Coloring labels by a continuous variable fixed

[timtreis]

As reported by @clwgg, there was a bug with the color of continuous annotations which resulted in them having incorrect colors. This is now fixed and the tests have been updated. Furthermore, within this fix I noticed our testing threshold was way too high and let a bunch of bugs pass. So I also lowered that and fixed the bugs this revealed.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 91.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 84.32%. Comparing base (ffd4a1f) to head (a9c7609).

Files with missing lines	Patch %	Lines
src/spatialdata_plot/pl/utils.py	91.66%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #344 +/- ##
==========================================
+ Coverage 84.27% 84.32% +0.05% 
==========================================
Files 8 8 Lines 1558 1557 -1 ==========================================
Hits 1313 1313 + Misses 245 244 -1 

Files with missing lines	Coverage Δ
src/spatialdata_plot/pl/utils.py	77.37% <91.66%> (+0.08%)	⬆️

[timtreis]

The two failing tests can probably be traced back to #324 and #303

[melonora]

hmm is the copy really required here? I thought that map array returns an array with same shape as the original array, but it would not have the same pointer

[timtreis]

Ah, I remember. Don't ask me why, but if I don't copy(), one of the tests fails with:

ValueError: buffer source array is read-only

[clwgg]

I had the same issue with the current main: whenever I tried to color labels by a column in table.obs, I'd get ValueError: buffer source array is read-only, and putting in those .copy() calls fixed it. This happened regardless of whether the SpatialData object was buffered to disk or not.

[melonora]

I see, this is probably a bug in skimage actually. For now we could keep it as it is, but please add a comment with the issue so we remove this when we can.

[melonora]

preapproving for the vmin vmax fix, but please if you can open an issue in skimage regarding the buffer error. I have seen similar errors in pandas at some point, I believe it had to do with some cython code not allowing for read only buffer, but this could be fixed I think and then prevents a copy.

[timtreis]

As reported by @clwgg, there was a bug with the color of continuous annotations which resulted in them having incorrect colors. This is now fixed and the tests have been updated. Furthermore, within this fix I noticed our testing threshold was way too high and let a bunch of bugs pass. So I also lowered that and fixed the bugs this revealed.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 91.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 84.32%. Comparing base (ffd4a1f) to head (a9c7609).

Files with missing lines	Patch %	Lines
src/spatialdata_plot/pl/utils.py	91.66%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #344 +/- ##
==========================================
+ Coverage 84.27% 84.32% +0.05% 
==========================================
Files 8 8 Lines 1558 1557 -1 ==========================================
Hits 1313 1313 + Misses 245 244 -1 

Files with missing lines	Coverage Δ
src/spatialdata_plot/pl/utils.py	77.37% <91.66%> (+0.08%)	⬆️

[timtreis]

The two failing tests can probably be traced back to #324 and #303

[melonora]

hmm is the copy really required here? I thought that map array returns an array with same shape as the original array, but it would not have the same pointer

[timtreis]

Ah, I remember. Don't ask me why, but if I don't copy(), one of the tests fails with:

ValueError: buffer source array is read-only

[clwgg]

I had the same issue with the current main: whenever I tried to color labels by a column in table.obs, I'd get ValueError: buffer source array is read-only, and putting in those .copy() calls fixed it. This happened regardless of whether the SpatialData object was buffered to disk or not.

[melonora]

I see, this is probably a bug in skimage actually. For now we could keep it as it is, but please add a comment with the issue so we remove this when we can.

[melonora]

preapproving for the vmin vmax fix, but please if you can open an issue in skimage regarding the buffer error. I have seen similar errors in pandas at some point, I believe it had to do with some cython code not allowing for read only buffer, but this could be fixed I think and then prevents a copy.

[timtreis]

As reported by @clwgg, there was a bug with the color of continuous annotations which resulted in them having incorrect colors. This is now fixed and the tests have been updated. Furthermore, within this fix I noticed our testing threshold was way too high and let a bunch of bugs pass. So I also lowered that and fixed the bugs this revealed.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 91.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 84.32%. Comparing base (ffd4a1f) to head (a9c7609).

Files with missing lines	Patch %	Lines
src/spatialdata_plot/pl/utils.py	91.66%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #344 +/- ##
==========================================
+ Coverage 84.27% 84.32% +0.05% 
==========================================
Files 8 8 Lines 1558 1557 -1 ==========================================
Hits 1313 1313 + Misses 245 244 -1 

Files with missing lines	Coverage Δ
src/spatialdata_plot/pl/utils.py	77.37% <91.66%> (+0.08%)	⬆️

[timtreis]

The two failing tests can probably be traced back to #324 and #303

[melonora]

hmm is the copy really required here? I thought that map array returns an array with same shape as the original array, but it would not have the same pointer

[timtreis]

Ah, I remember. Don't ask me why, but if I don't copy(), one of the tests fails with:

ValueError: buffer source array is read-only

[clwgg]

I had the same issue with the current main: whenever I tried to color labels by a column in table.obs, I'd get ValueError: buffer source array is read-only, and putting in those .copy() calls fixed it. This happened regardless of whether the SpatialData object was buffered to disk or not.

[melonora]

I see, this is probably a bug in skimage actually. For now we could keep it as it is, but please add a comment with the issue so we remove this when we can.

[melonora]

preapproving for the vmin vmax fix, but please if you can open an issue in skimage regarding the buffer error. I have seen similar errors in pandas at some point, I believe it had to do with some cython code not allowing for read only buffer, but this could be fixed I think and then prevents a copy.

[timtreis]

As reported by @clwgg, there was a bug with the color of continuous annotations which resulted in them having incorrect colors. This is now fixed and the tests have been updated. Furthermore, within this fix I noticed our testing threshold was way too high and let a bunch of bugs pass. So I also lowered that and fixed the bugs this revealed.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 91.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 84.32%. Comparing base (ffd4a1f) to head (a9c7609).

Files with missing lines	Patch %	Lines
src/spatialdata_plot/pl/utils.py	91.66%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #344 +/- ##
==========================================
+ Coverage 84.27% 84.32% +0.05% 
==========================================
Files 8 8 Lines 1558 1557 -1 ==========================================
Hits 1313 1313 + Misses 245 244 -1 

Files with missing lines	Coverage Δ
src/spatialdata_plot/pl/utils.py	77.37% <91.66%> (+0.08%)	⬆️

[timtreis]

The two failing tests can probably be traced back to #324 and #303

[melonora]

hmm is the copy really required here? I thought that map array returns an array with same shape as the original array, but it would not have the same pointer

[timtreis]

Ah, I remember. Don't ask me why, but if I don't copy(), one of the tests fails with:

ValueError: buffer source array is read-only

[clwgg]

I had the same issue with the current main: whenever I tried to color labels by a column in table.obs, I'd get ValueError: buffer source array is read-only, and putting in those .copy() calls fixed it. This happened regardless of whether the SpatialData object was buffered to disk or not.

[melonora]

I see, this is probably a bug in skimage actually. For now we could keep it as it is, but please add a comment with the issue so we remove this when we can.

[melonora]

preapproving for the vmin vmax fix, but please if you can open an issue in skimage regarding the buffer error. I have seen similar errors in pandas at some point, I believe it had to do with some cython code not allowing for read only buffer, but this could be fixed I think and then prevents a copy.

[timtreis]

As reported by @clwgg, there was a bug with the color of continuous annotations which resulted in them having incorrect colors. This is now fixed and the tests have been updated. Furthermore, within this fix I noticed our testing threshold was way too high and let a bunch of bugs pass. So I also lowered that and fixed the bugs this revealed.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 91.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 84.32%. Comparing base (ffd4a1f) to head (a9c7609).

Files with missing lines	Patch %	Lines
src/spatialdata_plot/pl/utils.py	91.66%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #344 +/- ##
==========================================
+ Coverage 84.27% 84.32% +0.05% 
==========================================
Files 8 8 Lines 1558 1557 -1 ==========================================
Hits 1313 1313 + Misses 245 244 -1 

Files with missing lines	Coverage Δ
src/spatialdata_plot/pl/utils.py	77.37% <91.66%> (+0.08%)	⬆️

[timtreis]

The two failing tests can probably be traced back to #324 and #303

[melonora]

hmm is the copy really required here? I thought that map array returns an array with same shape as the original array, but it would not have the same pointer

[timtreis]

Ah, I remember. Don't ask me why, but if I don't copy(), one of the tests fails with:

ValueError: buffer source array is read-only

[clwgg]

I had the same issue with the current main: whenever I tried to color labels by a column in table.obs, I'd get ValueError: buffer source array is read-only, and putting in those .copy() calls fixed it. This happened regardless of whether the SpatialData object was buffered to disk or not.

[melonora]

I see, this is probably a bug in skimage actually. For now we could keep it as it is, but please add a comment with the issue so we remove this when we can.

[melonora]

preapproving for the vmin vmax fix, but please if you can open an issue in skimage regarding the buffer error. I have seen similar errors in pandas at some point, I believe it had to do with some cython code not allowing for read only buffer, but this could be fixed I think and then prevents a copy.

[timtreis]

As reported by @clwgg, there was a bug with the color of continuous annotations which resulted in them having incorrect colors. This is now fixed and the tests have been updated. Furthermore, within this fix I noticed our testing threshold was way too high and let a bunch of bugs pass. So I also lowered that and fixed the bugs this revealed.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 91.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 84.32%. Comparing base (ffd4a1f) to head (a9c7609).

Files with missing lines	Patch %	Lines
src/spatialdata_plot/pl/utils.py	91.66%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #344 +/- ##
==========================================
+ Coverage 84.27% 84.32% +0.05% 
==========================================
Files 8 8 Lines 1558 1557 -1 ==========================================
Hits 1313 1313 + Misses 245 244 -1 

Files with missing lines	Coverage Δ
src/spatialdata_plot/pl/utils.py	77.37% <91.66%> (+0.08%)	⬆️

[timtreis]

The two failing tests can probably be traced back to #324 and #303

[melonora]

hmm is the copy really required here? I thought that map array returns an array with same shape as the original array, but it would not have the same pointer

[timtreis]

Ah, I remember. Don't ask me why, but if I don't copy(), one of the tests fails with:

ValueError: buffer source array is read-only

[clwgg]

I had the same issue with the current main: whenever I tried to color labels by a column in table.obs, I'd get ValueError: buffer source array is read-only, and putting in those .copy() calls fixed it. This happened regardless of whether the SpatialData object was buffered to disk or not.

[melonora]

I see, this is probably a bug in skimage actually. For now we could keep it as it is, but please add a comment with the issue so we remove this when we can.

[melonora]

preapproving for the vmin vmax fix, but please if you can open an issue in skimage regarding the buffer error. I have seen similar errors in pandas at some point, I believe it had to do with some cython code not allowing for read only buffer, but this could be fixed I think and then prevents a copy.

[timtreis]

As reported by @clwgg, there was a bug with the color of continuous annotations which resulted in them having incorrect colors. This is now fixed and the tests have been updated. Furthermore, within this fix I noticed our testing threshold was way too high and let a bunch of bugs pass. So I also lowered that and fixed the bugs this revealed.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 91.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 84.32%. Comparing base (ffd4a1f) to head (a9c7609).

Files with missing lines	Patch %	Lines
src/spatialdata_plot/pl/utils.py	91.66%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #344 +/- ##
==========================================
+ Coverage 84.27% 84.32% +0.05% 
==========================================
Files 8 8 Lines 1558 1557 -1 ==========================================
Hits 1313 1313 + Misses 245 244 -1 

Files with missing lines	Coverage Δ
src/spatialdata_plot/pl/utils.py	77.37% <91.66%> (+0.08%)	⬆️

[timtreis]

The two failing tests can probably be traced back to #324 and #303

[melonora]

hmm is the copy really required here? I thought that map array returns an array with same shape as the original array, but it would not have the same pointer

[timtreis]

Ah, I remember. Don't ask me why, but if I don't copy(), one of the tests fails with:

ValueError: buffer source array is read-only

[clwgg]

I had the same issue with the current main: whenever I tried to color labels by a column in table.obs, I'd get ValueError: buffer source array is read-only, and putting in those .copy() calls fixed it. This happened regardless of whether the SpatialData object was buffered to disk or not.

[melonora]

I see, this is probably a bug in skimage actually. For now we could keep it as it is, but please add a comment with the issue so we remove this when we can.

[melonora]

preapproving for the vmin vmax fix, but please if you can open an issue in skimage regarding the buffer error. I have seen similar errors in pandas at some point, I believe it had to do with some cython code not allowing for read only buffer, but this could be fixed I think and then prevents a copy.

[timtreis]

As reported by @clwgg, there was a bug with the color of continuous annotations which resulted in them having incorrect colors. This is now fixed and the tests have been updated. Furthermore, within this fix I noticed our testing threshold was way too high and let a bunch of bugs pass. So I also lowered that and fixed the bugs this revealed.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 91.66667% with 1 line in your changes missing coverage. Please review.

Project coverage is 84.32%. Comparing base (ffd4a1f) to head (a9c7609).

Files with missing lines	Patch %	Lines
src/spatialdata_plot/pl/utils.py	91.66%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #344 +/- ##
==========================================
+ Coverage 84.27% 84.32% +0.05% 
==========================================
Files 8 8 Lines 1558 1557 -1 ==========================================
Hits 1313 1313 + Misses 245 244 -1 

Files with missing lines	Coverage Δ
src/spatialdata_plot/pl/utils.py	77.37% <91.66%> (+0.08%)	⬆️

[timtreis]

The two failing tests can probably be traced back to #324 and #303

[melonora]

hmm is the copy really required here? I thought that map array returns an array with same shape as the original array, but it would not have the same pointer

[timtreis]

Ah, I remember. Don't ask me why, but if I don't copy(), one of the tests fails with:

ValueError: buffer source array is read-only

[clwgg]

I had the same issue with the current main: whenever I tried to color labels by a column in table.obs, I'd get ValueError: buffer source array is read-only, and putting in those .copy() calls fixed it. This happened regardless of whether the SpatialData object was buffered to disk or not.

[melonora]

I see, this is probably a bug in skimage actually. For now we could keep it as it is, but please add a comment with the issue so we remove this when we can.

[melonora]

preapproving for the vmin vmax fix, but please if you can open an issue in skimage regarding the buffer error. I have seen similar errors in pandas at some point, I believe it had to do with some cython code not allowing for read only buffer, but this could be fixed I think and then prevents a copy.