Pure-Go (CGO=0) plotting for go-ndarray: matplotlib-class figures to SVG and PNG, with real TeX in the labels.
🌐 Website · 📚 Documentation
The object model and the defaults are matplotlib's — figure size, fonts, colour
cycle, margins, tick placement and tick labels — so a figure looks the way a
matplotlib user expects, and the parts that can be equal are tested equal to
matplotlib: tick locations and labels on 408 real axes, eight colormaps entry
for entry, and the layout of 30 figures (Axes positions, view limits, tick
labels, and positions after TightLayout). The API is the object-oriented one
that most notebooks use (fig, ax = plt.subplots(); ax.plot(...)). What Go adds: one static binary, js/wasm, no Python, and TeX labels
set by a real TeX math typesetter (go-tex/math).
package main
import (
"math"
"os"
"github.com/go-plot/plot"
)
func main() {
x := make([]float64, 200)
y := make([]float64, 200)
for i := range x {
x[i] = float64(i) / 20
y[i] = math.Sin(x[i])
}
fig, axs := plot.Subplots(1, 1)
ax := axs[0][0]
ax.Plot(x, y, plot.Label(`$\sin t$`))
ax.Plot(x, y, plot.Marker("o"), plot.LineStyle("None"), plot.MarkerSize(2))
ax.SetXLabel(`time $t$ (s)`)
ax.Legend()
fig.SavePNG(os.Stdout)
}Data arguments take []float64, other numeric slices, or a *ndarray.Array of
any dtype. A notebook kernel calls fig.Display(), which returns the figure as
image/svg+xml and image/png.
| figures | NewFigure, Subplots(rows, cols), AddAxes, size in inches and dpi, TightLayout |
| plots | Plot (lines, 17 markers, 4 line styles), Scatter (size and colour per point, colormaps), Bar, Barh, Hist, FillBetween, ErrorBar, Imshow (2-D data or RGB(A)), Contour, Contourf, Axhline, Axvline, Text, Annotate |
| axes | titles, labels, Legend (with "best"), Grid, SetXLim/SetYLim, linear and log scales, TwinX, manual ticks, aspect |
| colour | matplotlib's named colours, tab10 cycle, viridis, plasma, inferno, magma, cividis, gray, coolwarm, RdBu (and _r), Colorbar, LogNorm |
| output | SaveSVG, SavePNG, Display |
Not yet: symlog/date axes, constrained_layout, boxplot/pie/violin, hatching, windows
(go-widgets/window), PDF. The full list, and every place where go-plot differs
from matplotlib, is in docs/design.md.
Rendering: twelve reference figures drawn by both (same data, size, dpi) put
every axes frame and every tick mark on the same pixel, and after
TightLayout within 2 pixels; the curves overlap to 0.970–1.000 (see docs/design.md, with the
negative controls the comparison is required to fail).
Speed, in-process, medians of 9 interleaved fresh-process runs on a loaded 16-core host (docs/perf.md):
| figure | go-plot | matplotlib 3.11.2 | |
|---|---|---|---|
| 1e6-point line → PNG | 0.055 s | 0.065 s | 1.2x |
| 1e6-point line → SVG | 0.019 s | 0.053 s | 2.8x |
| 1000² imshow → PNG | 0.025 s | 0.042 s | 1.7x |
| 1000² imshow → SVG | 0.020 s | 0.042 s | 2.1x |
Python's start and import matplotlib.pyplot (about 0.5 s) are not counted.
go-widgets/painter (compositing), go-gfx/gfx (anti-aliased coverage), go-opentype (DejaVu Sans, glyph outlines, kerning), go-tex/math (TeX), go-ndarray (data).
go test ./... # includes the matplotlib fixtures
uv run --with matplotlib --with numpy python tools/gen_ticks.py # regenerate a fixture
cd tools/compare && go run . -out DIR && uv run --with matplotlib --with numpy python figures.py DIR \
&& uv run --with matplotlib --with numpy python compare.py DIR line=line ...Write rendered images to a directory outside the repository.
BSD 3-Clause. See LICENSE.