Plot gallery

Build retained multi-axis figures, render them through OA’s Vulkan compositor, and reproduce six checked presentation assets from matching C++ and Python programs.

TuPaired source6 outputsUi

Source

Surface	Repository path	SHA-256
C++	sdk/cpp/tutorials/plot/tuPlotGallery.cpp	0694b990fb1b8b76
Python	sdk/py/tutorials/plot/tuPlotGallery.py	8f3d7b1e1512e3e6
Presentation	sdk/asset/docs/plot/*.png	Verified per image

Example

A figure owns layout and axes state. Each axes records semantic artists; rendering lowers the retained figure through the shared compositor and returns an oa::Image. The language switch replaces the complete checked source fragment.

1oa::plot::Figure introFigure() {
2	oa::plot::Figure figure({
3		.title = "OA Plot intro",
4		.rows = 1,
5		.cols = 2,
6		.width = 1280U,
7		.height = 560U,
8		.hSpacing = 36,
9		.padding = 34,
10		.theme = oa::plot::Theme::Dark,
11	});
12	figure.title("One retained figure - C++ and Python parity");
13
14	constexpr oa::Array<oa::F32, 8> steps{
15		0.0F, 1.0F, 2.0F, 3.0F, 4.0F, 5.0F, 6.0F, 7.0F};
16	constexpr oa::Array<oa::F32, 8> train{
17		1.00F, 0.78F, 0.61F, 0.48F, 0.37F, 0.29F, 0.23F, 0.19F};
18	constexpr oa::Array<oa::F32, 8> validation{
19		1.04F, 0.83F, 0.66F, 0.53F, 0.43F, 0.35F, 0.30F, 0.27F};
20	constexpr oa::Array<oa::F32, 5> ideal{
21		0.0F, 0.25F, 0.50F, 0.75F, 1.0F};
22	constexpr oa::Array<oa::F32, 5> confidence{
23		0.10F, 0.30F, 0.50F, 0.70F, 0.90F};
24	constexpr oa::Array<oa::F32, 5> accuracy{
25		0.08F, 0.34F, 0.47F, 0.74F, 0.88F};
26
27	auto& curves = figure.ax(0, 0);
28	curves.title("training curves");
29	curves.xLabel("optimizer step");
30	curves.yLabel("cross entropy");
31	curves.limits(0.0F, 7.0F, 0.0F, 1.1F);
32	curves.plot(steps, train,
33		{.color = oa::Color::accent(), .label = "train", .width = 1.6F});
34	curves.plot(steps, validation,
35		{.color = oa::Color::success(), .label = "validation", .width = 1.6F});
36
37	auto& quality = figure.ax(0, 1);
38	quality.title("Calibration");
39	quality.xLabel("confidence");
40	quality.yLabel("observed accuracy");
41	quality.limits(0.0F, 1.0F, 0.0F, 1.0F);
42	quality.plot(ideal, ideal,
43		{.color = oa::Color{0.565F, 0.565F, 0.565F, 1.0F},
44		 .label = "ideal"});
45	quality.scatter(confidence, accuracy,
46		{.color = oa::Color::cyan(), .label = "model", .radius = 3.5F});
47	return figure;
48}
49

Present

Six checked renders cover dark and light diagnostics, evaluation curves, dense training layouts, and projected scalar-field landscapes.