uPlot integration

Embed lightweight, fast uPlot charts directly in the GUI interface.

This example demonstrates real-time data visualization using viser.GuiApi.add_uplot() with multiple animated sine wave plots. uPlot provides efficient rendering for time-series data with smooth 60fps updates.

Features demonstrated:

• Multiple synchronized plots with viser.uplot.Series configuration

• Real-time data streaming and chart updates

• Custom scales and styling options

• Legend display with viser.uplot.Legend

Source: examples/02_gui/09_uplot.py

Code

from __future__ import annotations

import time

import numpy as np
import tyro

import viser
import viser.uplot


def y(x: np.ndarray, offset: float) -> np.ndarray:
    return np.sin(2 * np.pi * x + offset)


def main(num_plots: int = 5, num_timesteps: int = 100, num_lines: int = 3) -> None:
    server = viser.ViserServer()
    time_step = 1.0 / 60.0

    # Data for uPlot: tuple of arrays where first is x-data, rest are y-data.
    x_data = time_step * np.arange(num_timesteps, dtype=np.float64)
    data = (x_data, *[y(x_data, i) for i in range(num_lines)])
    print("Data shapes:", [arr.shape for arr in data])

    uplot_handles: list[viser.GuiUplotHandle] = []
    for _ in range(num_plots):
        uplot_handles.append(
            server.gui.add_uplot(
                data=data,
                series=(
                    viser.uplot.Series(label="time"),
                    *[
                        viser.uplot.Series(
                            label=f"y{i}",
                            stroke=["red", "green", "blue"][i % 3],
                            width=2,
                        )
                        for i in range(num_lines)
                    ],
                ),
                scales={
                    "x": viser.uplot.Scale(
                        time=False,
                        auto=True,
                    ),
                    "y": viser.uplot.Scale(range=(-1.5, 2.5)),
                },
                legend=viser.uplot.Legend(show=True),
                aspect=1.0,
            )
        )

    while True:
        # Update the line plot.
        x_data = x_data + time_step
        for uplot_handle in uplot_handles:
            uplot_handle.data = (x_data, *[y(x_data, i) for i in range(num_lines)])
        time.sleep(time_step)


if __name__ == "__main__":
    tyro.cli(main)