Release v0.20.10

[asinghvi17]

Description

• Update to DelaunayTriangulation v1
• Add a Cairo scatter marker conversion override
• Fix categorical colorbar ticks

TODO:

• Update reference images
• Bump versions

[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(fig)

	using	create	display	create	display
GLMakie	3.92s (3.86, 4.04) 0.06+-	166.56ms (163.30, 172.83) 3.89+-	385.53ms (376.03, 399.48) 9.01+-	7.29ms (7.17, 7.57) 0.13+-	25.45ms (25.29, 26.06) 0.27+-
master	4.25s (4.17, 4.39) 0.08+-	163.61ms (161.15, 170.47) 3.20+-	416.86ms (411.96, 434.55) 7.87+-	7.29ms (7.15, 7.71) 0.20+-	25.64ms (25.33, 26.12) 0.33+-
evaluation	1.09x faster✅, -0.33s (-4.76d, 0.00p, 0.07std)	0.98x invariant, 2.95ms (0.83d, 0.15p, 3.55std)	1.08x faster✅, -31.34ms (-3.70d, 0.00p, 8.44std)	1.00x invariant, 0.0ms (0.01d, 0.98p, 0.16std)	1.01x invariant, -0.19ms (-0.62d, 0.27p, 0.30std)
CairoMakie	3.60s (3.55, 3.64) 0.04+-	170.33ms (168.01, 177.27) 3.20+-	103.50ms (101.00, 108.27) 2.37+-	7.49ms (7.39, 7.60) 0.08+-	887.24μs (881.79, 894.48) 4.49+-
master	3.96s (3.92, 4.02) 0.04+-	171.10ms (167.48, 178.89) 3.89+-	104.40ms (101.72, 110.78) 3.82+-	7.66ms (7.47, 7.99) 0.22+-	897.17μs (891.64, 904.48) 4.77+-
evaluation	1.10x faster✅, -0.36s (-10.02d, 0.00p, 0.04std)	1.00x invariant, -0.77ms (-0.22d, 0.69p, 3.54std)	1.01x invariant, -0.91ms (-0.29d, 0.60p, 3.09std)	1.02x invariant, -0.17ms (-1.04d, 0.09p, 0.15std)	1.01x faster ✓, -9.94μs (-2.14d, 0.00p, 4.63std)
WGLMakie	4.06s (3.94, 4.32) 0.12+-	178.09ms (165.13, 193.72) 10.45+-	9.13s (8.59, 9.64) 0.40+-	9.77ms (9.27, 10.14) 0.34+-	69.54ms (68.16, 71.06) 1.25+-
master	4.59s (4.39, 4.97) 0.19+-	174.33ms (166.03, 193.43) 10.54+-	9.11s (8.87, 9.67) 0.29+-	9.63ms (9.33, 10.09) 0.29+-	69.25ms (67.66, 69.99) 0.83+-
evaluation	1.13x faster✅, -0.53s (-3.24d, 0.00p, 0.16std)	0.98x invariant, 3.76ms (0.36d, 0.52p, 10.49std)	1.00x invariant, 0.01s (0.04d, 0.94p, 0.34std)	0.99x invariant, 0.13ms (0.42d, 0.45p, 0.32std)	1.00x invariant, 0.29ms (0.27d, 0.62p, 1.04std)

[SimonDanisch]

@DanielVandH you know what this is about?

│ Closest candidates are:
│   triangulate(::P; segments, boundary_nodes, weights, IntegerType, EdgeType, TriangleType, EdgesType, TrianglesType, randomise, delete_ghosts, delete_empty_features, try_last_inserted_point, skip_points, num_sample_rule, rng, insertion_order, recompute_representative_points, delete_holes, check_arguments, full_polygon_hierarchy, boundary_enricher, boundary_curves, polygonise_n, coarse_n, enrich) where P got unsupported keyword argument "edges"

It comes from tricontourf.md:

n = 20
angles = range(0, 2pi, length = n+1)[1:end-1]
x = [cos.(angles); 2 .* cos.(angles .+ pi/n)]
y = [sin.(angles); 2 .* sin.(angles .+ pi/n)]
z = (x .- 0.5).^2 + (y .- 0.5).^2 .+ 0.5.*randn.()

inner = [n:-1:1; n] # clockwise inner 
outer = [(n+1):(2n); n+1] # counter-clockwise outer
boundary_nodes = [[outer], [inner]]
points = [x'; y']
tri = triangulate(points; boundary_nodes = boundary_nodes)
f, ax, _ = tricontourf(tri, z;
    axis = (; aspect = 1, title = "Constrained triangulation\nvia DelaunayTriangulation.jl"))
scatter!(x, y, color = z, strokewidth = 1, strokecolor = :black)
f

there are a few more calls to triangulate in the docs...

[DanielVandH]

Oh, sorry - I forgot about the .md files when making the changes... It needs to be segments = instead of edges = now.

Oh this example doesn't even use edges = at all?? I'll look into it.

[DanielVandH]

Don't think I can make suggestions to an unchanged file in the PR, so I just made the changes in #3828 that should hopefully fix it @SimonDanisch

[DanielVandH]

Sorry. Can you also change the get_total_area in triplot.md to get_area?