Version 0.2#

DataLab-Kernel Version 0.2.11 (2026-02-15)#

Fixed ROI labels showing generic “ROI” text instead of user-defined titles in both Plotly and Matplotlib backends — ROI shapes now display the title attribute when set, falling back to “ROI 1”, “ROI 2”, etc.
Fixed Plotly multi-signal overlay showing “new text” on the second ROI instead of a proper label (the annotation text was inadvertently set to None, triggering Plotly’s default placeholder)
Fixed missing ROI title annotations on image ROI shapes (rectangles, circles, polygons) in both Plotly and Matplotlib backends
Fixed non-uniform image detection using the wrong attribute name (uniform_coords instead of is_uniform_coords), which caused non-uniform coordinate images to render incorrectly

Revamped the Advanced Showcase notebook with more relevant examples: double- Gaussian peaks for FWHM demo, Butterworth bandpass filter for log-scale display, per-ROI statistics and centroid overlays in analysis results
Added requirements.txt with package dependencies
CI: PyPI deployment now gates on test suite passing before publishing

Multi-object plotting: plotter.plot() now accepts lists of signals or images, enabling side-by-side visualization of multiple objects in a single call
Analysis result display: Geometry and table results now display as separate HTML tables below the plot instead of cluttering the figure with on-plot text annotations — especially important when working with multiple ROIs
Consistent figure sizing: Introduced DEFAULT_PLOT_WIDTH constant for uniform figure dimensions across Plotly and Matplotlib backends
User-defined figure dimensions: Plotly plots now accept custom width and height keyword arguments for fine-grained control over figure sizing
Advanced plotting showcase notebook: Added a comprehensive notebook demonstrating advanced features: multiple curve styles, colormaps, ROI-based analysis, geometry/table results, and interactive Plotly visualization

Fixed legend display in Plotly plots — legends are now hidden by default for cleaner visualization
Fixed default Plotly signal figure dimensions to match Matplotlib’s aspect ratio
Fixed figure serialization performance in Plotly backend
Fixed image rendering edge cases in Matplotlib backend

Updated documentation with structured “Try it Online” section linking to both Quick Start and Advanced Showcase notebooks
Improved portability of development scripts by using relative paths

Added a new Plotly interactive visualization backend as an alternative to the existing Matplotlib static backend. Plotly figures provide rich interactivity (zoom, pan, hover tooltips) directly in Jupyter notebooks.
Install with pip install datalab-kernel[plotly].

When both Plotly and Matplotlib are available, Plotly is automatically preferred for its richer interactive experience.
Switch backends at runtime with plotter.set_backend("matplotlib") or plotter.set_backend("plotly").
Override the default backend via the DATALAB_PLOTTER_BACKEND environment variable.

Enhanced figure height customization for both backends, with updated documentation.

Plotly figures now produce proper MIME bundles for native Jupyter display integration.

Removed references to the legacy XML-RPC connection method; documentation now emphasizes the Web API as the primary (and recommended) connection method.
Updated README with visualization backend documentation, usage examples, and runtime switching instructions.
Updated online demo badge link.

Added plotly optional dependency group in pyproject.toml (pip install datalab-kernel[plotly]).

Enhanced wheel building script (serve_dev_wheels.py) to support package selection via --packages argument.
New VS Code tasks for interactive plot tests, dev wheel serving, and backend-specific testing.

Fixed test compatibility with Python 3.9 by using ExitStack instead of multiple mock.patch context managers.
Fixed test module shadowing issue by using mock.patch.object instead of mock.patch with string targets.
Fixed various Ruff and Pylint warnings across the codebase.
Fixed test workflow branch references for DataLab installation.

Added 65 tests for the Matplotlib backend covering all rendering features, backward-compatibility aliases, and edge cases.
Added 40 tests for the Plotly backend covering all rendering features and edge cases.
Added backend selection and resolution logic tests.
Added --gui interactive gallery viewers for visual test validation (tkinter for Matplotlib, tabbed HTML for Plotly).

These early development releases established the core architecture of DataLab-Kernel, including: