Visualization

Visualization#

How do I visualize tissue, tiles, and annotations?#

zs.pl.tissue(wsi)
zs.pl.tiles(wsi, linewidth=0.5)
zs.pl.annotations(wsi, key="annotations", color="class")

Use the same keys that generated the result:

zs.pl.tissue(wsi, tissue_key="tissues_entropy")
zs.pl.tiles(wsi, tile_key="tiles_20x")

How do I color tiles by a stored column?#

When the value is a column in the tile shape table, pass color without feature_key:

zs.pl.tiles(wsi, color="tissue_id", palette="tab10")

For an embedding or prediction table, provide its feature key:

zs.pl.tiles(wsi, feature_key="uni", color="0", cmap="viridis")

Inspect the table’s var_names or columns if you do not know the available color names.

How do I focus on one tissue or region?#

zs.pl.tissue(wsi, tissue_id=0)
zs.pl.tiles(wsi, tissue_id=0)

Or pass a viewport in level-0 coordinates:

zs.pl.tiles(wsi, zoom=(10_000, 20_000, 5_000, 15_000))

How do I create a publication-quality figure?#

Provide an axis and match target_dpi to the export DPI:

import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(6, 6))
zs.pl.tiles(
    wsi,
    feature_key="uni",
    color="0",
    target_dpi=300,
    ax=ax,
)
fig.savefig("feature-map.png", dpi=300, bbox_inches="tight")

Why is my plot blank or misaligned?#

Check that:

  • the requested shape and feature keys exist;

  • the feature table is associated with the selected tile_key;

  • imported annotations use level-0 pixel coordinates;

  • in_bounds matches whether geometries include slide-bound offsets;

  • the selected tissue_id, zoom window, or value range contains data.

Start by plotting shapes without feature coloring, then add one option at a time.