Add colour and palettes tutorial by timtreis · Pull Request #4 · scverse/spatialdata-plot-tutorials

timtreis · 2026-05-12T11:45:33Z

Summary

Adds tutorials/color_and_palette.ipynb (33 cells, executes end-to-end). The notebook covers:

The color= resolution pipeline (literal → column → .uns['<col>_colors']) and the deliberate shorthand-rejection layer.
Categorical vs continuous coloring.
The v0.3.0 groups / na_color behaviour change.
make_palette(n, palette=..., method=...) — including the 'contrast' and CVD-aware orderings.
make_palette_from_data(...) — dict round-trip into render_shapes.
A per-render_* parameter matrix.

Uses the synthetic blobs dataset and mutates it in place to add a 5-category cell_type on the table and a 3-category region_type on blobs_polygons so palette demos have enough levels.

Companion card and thumbnail in scverse/spatialdata-plot will land in a follow-up once this merges so the toctree link resolves.

Addresses scverse/spatialdata-plot#463. Tutorial sidesteps a labels-related gap in make_palette_from_data (tracked at scverse/spatialdata-plot#662); §5 uses blobs_polygons instead.

Tutorial covering the color resolution pipeline, v0.3.0 `groups` / `na_color` behaviour, and the `make_palette` / `make_palette_from_data` helpers. Uses the synthetic `blobs` dataset; mutates it in place to add a 5-category `cell_type` on the table and a 3-category `region_type` on `blobs_polygons` so palette demos have enough levels. Addresses scverse/spatialdata-plot#463.

review-notebook-app · 2026-05-12T11:45:41Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

Defines a default pixi environment with jupyterlab + ipykernel + nbconvert + watermark from conda-forge and spatialdata-plot + squidpy from PyPI. Tasks: `kernel-install` registers a Jupyter kernel named "sdata-plot-notebooks" pointing at the pixi env's Python; `lab` opens JupyterLab; `execute` re-runs every notebook in place. The existing `[exec]` pip extra is retained for users who prefer plain venv-based setups.

github-actions · 2026-05-12T12:01:12Z

📖 Docs preview: https://scverse.org/spatialdata-plot-notebooks/pr-4/gallery.html

Built from 8df0dcd; redeployed on every push.

Notebook reworked end-to-end; thumbnail regenerated from the publication-style final figure (gene-expression-coloured spots on the H&E tissue). Path in zarr-store output redacted to the relative form used elsewhere in the repo.

`>=0.3.4` includes recent fixes used by the tutorials. Adds `.pixi/` (env cache) and `pixi.lock` to `.gitignore`; pixi solves dependencies fresh per contributor, no need to commit the lockfile.

CI uses the default python3 kernel installed by pip; the local pixi kernel name baked into the notebook by an interactive re-execution does not exist in the CI environment, causing jupyter_client.kernelspec.NoSuchKernel.

Pin nbconvert to CI's python3 kernel via --ExecutePreprocessor.kernel_name=python3, so notebooks committed with a contributor-local kernel name (e.g. pixi's sdata-plot-notebooks kernel from kernel-install) still execute correctly in CI.

timtreis · 2026-05-12T11:45:33Z

Summary

Adds tutorials/color_and_palette.ipynb (33 cells, executes end-to-end). The notebook covers:

The color= resolution pipeline (literal → column → .uns['<col>_colors']) and the deliberate shorthand-rejection layer.
Categorical vs continuous coloring.
The v0.3.0 groups / na_color behaviour change.
make_palette(n, palette=..., method=...) — including the 'contrast' and CVD-aware orderings.
make_palette_from_data(...) — dict round-trip into render_shapes.
A per-render_* parameter matrix.

Uses the synthetic blobs dataset and mutates it in place to add a 5-category cell_type on the table and a 3-category region_type on blobs_polygons so palette demos have enough levels.

Companion card and thumbnail in scverse/spatialdata-plot will land in a follow-up once this merges so the toctree link resolves.

Addresses scverse/spatialdata-plot#463. Tutorial sidesteps a labels-related gap in make_palette_from_data (tracked at scverse/spatialdata-plot#662); §5 uses blobs_polygons instead.

Tutorial covering the color resolution pipeline, v0.3.0 `groups` / `na_color` behaviour, and the `make_palette` / `make_palette_from_data` helpers. Uses the synthetic `blobs` dataset; mutates it in place to add a 5-category `cell_type` on the table and a 3-category `region_type` on `blobs_polygons` so palette demos have enough levels. Addresses scverse/spatialdata-plot#463.

review-notebook-app · 2026-05-12T11:45:41Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

Defines a default pixi environment with jupyterlab + ipykernel + nbconvert + watermark from conda-forge and spatialdata-plot + squidpy from PyPI. Tasks: `kernel-install` registers a Jupyter kernel named "sdata-plot-notebooks" pointing at the pixi env's Python; `lab` opens JupyterLab; `execute` re-runs every notebook in place. The existing `[exec]` pip extra is retained for users who prefer plain venv-based setups.

github-actions · 2026-05-12T12:01:12Z

📖 Docs preview: https://scverse.org/spatialdata-plot-notebooks/pr-4/gallery.html

Built from 8df0dcd; redeployed on every push.

Notebook reworked end-to-end; thumbnail regenerated from the publication-style final figure (gene-expression-coloured spots on the H&E tissue). Path in zarr-store output redacted to the relative form used elsewhere in the repo.

`>=0.3.4` includes recent fixes used by the tutorials. Adds `.pixi/` (env cache) and `pixi.lock` to `.gitignore`; pixi solves dependencies fresh per contributor, no need to commit the lockfile.

CI uses the default python3 kernel installed by pip; the local pixi kernel name baked into the notebook by an interactive re-execution does not exist in the CI environment, causing jupyter_client.kernelspec.NoSuchKernel.

Pin nbconvert to CI's python3 kernel via --ExecutePreprocessor.kernel_name=python3, so notebooks committed with a contributor-local kernel name (e.g. pixi's sdata-plot-notebooks kernel from kernel-install) still execute correctly in CI.

timtreis · 2026-05-12T11:45:33Z

Summary

Adds tutorials/color_and_palette.ipynb (33 cells, executes end-to-end). The notebook covers:

The color= resolution pipeline (literal → column → .uns['<col>_colors']) and the deliberate shorthand-rejection layer.
Categorical vs continuous coloring.
The v0.3.0 groups / na_color behaviour change.
make_palette(n, palette=..., method=...) — including the 'contrast' and CVD-aware orderings.
make_palette_from_data(...) — dict round-trip into render_shapes.
A per-render_* parameter matrix.

Uses the synthetic blobs dataset and mutates it in place to add a 5-category cell_type on the table and a 3-category region_type on blobs_polygons so palette demos have enough levels.

Companion card and thumbnail in scverse/spatialdata-plot will land in a follow-up once this merges so the toctree link resolves.

Addresses scverse/spatialdata-plot#463. Tutorial sidesteps a labels-related gap in make_palette_from_data (tracked at scverse/spatialdata-plot#662); §5 uses blobs_polygons instead.

Tutorial covering the color resolution pipeline, v0.3.0 `groups` / `na_color` behaviour, and the `make_palette` / `make_palette_from_data` helpers. Uses the synthetic `blobs` dataset; mutates it in place to add a 5-category `cell_type` on the table and a 3-category `region_type` on `blobs_polygons` so palette demos have enough levels. Addresses scverse/spatialdata-plot#463.

review-notebook-app · 2026-05-12T11:45:41Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

Defines a default pixi environment with jupyterlab + ipykernel + nbconvert + watermark from conda-forge and spatialdata-plot + squidpy from PyPI. Tasks: `kernel-install` registers a Jupyter kernel named "sdata-plot-notebooks" pointing at the pixi env's Python; `lab` opens JupyterLab; `execute` re-runs every notebook in place. The existing `[exec]` pip extra is retained for users who prefer plain venv-based setups.

github-actions · 2026-05-12T12:01:12Z

📖 Docs preview: https://scverse.org/spatialdata-plot-notebooks/pr-4/gallery.html

Built from 8df0dcd; redeployed on every push.

Notebook reworked end-to-end; thumbnail regenerated from the publication-style final figure (gene-expression-coloured spots on the H&E tissue). Path in zarr-store output redacted to the relative form used elsewhere in the repo.

`>=0.3.4` includes recent fixes used by the tutorials. Adds `.pixi/` (env cache) and `pixi.lock` to `.gitignore`; pixi solves dependencies fresh per contributor, no need to commit the lockfile.

CI uses the default python3 kernel installed by pip; the local pixi kernel name baked into the notebook by an interactive re-execution does not exist in the CI environment, causing jupyter_client.kernelspec.NoSuchKernel.

Pin nbconvert to CI's python3 kernel via --ExecutePreprocessor.kernel_name=python3, so notebooks committed with a contributor-local kernel name (e.g. pixi's sdata-plot-notebooks kernel from kernel-install) still execute correctly in CI.

timtreis · 2026-05-12T11:45:33Z

Summary

Adds tutorials/color_and_palette.ipynb (33 cells, executes end-to-end). The notebook covers:

The color= resolution pipeline (literal → column → .uns['<col>_colors']) and the deliberate shorthand-rejection layer.
Categorical vs continuous coloring.
The v0.3.0 groups / na_color behaviour change.
make_palette(n, palette=..., method=...) — including the 'contrast' and CVD-aware orderings.
make_palette_from_data(...) — dict round-trip into render_shapes.
A per-render_* parameter matrix.

Uses the synthetic blobs dataset and mutates it in place to add a 5-category cell_type on the table and a 3-category region_type on blobs_polygons so palette demos have enough levels.

Companion card and thumbnail in scverse/spatialdata-plot will land in a follow-up once this merges so the toctree link resolves.

Addresses scverse/spatialdata-plot#463. Tutorial sidesteps a labels-related gap in make_palette_from_data (tracked at scverse/spatialdata-plot#662); §5 uses blobs_polygons instead.

Tutorial covering the color resolution pipeline, v0.3.0 `groups` / `na_color` behaviour, and the `make_palette` / `make_palette_from_data` helpers. Uses the synthetic `blobs` dataset; mutates it in place to add a 5-category `cell_type` on the table and a 3-category `region_type` on `blobs_polygons` so palette demos have enough levels. Addresses scverse/spatialdata-plot#463.

review-notebook-app · 2026-05-12T11:45:41Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

Defines a default pixi environment with jupyterlab + ipykernel + nbconvert + watermark from conda-forge and spatialdata-plot + squidpy from PyPI. Tasks: `kernel-install` registers a Jupyter kernel named "sdata-plot-notebooks" pointing at the pixi env's Python; `lab` opens JupyterLab; `execute` re-runs every notebook in place. The existing `[exec]` pip extra is retained for users who prefer plain venv-based setups.

github-actions · 2026-05-12T12:01:12Z

📖 Docs preview: https://scverse.org/spatialdata-plot-notebooks/pr-4/gallery.html

Built from 8df0dcd; redeployed on every push.

Notebook reworked end-to-end; thumbnail regenerated from the publication-style final figure (gene-expression-coloured spots on the H&E tissue). Path in zarr-store output redacted to the relative form used elsewhere in the repo.

`>=0.3.4` includes recent fixes used by the tutorials. Adds `.pixi/` (env cache) and `pixi.lock` to `.gitignore`; pixi solves dependencies fresh per contributor, no need to commit the lockfile.

CI uses the default python3 kernel installed by pip; the local pixi kernel name baked into the notebook by an interactive re-execution does not exist in the CI environment, causing jupyter_client.kernelspec.NoSuchKernel.

Pin nbconvert to CI's python3 kernel via --ExecutePreprocessor.kernel_name=python3, so notebooks committed with a contributor-local kernel name (e.g. pixi's sdata-plot-notebooks kernel from kernel-install) still execute correctly in CI.

timtreis · 2026-05-12T11:45:33Z

Summary

Adds tutorials/color_and_palette.ipynb (33 cells, executes end-to-end). The notebook covers:

The color= resolution pipeline (literal → column → .uns['<col>_colors']) and the deliberate shorthand-rejection layer.
Categorical vs continuous coloring.
The v0.3.0 groups / na_color behaviour change.
make_palette(n, palette=..., method=...) — including the 'contrast' and CVD-aware orderings.
make_palette_from_data(...) — dict round-trip into render_shapes.
A per-render_* parameter matrix.

Uses the synthetic blobs dataset and mutates it in place to add a 5-category cell_type on the table and a 3-category region_type on blobs_polygons so palette demos have enough levels.

Companion card and thumbnail in scverse/spatialdata-plot will land in a follow-up once this merges so the toctree link resolves.

Addresses scverse/spatialdata-plot#463. Tutorial sidesteps a labels-related gap in make_palette_from_data (tracked at scverse/spatialdata-plot#662); §5 uses blobs_polygons instead.

Tutorial covering the color resolution pipeline, v0.3.0 `groups` / `na_color` behaviour, and the `make_palette` / `make_palette_from_data` helpers. Uses the synthetic `blobs` dataset; mutates it in place to add a 5-category `cell_type` on the table and a 3-category `region_type` on `blobs_polygons` so palette demos have enough levels. Addresses scverse/spatialdata-plot#463.

review-notebook-app · 2026-05-12T11:45:41Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

Defines a default pixi environment with jupyterlab + ipykernel + nbconvert + watermark from conda-forge and spatialdata-plot + squidpy from PyPI. Tasks: `kernel-install` registers a Jupyter kernel named "sdata-plot-notebooks" pointing at the pixi env's Python; `lab` opens JupyterLab; `execute` re-runs every notebook in place. The existing `[exec]` pip extra is retained for users who prefer plain venv-based setups.

github-actions · 2026-05-12T12:01:12Z

📖 Docs preview: https://scverse.org/spatialdata-plot-notebooks/pr-4/gallery.html

Built from 8df0dcd; redeployed on every push.

Notebook reworked end-to-end; thumbnail regenerated from the publication-style final figure (gene-expression-coloured spots on the H&E tissue). Path in zarr-store output redacted to the relative form used elsewhere in the repo.

`>=0.3.4` includes recent fixes used by the tutorials. Adds `.pixi/` (env cache) and `pixi.lock` to `.gitignore`; pixi solves dependencies fresh per contributor, no need to commit the lockfile.

CI uses the default python3 kernel installed by pip; the local pixi kernel name baked into the notebook by an interactive re-execution does not exist in the CI environment, causing jupyter_client.kernelspec.NoSuchKernel.

Pin nbconvert to CI's python3 kernel via --ExecutePreprocessor.kernel_name=python3, so notebooks committed with a contributor-local kernel name (e.g. pixi's sdata-plot-notebooks kernel from kernel-install) still execute correctly in CI.

timtreis · 2026-05-12T11:45:33Z

Summary

Adds tutorials/color_and_palette.ipynb (33 cells, executes end-to-end). The notebook covers:

The color= resolution pipeline (literal → column → .uns['<col>_colors']) and the deliberate shorthand-rejection layer.
Categorical vs continuous coloring.
The v0.3.0 groups / na_color behaviour change.
make_palette(n, palette=..., method=...) — including the 'contrast' and CVD-aware orderings.
make_palette_from_data(...) — dict round-trip into render_shapes.
A per-render_* parameter matrix.

Uses the synthetic blobs dataset and mutates it in place to add a 5-category cell_type on the table and a 3-category region_type on blobs_polygons so palette demos have enough levels.

Companion card and thumbnail in scverse/spatialdata-plot will land in a follow-up once this merges so the toctree link resolves.

Addresses scverse/spatialdata-plot#463. Tutorial sidesteps a labels-related gap in make_palette_from_data (tracked at scverse/spatialdata-plot#662); §5 uses blobs_polygons instead.

Tutorial covering the color resolution pipeline, v0.3.0 `groups` / `na_color` behaviour, and the `make_palette` / `make_palette_from_data` helpers. Uses the synthetic `blobs` dataset; mutates it in place to add a 5-category `cell_type` on the table and a 3-category `region_type` on `blobs_polygons` so palette demos have enough levels. Addresses scverse/spatialdata-plot#463.

review-notebook-app · 2026-05-12T11:45:41Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

Defines a default pixi environment with jupyterlab + ipykernel + nbconvert + watermark from conda-forge and spatialdata-plot + squidpy from PyPI. Tasks: `kernel-install` registers a Jupyter kernel named "sdata-plot-notebooks" pointing at the pixi env's Python; `lab` opens JupyterLab; `execute` re-runs every notebook in place. The existing `[exec]` pip extra is retained for users who prefer plain venv-based setups.

github-actions · 2026-05-12T12:01:12Z

📖 Docs preview: https://scverse.org/spatialdata-plot-notebooks/pr-4/gallery.html

Built from 8df0dcd; redeployed on every push.

Notebook reworked end-to-end; thumbnail regenerated from the publication-style final figure (gene-expression-coloured spots on the H&E tissue). Path in zarr-store output redacted to the relative form used elsewhere in the repo.

`>=0.3.4` includes recent fixes used by the tutorials. Adds `.pixi/` (env cache) and `pixi.lock` to `.gitignore`; pixi solves dependencies fresh per contributor, no need to commit the lockfile.

CI uses the default python3 kernel installed by pip; the local pixi kernel name baked into the notebook by an interactive re-execution does not exist in the CI environment, causing jupyter_client.kernelspec.NoSuchKernel.

Pin nbconvert to CI's python3 kernel via --ExecutePreprocessor.kernel_name=python3, so notebooks committed with a contributor-local kernel name (e.g. pixi's sdata-plot-notebooks kernel from kernel-install) still execute correctly in CI.

timtreis · 2026-05-12T11:45:33Z

Summary

Adds tutorials/color_and_palette.ipynb (33 cells, executes end-to-end). The notebook covers:

The color= resolution pipeline (literal → column → .uns['<col>_colors']) and the deliberate shorthand-rejection layer.
Categorical vs continuous coloring.
The v0.3.0 groups / na_color behaviour change.
make_palette(n, palette=..., method=...) — including the 'contrast' and CVD-aware orderings.
make_palette_from_data(...) — dict round-trip into render_shapes.
A per-render_* parameter matrix.

Uses the synthetic blobs dataset and mutates it in place to add a 5-category cell_type on the table and a 3-category region_type on blobs_polygons so palette demos have enough levels.

Companion card and thumbnail in scverse/spatialdata-plot will land in a follow-up once this merges so the toctree link resolves.

Addresses scverse/spatialdata-plot#463. Tutorial sidesteps a labels-related gap in make_palette_from_data (tracked at scverse/spatialdata-plot#662); §5 uses blobs_polygons instead.

Tutorial covering the color resolution pipeline, v0.3.0 `groups` / `na_color` behaviour, and the `make_palette` / `make_palette_from_data` helpers. Uses the synthetic `blobs` dataset; mutates it in place to add a 5-category `cell_type` on the table and a 3-category `region_type` on `blobs_polygons` so palette demos have enough levels. Addresses scverse/spatialdata-plot#463.

review-notebook-app · 2026-05-12T11:45:41Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

Defines a default pixi environment with jupyterlab + ipykernel + nbconvert + watermark from conda-forge and spatialdata-plot + squidpy from PyPI. Tasks: `kernel-install` registers a Jupyter kernel named "sdata-plot-notebooks" pointing at the pixi env's Python; `lab` opens JupyterLab; `execute` re-runs every notebook in place. The existing `[exec]` pip extra is retained for users who prefer plain venv-based setups.

github-actions · 2026-05-12T12:01:12Z

📖 Docs preview: https://scverse.org/spatialdata-plot-notebooks/pr-4/gallery.html

Built from 8df0dcd; redeployed on every push.

Notebook reworked end-to-end; thumbnail regenerated from the publication-style final figure (gene-expression-coloured spots on the H&E tissue). Path in zarr-store output redacted to the relative form used elsewhere in the repo.

`>=0.3.4` includes recent fixes used by the tutorials. Adds `.pixi/` (env cache) and `pixi.lock` to `.gitignore`; pixi solves dependencies fresh per contributor, no need to commit the lockfile.

CI uses the default python3 kernel installed by pip; the local pixi kernel name baked into the notebook by an interactive re-execution does not exist in the CI environment, causing jupyter_client.kernelspec.NoSuchKernel.

Pin nbconvert to CI's python3 kernel via --ExecutePreprocessor.kernel_name=python3, so notebooks committed with a contributor-local kernel name (e.g. pixi's sdata-plot-notebooks kernel from kernel-install) still execute correctly in CI.

timtreis · 2026-05-12T11:45:33Z

Summary

Adds tutorials/color_and_palette.ipynb (33 cells, executes end-to-end). The notebook covers:

The color= resolution pipeline (literal → column → .uns['<col>_colors']) and the deliberate shorthand-rejection layer.
Categorical vs continuous coloring.
The v0.3.0 groups / na_color behaviour change.
make_palette(n, palette=..., method=...) — including the 'contrast' and CVD-aware orderings.
make_palette_from_data(...) — dict round-trip into render_shapes.
A per-render_* parameter matrix.

Uses the synthetic blobs dataset and mutates it in place to add a 5-category cell_type on the table and a 3-category region_type on blobs_polygons so palette demos have enough levels.

Companion card and thumbnail in scverse/spatialdata-plot will land in a follow-up once this merges so the toctree link resolves.

Addresses scverse/spatialdata-plot#463. Tutorial sidesteps a labels-related gap in make_palette_from_data (tracked at scverse/spatialdata-plot#662); §5 uses blobs_polygons instead.

Tutorial covering the color resolution pipeline, v0.3.0 `groups` / `na_color` behaviour, and the `make_palette` / `make_palette_from_data` helpers. Uses the synthetic `blobs` dataset; mutates it in place to add a 5-category `cell_type` on the table and a 3-category `region_type` on `blobs_polygons` so palette demos have enough levels. Addresses scverse/spatialdata-plot#463.

review-notebook-app · 2026-05-12T11:45:41Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

Defines a default pixi environment with jupyterlab + ipykernel + nbconvert + watermark from conda-forge and spatialdata-plot + squidpy from PyPI. Tasks: `kernel-install` registers a Jupyter kernel named "sdata-plot-notebooks" pointing at the pixi env's Python; `lab` opens JupyterLab; `execute` re-runs every notebook in place. The existing `[exec]` pip extra is retained for users who prefer plain venv-based setups.

github-actions · 2026-05-12T12:01:12Z

📖 Docs preview: https://scverse.org/spatialdata-plot-notebooks/pr-4/gallery.html

Built from 8df0dcd; redeployed on every push.

Notebook reworked end-to-end; thumbnail regenerated from the publication-style final figure (gene-expression-coloured spots on the H&E tissue). Path in zarr-store output redacted to the relative form used elsewhere in the repo.

`>=0.3.4` includes recent fixes used by the tutorials. Adds `.pixi/` (env cache) and `pixi.lock` to `.gitignore`; pixi solves dependencies fresh per contributor, no need to commit the lockfile.

CI uses the default python3 kernel installed by pip; the local pixi kernel name baked into the notebook by an interactive re-execution does not exist in the CI environment, causing jupyter_client.kernelspec.NoSuchKernel.

Pin nbconvert to CI's python3 kernel via --ExecutePreprocessor.kernel_name=python3, so notebooks committed with a contributor-local kernel name (e.g. pixi's sdata-plot-notebooks kernel from kernel-install) still execute correctly in CI.

timtreis mentioned this pull request May 12, 2026
Add Colour and palettes card to gallery scverse/spatialdata-plot#663
Merged

github-actionsBot pushed a commit that referenced this pull request May 12, 2026
Preview for PR#40ec6890
1d5d13b

timtreis added 4 commits May 12, 2026 14:10

Update Visium tutorial and refresh thumbnail
f1ab577
Notebook reworked end-to-end; thumbnail regenerated from the publication-style final figure (gene-expression-coloured spots on the H&E tissue). Path in zarr-store output redacted to the relative form used elsewhere in the repo.

Bump spatialdata-plot floor to 0.3.4 and ignore pixi local state
7612f4e
`>=0.3.4` includes recent fixes used by the tutorials. Adds `.pixi/` (env cache) and `pixi.lock` to `.gitignore`; pixi solves dependencies fresh per contributor, no need to commit the lockfile.

Reset visium notebook kernel to python3
b4cad04
CI uses the default python3 kernel installed by pip; the local pixi kernel name baked into the notebook by an interactive re-execution does not exist in the CI environment, causing jupyter_client.kernelspec.NoSuchKernel.

Harden execute workflow against local kernel names
ed6cb69
Pin nbconvert to CI's python3 kernel via --ExecutePreprocessor.kernel_name=python3, so notebooks committed with a contributor-local kernel name (e.g. pixi's sdata-plot-notebooks kernel from kernel-install) still execute correctly in CI.

github-actionsBot pushed a commit that referenced this pull request May 12, 2026
Preview for PR#48df0dcd
b1cf9f3

timtreis merged commit 3e14801 into mainMay 12, 2026
4 checks passed

timtreis deleted the docs/issue-463-color-palette branch May 12, 2026 12:29