@spatialdata/vis

High-level React-components and deck.gl layers for visualising SpatialData. This package is explicitly less stable and more experimental than the related core and react packages - the dependency on the viv/deck.gl/luma.gl stack means that installing it into an app (particularly one that already uses anything from this ecosystem) entails a certain amount of care around potential conflicts or breaking changes between versions.

It is hoped that it will be robust and of a high quality, but especially at the early stage of development, the goal is not to have a lean bundle size, guarantees of API stability between releases, etc.

It is used to provide working examples for displaying in the docs site, as well as a sample app primarily for prototyping while developing features.

Browser codec workers

SpatialCanvas enables the bundled zarrextra codec worker automatically in browser contexts. Apps using normal vis components can load JP2K (imagecodecs_jpeg2k) and OpenJPH HTJ2K (experimental.openjph_htj2k) backed Zarr images without passing a workerUrl or calling zarrextra/workers directly.

ensureCodecWorkers() is exported for hosts that want to activate this path before mounting UI. The helper is idempotent, so repeated calls do not replace the worker pool.

Local fixture demos

pnpm test:fixtures:generate:0.8.0
pnpm test:fixtures:generate:codecs
pnpm --filter @spatialdata/vis dev

• http://127.0.0.1:5173/headless loads the versioned blobs.zarr fixture.
• http://127.0.0.1:5173/codec loads local codec fixtures (JP2K smoke test, HTJ2K Mandelbulb volume, HTJ2K encode demo).