plot

Header-only C++ SVG charts — heatmap, line, scatter — GR-style API + Solarized themes.

OS / Compiler	GCC 13	GCC 14	Clang 18	Clang 20	Apple Clang	MSVC
Ubuntu 22.04
Ubuntu 24.04
macOS 15
Windows

Badge and documentation URLs resolve only once GitHub Pages is enabled for this (private) repository — the CI badge job publishes to the gh-pages branch under badges/ (release) and badges-staging/ (staging).

Quick start

#include <plot/all>
#include <vector>

int main() {
  plot::theme(plot::solarized_dark);          // process default; per-call: plot::use{t}

  std::vector<double> x{1, 2, 3, 4, 5};
  plot::line("throughput.svg", x,
      { {"scalar", {820, 1500, 2600, 3400, 3900}},
        {"simd",   {1900, 4100, 7200, 9800, 11500}} },
      plot::title("throughput"), plot::xlabel{"n"}, plot::ylabel{"MB/s"});
}

Features

• GR-style API — plot::line, plot::scatter, and plot::heatmap, each taking an output stream or filename plus order-independent named options (title, xlabel/ylabel, xlog/ylog, legend, width/height, margin).
• Composable view + grid — the no-os form (plot::line(x, y, …)) returns a deferred, movable view rendered via plot::save("f.svg", view) / plot::render(os, view); plot::grid(file|os, plot::rows{r}, plot::cols{c}, view…) tiles several views into one figure (auto layout if rows/cols omitted).
• Solarized themes — plot::solarized_dark / plot::solarized_light; set the global default with plot::theme(t) or override per call with plot::use{t}.
• Dependency-free C++23 — standard library only; no HDF5, zlib, boost, or third-party plotting/matrix/formatting libraries.
• CMake-friendly — find_package(plot) exposes the plot::plot target.

Themes

Set the process-global default with plot::theme(t), or override a single render with plot::use{t}:

plot::theme(plot::tokyo_night);                 // process-global
plot::line("bw.svg", x, y, plot::use{plot::dracula});   // per-call override

Presets shipping in theme.hpp (added in a parallel lane): solarized_dark, solarized_light, dark_plus, light_plus, monokai, dracula, nord, one_dark, gruvbox_dark, gruvbox_light, tomorrow_night, night_owl, material, tokyo_night.

Build & install

cmake -S . -B build -DCMAKE_BUILD_TYPE=Release
cmake --build build -j
ctest --test-dir build              # plot_gr

plot is header-only — drop include/plot on your include path, or consume it via CMake:

find_package(plot CONFIG REQUIRED)
target_link_libraries(my_app PRIVATE plot::plot)

Releases & the single-header drop

Each tagged release ships prebuilt packages (built by .github/workflows/package.yml):

Platform	Artifact
Debian / Ubuntu	plot-dev_<ver>_all.deb
RedHat / Fedora	plot-devel-<ver>.noarch.rpm
Windows	plot-v<ver>-windows.exe (NSIS)
macOS	plot-v<ver>-darwin.pkg
any	plot-v<ver>-<os>.tar.gz / .zip

The payload is a single amalgamated header plot.hpp (generated from include/plot/ by scripts/amalgamate.py). Once installed — or just dropped in your tree — it is all you need:

#include <plot.hpp>   // no -I plot/, no other files

Regenerate it manually with python3 scripts/amalgamate.py include/plot plot.hpp.

Examples

-DPLOTS_BUILD_EXAMPLES=ON builds the demos under examples/; each writes an .svg:

• examples/line_chart → line_chart.svg
• examples/scatter → scatter.svg
• examples/heatmap → heatmap.svg
• examples/dashboard → dashboard.svg (a 2×2 view/grid figure)

Gallery

Every shipping driver, rendered with the Solarized theme. Each image follows your GitHub color scheme (light or dark) automatically.

Documentation

Full API reference: vargalabs.github.io/plots

License

MIT — see LICENSE.