Skip to content

Coloring labels by a continuous variable appears to be broken (labels associated with wrong colors) #342

@clwgg

Description

@clwgg

My testing code:

import spatialdata as sd
import spatialdata_plot
import matplotlib.pyplot as plt
import numpy as np

sdata = sd.datasets.blobs()
sdata.pl.render_labels("blobs_labels", color="channel_0_sum").pl.show()

Previously (here I used commit c6575e2) this would look something like this:

However, since e97a7fc, this looks like this:

Just to make sure that the first plot is correct, I extracted the labels where channel_0_sum is high, which shows that the two labels shown in yellow and green in the first plot are indeed the ones with the highest values:

highid = sdata.table.obs[sdata.table[:,'channel_0_sum'].X.toarray() > 1000].instance_id
plt.imshow(np.isin(sdata['blobs_labels'], highid))

Metadata

Metadata

Assignees

Labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions