napari: a multi-dimensional image viewer for Python

Le résumé fourni par la source

napari 0.7.1 Tue, Jun 16, 2026 We're happy to announce the release of napari 0.7.1! napari is a fast, interactive, multi-dimensional image viewer for Python. It's designed for browsing, annotating, and analyzing large multi-dimensional images. It's built on top of Qt (for the GUI), vispy (for performant GPU-based rendering), and the scientific Python stack (numpy, scipy). For more information, examples, and documentation, please visit our website, https://napari.org. napari follows EffVer (Intended Effort Versioning); this is a Meso release containing awesome new features, but some effort may be needed when updating previous projects to use this version. Highlights The napari 0.7.1 release includes many new features and improvements. Here are some of the highlights: Signed Windows bundle Starting with the napari 0.7.1 release, our bundle on Windows is now signed with a NumFOCUS certificate, like our macOS bundle has been. This means that you should be able to run napari without any warnings about the application being from an unknown publisher. This is an important step forward for our Windows users, as it enhances security and trust in our application, especially in managed IT environments where unsigned applications may be blocked by default. The certificate is issued for the NumFOCUS foundation, which supports napari and a large number of other open source projects. If you encounter any issues related to this change, please let us know! Selection of the rendered level for multiscale layers Prior to napari 0.7.1, when rendering a multiscale layer: in 3D display, napari would always render the lowest resolution level of the pyramid in 2D display, the pyramid level would be automatically selected based on the viewport Thanks to #8917, users can now select a multiscale level to render (check out the awesome new example!). Importantly, this allows choosing a higher resolution rendering than before in 3D, as long as the selected level can fit within GPU texture limits. Meanwhile, for 2D display, you can fix the resolution level, which can be useful for annotation or previewing data prior to an analysis or export step. You can set this resolution using the resolution widget in the layer controls: Colorbars for points layer The points allows coloring points based of their feature values; in napari 0.7.1 we added support for colorbars (#8624), so you can now easily see the mapping between feature values and colors, just like the colorbars for image colormapping. This is especially useful when you have a large number of points and want to quickly understand the distribution of feature values. Scalebar with units In previous versions of napari, if you added a scale bar using View > Scale Bar > Scale Bar visible, it was shown with no units. In napari 0.7.1 we now set default unit to pixel in #8900 and also add calculation of units for scale bar based on currently added layers in #8907 and #9007, if they have units set and are logically consistent across layers. We've also added a guide about unit and scale aware rendering (#1032). Lock layer to prevent accidental deletions In #8736 we added the initial implementation for a lock mechanism for layers. Now, when a layer is locked, it cannot be accidentally deleted or destructively modified. This is especially useful when you have a complex project with many layers and want to prevent accidental changes to important layers. Note: the layer controls are not affected by the lock at this time. In the future we plan to expand this feature to prevent not only deletion but also other modifications. New Features Add points layer face and border colorbar (#8624) UX: Add Layer.locked property to protect layers from accidental deletion (#8736) Improvements Example: Enhance theme sample widget with theme colors, widgets (#8662) Do not connect to children events of EventedDict items if there is no callback (#8673) ENH: For multiscale 2d: store materialized thumbnail_level (#8715) Canvas based font size (#8770) Enh, minor, builtins: use urllib (and defer) instead of requests if reading a remote script path (#8785) Perf: Defer importing scipy.spatial (#8789) Revert the switch from Welcome widget to Welcome overlay from #8117 (#8793) Update text in mac installer to improve UX (#8806) Fix: Use Python's tokenize script decoding for builtin reader (#8838) Add note for user how to reset the reader preference (#8848) Ensure all napari and plugin commands use . instead of : (#8883) Improve dask check to support more dask-backed array types (#8896) Fix floating widgets and use custom title bar in all cases (#8898) Set pixel as default scalebar unit (#8900) ScaleBar units from layers (#8907) Turn off contrast/color controls when surface has vertex_colors (#8909) Actually use theme type from npe contrib (#8915) Add multiscale level lock for scalar field layers (#8917) Do not add a new colormap if one already exists in napari (#8924) Adjust pt argument of font size increase/decrease based on platform (#8950) Improve sizing of dims ordering popup (#8952) Add float precision setting (#8975) typing: add type hints to napari/utils/shortcuts.py (#8983) Use QSS instead of hard-coding style (including font) for the notification action button (#9001) Use actual font-size in the keybind-clash popup (#9002) Performance ENH: For multiscale 2d: store materialized thumbnail_level (#8715) Enh, minor, builtins: use urllib (and defer) instead of requests if reading a remote script path (#8785) Perf: Defer importing scipy.spatial (#8789) Revert the switch from Welcome widget to Welcome overlay from #8117 (#8793) Improve dask check to support more dask-backed array types (#8896) Use pep562 to defer AVAILABLE_LABELS_COLORMAPS which imports skimage.color -> scipy.linalg (#8903) Bug Fixes Add max depth_value to avoid clipping volumes in 3d (#8809) (#8810) Fix TypeError with anisotropic data in 3D ray intersections (#8812) Reuse QMarginSlidersPopup between rightclicks (#8819) Fix: Use Python's tokenize script decoding for builtin reader (#8838) Fix: Blocks dims slider widget creation feedback to dims model (#8840) Fix: play button loop mode duplication (#8841) Improve dask check to support more dask-backed array types (#8896) Fix floating widgets and use custom title bar in all cases (#8898) Wrap Labels multiscale data in MultiScaleData object in setter (#8922) Fix Labels show_selected_label being silently dropped after color shuffle (#8947) Fix Volume visual crash when adding invisible scalar field in 3D (#8968) Auto-apply XCB/GLX workaround on Linux+Wayland at startup (#9009) Fix update of Shapes._value when remove shape (#9041) Set Shapes._valueto None, None on delete of shape (#9044) Build Tools ci(deps): bump aiohttp from 3.13.5 to 3.14.0 in /resources (#9028) Documentation Fix typo in 0.7.0 release notes (docs#967) Remove draft note from 0.7.0 release notes. (docs#968) Reorganize cards in homepage (docs#970) Rename "latest" version switch to "dev" (docs#971) Use stable version for dev docs bundle links (docs#973) Remove mention of unmaintained plugin in quick start (docs#976) Rename navbar entries (API and Contribute) (docs#978) Update team page (docs#980) Add "Edit on Github" to secondary sidebar (docs#982) Add deprecation warning section in contributing guide (docs#984) Enable search as you type and remove sidebar search (docs#989) Use flexible search field for navbar (docs#991) Add more info about example tags and _.py (docs#994) Remove napari-hub from navbar and add to sidebar (docs#995) Update deprecation warning guidance to use FutureWarning (docs#997) Update Jupyter notebook example screenshot (docs#1003) Update multiscale documentation to describe new level selection (docs#1004) Add release notes for 0.7.1a1 (docs#1009) Add sample databases (docs#1010) Minimal all-contributors setup (docs#1011) Add Carlos to Core Team page (docs#1012) Update release notes for 0.7.1 (docs#1019) Add Curtis and Brian to SC members (docs#1020) Move download links

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.