Fix 3D heatmaps in CairoMakie

[ffreyer]

Description

Fixes #2889, though the heatmap tiles aren't exactly aligned and image still doesn't work

Type of change

• Bug fix (non-breaking change which fixes an issue)

Checklist

• Added an entry in NEWS.md (for new features and breaking changes)
• Added or changed relevant sections in the documentation
• Added unit tests for new algorithms, conversion methods, etc.
• Added reference image tests for new plotting functions, recipes, visual options, etc.

[MakieBot]

Compile Times benchmark

Note, that these numbers may fluctuate on the CI servers, so take them with a grain of salt. All benchmark results are based on the mean time and negative percent mean faster than the base branch. Note, that GLMakie + WGLMakie run on an emulated GPU, so the runtime benchmark is much slower. Results are from running:

using_time = @ctime using Backend
# Compile time
create_time = @ctime fig = scatter(1:4; color=1:4, colormap=:turbo, markersize=20, visible=true)
display_time = @ctime Makie.colorbuffer(display(fig))
# Runtime
create_time = @benchmark fig = scatter(1:4; color=1:4, colormap=:turbo, markersize=20, visible=true)
display_time = @benchmark Makie.colorbuffer(display(fig))

	using	create	display	create	display
GLMakie	14.63s (13.75, 15.30) 0.58+-	1.43s (1.30, 1.52) 0.07+-	1.04s (0.98, 1.18) 0.07+-	16.84ms (16.37, 17.37) 0.34+-	124.85ms (121.63, 127.04) 1.82+-
master	14.58s (13.92, 15.32) 0.43+-	1.34s (1.27, 1.45) 0.06+-	986.10ms (947.44, 1036.17) 26.74+-	16.59ms (16.10, 16.88) 0.26+-	123.65ms (121.12, 127.14) 2.16+-
evaluation	+0.32%, 0.05s invariant (0.09d, 0.87p, 0.50std)	+6.59%, 0.09s slower❌ (1.44d, 0.02p, 0.07std)	+5.11%, 0.05s noisy🤷‍♀️ (1.00d, 0.10p, 0.05std)	+1.47%, 0.25ms invariant (0.81d, 0.16p, 0.30std)	+0.96%, 1.2ms invariant (0.60d, 0.28p, 1.99std)
CairoMakie	12.31s (12.15, 12.47) 0.11+-	1.42s (1.38, 1.44) 0.02+-	272.69ms (268.13, 279.02) 4.23+-	12.53ms (12.31, 12.81) 0.19+-	7.16ms (7.06, 7.29) 0.07+-
master	12.31s (12.19, 12.49) 0.09+-	1.41s (1.40, 1.43) 0.01+-	271.63ms (266.35, 275.52) 3.72+-	12.72ms (12.51, 12.92) 0.15+-	7.20ms (7.08, 7.33) 0.10+-
evaluation	-0.03%, -0.0s invariant (-0.04d, 0.95p, 0.10std)	+0.12%, 0.0s invariant (0.10d, 0.85p, 0.02std)	+0.39%, 1.06ms invariant (0.27d, 0.63p, 3.98std)	-1.57%, -0.2ms invariant (-1.15d, 0.05p, 0.17std)	-0.52%, -0.04ms invariant (-0.43d, 0.44p, 0.09std)
WGLMakie	15.91s (15.78, 16.07) 0.10+-	1.67s (1.54, 1.93) 0.12+-	14.36s (14.10, 14.79) 0.27+-	19.38ms (17.87, 23.56) 2.01+-	1.45s (1.35, 1.58) 0.07+-
master	15.93s (15.30, 16.55) 0.51+-	1.68s (1.57, 1.89) 0.11+-	14.50s (13.67, 15.03) 0.56+-	18.05ms (16.45, 19.46) 1.01+-	1.43s (1.35, 1.51) 0.05+-
evaluation	-0.08%, -0.01s invariant (-0.03d, 0.95p, 0.31std)	-0.34%, -0.01s invariant (-0.05d, 0.93p, 0.12std)	-0.98%, -0.14s invariant (-0.32d, 0.56p, 0.41std)	+6.85%, 1.33ms noisy🤷‍♀️ (0.84d, 0.15p, 1.51std)	+1.41%, 0.02s invariant (0.33d, 0.54p, 0.06std)

[SimonDanisch]

I think this is ready, no? :)

[ffreyer]

It fixes the heatmap issue but the same issue still exists for image. I'm not sure what problems I ran into trying to fix it. I think I had issue getting the image to transform like the rect it's drawn on?

[ffreyer]

I tried used a poylgon defined with line_to + image surface transformed by a matrix. That doesn't work because of perspective projection:

[ffreyer]

I added the example from the issue as a refimg test. Maybe we should also add one for image and disable it in CairoMakie for now?

[github-actions]

Missing reference images

Found 1 new images without existing references.
Upload new reference images before merging this PR.