-based approach with more user-friendly percentage values

  • Default remains 0.2% (equivalent to previous 0.002) for backward compatibility

  • Includes validation to prevent unreasonable values (0-50% range)

• Image statistics tool improvements:

  • Enhanced get_stats function to display delta (Δ) values for coordinate ranges

  • Now shows Δx, Δy, and Δz values alongside the min/max ranges for better analysis

• Added optional “Axes” tab control in Parameters dialog:

  • New show_axes_tab option in BasePlotOptions and PlotOptions (default: True)

  • When set to False, the “Axes” tab is hidden from item parameter dialogs

  • This allows applications to provide their own axes management while using PlotPy

  • Can be configured during plot creation or changed at runtime using plot.set_show_axes_tab(False)

• Issue #45 - Add support for new curve Y-range cursor tool YRangeCursorTool:

  • This tool is similar to the existing CurveStatsTool, but it simply shows the Y-range values (min, max and interval).

  • It can be added to the plot widget using plot_widget.manager.add_tool(YRangeCursorTool)

• Update color configurations defaults for improved visibility

• Item list:

  • Added a new “Rename” context menu entry to rename the selected item

  • This entry is only available for editable items

• X and Y range selection items:

  • Added support for item title in parameters data set (RangeShapeParam)

  • This concerns the XRangeSelection and YRangeSelection items

• New annotated X and Y range selection items:

  • Added AnnotatedXRange and AnnotatedYRange items

  • These items provide X and Y range selection with an annotation label

  • They can be created using make.annotated_xrange and make.annotated_yrange functions

• New SyncPlotDialog class:

  • This class provides a dialog for displaying synchronized plots.

  • This is a complementary class to SyncPlotWindow, providing a modal dialog interface for synchronized plotting.

• Native datetime axis support:

  • Added BasePlot.set_axis_datetime() method to easily configure an axis for datetime display

  • Added BasePlot.set_axis_limits_from_datetime() method to set axis limits using datetime objects directly

  • Supports customizable datetime format strings using Python’s strftime format codes

  • Configurable label rotation and spacing for optimal display

  • Example: plot.set_axis_datetime("bottom", format="%H:%M:%S") for time-only display

  • Example: plot.set_axis_limits_from_datetime("bottom", dt1, dt2) to zoom to a specific time range

  • Added "datetime" as a valid scale type (alongside "lin" and "log") for axis configuration

  • Added datetime coordinate formatting support throughout PlotPy:

    • Cursor tools (VCursorTool, HCursorTool, XCursorTool) now display datetime-formatted X/Y coordinates

    • CurveStatsTool now displays datetime-formatted X coordinates for statistical computations

    • Marker labels automatically format coordinates as datetime when axis uses datetime scale

    • Coordinate display in the plot canvas now shows datetime format when appropriate

    • Refactored ObjectInfo base class to provide shared datetime formatting methods for code reuse

🧹 API cleanup: removed deprecated update methods (use update_item instead)

• Removed AnnotationParam.update_annotation method

• Removed AxesShapeParam.update_axes method

• Removed AxesParam.update_axes method

• Removed ImageAxesParam.update_axes method

• Removed LabelParam.update_label method

• Removed MarkerParam.update_marker method

• Removed RangeShapeParam.update_range method

• Removed ShapeParam.update_shape method

🛠️ Bug fixes:

• Fixed RuntimeError in SyncPlotWindow and SyncPlotDialog when closing widgets quickly:

  • Fixed “wrapped C/C++ object of type QwtScaleWidget has been deleted” error that occurred when widgets were closed before the deferred plot rescaling operation could complete

  • Replaced QTimer.singleShot() with controllable QTimer instances that can be stopped on widget close

  • Added handle_show_event() and handle_close_event() methods to BaseSyncPlot for proper timer lifecycle management

  • Refactored showEvent() and closeEvent() in both SyncPlotWindow and SyncPlotDialog to eliminate code duplication

  • Added early exit check in rescale_plots() to prevent execution if the timer has been stopped

  • This fix ensures clean widget shutdown and prevents Qt from attempting to access deleted C++ objects

• Cross-section panels: Fixed autoscaling logic in BaseCrossSectionPlot

  • Streamlined handling of autoscale_mode and lockscales options for consistent scaling behavior across all code paths

  • The update_plot() method now delegates all scaling logic to plot_axis_changed() to avoid code duplication and ensure consistency

  • Fixed issue where Y cross-section plots for rectangular images with non-uniform axes (e.g., Y = f(X)) were not properly scaled on initial display

  • The lockscales mode now correctly syncs the cross-section axis (CS_AXIS) to the image plot while autoscaling the intensity axis (Z_AXIS)

• Issue #49 - Fixed multiple coordinate handling bugs in XYImageItem:

  • Root cause: XYImageItem internally stores bin edges (length n+1) but several methods were incorrectly treating them as pixel centers (length n)

  • Fixed get_x_values() and get_y_values() to correctly compute and return pixel centers from stored bin edges: (edge[i] + edge[i+1]) / 2

  • Fixed get_pixel_coordinates() to correctly convert plot coordinates to pixel indices using searchsorted() with proper edge-to-index adjustment

  • Fixed get_plot_coordinates() to return pixel center coordinates instead of bin edge coordinates

  • Fixed get_closest_coordinates() to return pixel center coordinates instead of bin edge coordinates

  • Added comprehensive docstring documentation explaining that XYImageItem.x and XYImageItem.y store bin edges, not pixel centers

  • Removed redundant pixel centering code in CrossSectionItem.update_curve_data() that was working around these bugs

  • This fixes the reported issue where using cross-section tools progressively translated image data to the bottom-right corner

  • All coordinate-related methods now properly handle the bin edge vs pixel center distinction throughout the XYImageItem API

• Fixed index bounds calculation for image slicing compatibility:

  • Corrected the calculation of maximum indices in get_plot_coordinates to ensure proper bounds when using NumPy array slicing

  • Previously, the maximum indices were off by one, which could cause issues when extracting image data using the returned coordinates

  • Now returns indices that correctly align with Python/NumPy slicing conventions (e.g., [i1:i2+1, j1:j2+1])

  • This fixes an historic bug that could lead to off-by-one errors when users extracted image data using the coordinates provided by this function

• Fixed plot update after inserting a point using the EditPointTool on non-Windows platforms

• Issue #46 - Contrast adjustment with ‘Eliminate outliers’ failed for float images with high dynamic range

• Issue #29 - SelectTool: Selecting Another Shape Without Unselection

  • Fixed direct selection between different shapes without requiring intermediate click on empty space

  • Users can now click directly from one shape to another for immediate selection

  • Maintains all existing functionality including multi-selection (Ctrl+click), moving, and resizing

• Fixed ErrorBarCurveItem handling of all-NaN data:

  • Fixed ValueError: zero-size array to reduction operation minimum which has no identity when error bar curves contain only NaN values

  • Added proper checks in boundingRect() and draw() methods to handle empty arrays gracefully

  • Error bar curves with all-NaN data now fall back to parent class behavior instead of crashing

• Item list: refresh tree when item parameters are changed:

  • Added SIG_ITEM_PARAMETERS_CHANGED signal to BasePlot class

  • This signal is emitted when the parameters of an item are changed using the parameters dialog, or a specific tool (e.g. the colormap selection tool, or the lock/unlock tool for image items)

  • The signal is emitted with the item as argument

  • The ItemListWidget now listens to this signal and refreshes the item list accordingly

• Edit tools (Edit data, center image position):

  • Exclude read-only items from the list of editable items

  • It is no longer possible to use those tools on read-only items

• Marker items (markers, cursors):

  • Setting item movable state now also sets the resizable state:

    • The PlotPy event system won’t prevent the user from moving the item by dragging the handles if the item is just not movable: it has to be not resizable, which is not intuitive.

    • This is now fixed

• Range selection items (XRangeSelection, YRangeSelection):

  • Handles are now displayed only when the item is resizable

  • If the item is set as not resizable (using the set_resizable method), the handles will be hidden

• Fixed cursor label formatting error with percentage symbols:

  • Fixed ValueError: unsupported format character '=' (0x3d) when cursor labels contain percentage signs followed by format specifiers (e.g., “Crossing at 20.0% = %g”)

  • The issue occurred because old-style string formatting (label % value) treated the % in percentage displays as format specifiers

  • Added robust fallback mechanism that tests label format once during cursor creation and uses regex-based replacement when needed

  • Performance optimized: format validation is done once at cursor creation time, not on every callback execution

  • Affects vcursor, hcursor, and xcursor methods in CurveMarkerBuilder

Other changes:

• API breakage: renamed annotations AnnotatedShape.get_infos method to get_info

• Updated dependencies following the latest security advisories (NumPy >= 1.22)

• Added pre-commit hook to run ruff (both ruff check and ruff format) on commit

• Added missing build optional dependency to development dependencies in pyproject.toml

• Visual Studio Code tasks:

  • Major overhaul (cleanup and simplification)

  • Removal of no longer used batch files