hebi-charts 0.9.4

Real-time plotting, 3D scenes, and control UIs for robotics.

HEBI Charts

2D and 3D visualization for low-overhead rendering of high-speed telemetry, with a focus on robotics. Data ingestion is fully decoupled from the UI thread, so you can push updates from a 1 MHz control loop without stuttering the 60 FPS render thread.

Standalone and hardware-agnostic. It takes raw numbers, so it works with any actuator, sensor, or data source.

pip install hebi-charts

Quick start

Runnable examples and the full documentation live in hebi-charts-examples.

import time
import random
import hebi_charts

window = hebi_charts.GridWindow(title="Python", size=(800, 600))
chart = window.add_line_chart(title="Random Walk", xlabel="sample", ylabel="rnd(x)")
line = chart.add_line("walk")

x, y = 0, 0
window.show()
while window.is_showing():
    time.sleep(0.001)          # optional: reduce the rate
    x += 1
    y += random.uniform(-1, 1)
    line.add_point(x, y)       # non-blocking

Ctrl/Cmd + C copies the window to the clipboard, Ctrl/Cmd + S saves a PNG.

macOS

Cocoa must own the main thread, so you must wrap the entry point with run_application for a thread hand-off. On other platforms it calls the application code directly, so the code below is portable on all platforms.

def main():
    ...  # your application logic

if __name__ == "__main__":
    hebi_charts.run_application(main)

Requirements

The native library is bundled in the wheel. There is no compilation step or extra download step.

Platform	Wheel
Windows x86-64	win_amd64
Linux x86-64 (glibc)	manylinux_*_x86_64
macOS Intel	macosx_*_x86_64
macOS Apple Silicon	macosx_*_arm64

The target system needs local rendering with hardware GPU acceleration. Headless and software rendering are not supported yet. On Linux the library links against the system GTK 3, X11, and OpenGL stack.

There is no source distribution, so an unsupported platform fails at install time with pip's "no matching distribution" error rather than at import.

For the bulk array APIs, install NumPy alongside it:

pip install "hebi-charts[numpy]"

License

Free to use, including commercially, but not open source — the library is distributed in binary form only. See the LICENSE file included in the distribution, or https://www.hebirobotics.com/software-license.