The animation engine for agents: blazingly fast math and 3D videos with Manim's Python API, on a Rust GPU core.
Python
import math
import numpy as np
import manimgx as m
def curve_pt(theta: float) -> np.ndarray:
r = 1.6 + 0.5 * math.cos(2 * theta)
return np.array([r * math.cos(theta), r * math.sin(theta), 0.0])
SPEEDS = [1.0, 1.4, 0.7, 1.6]
PHASES = [0.0, math.pi / 2, math.pi, 3 * math.pi / 2]
COLORS = [m.YELLOW, m.GREEN, m.RED, m.BLUE]
class TeacherScene(m.Scene):
def construct(self) -> None:
loop = m.ParametricFunction(
curve_pt,
t_range=[0, 2 * math.pi],
color=m.WHITE,
stroke_width=2,
)
self.play(m.Create(loop))
title = (
m.Tex(
"Four points moving at different speeds along a closed loop",
)
.scale(0.75)
.to_edge(m.UP, buff=0.4)
)
self.play(m.Write(title))
t = m.ValueTracker(0.0)
def dot_pos(i: int) -> np.ndarray:
theta = (t.get_value() * SPEEDS[i] + PHASES[i]) % (2 * math.pi)
return curve_pt(theta)
dots = [
m.always_redraw(
lambda i=i: m.Dot(dot_pos(i), color=COLORS[i], radius=0.1),
)
for i in range(4)
]
polygon = m.always_redraw(
lambda: m.Polygon(
*[dot_pos(i) for i in range(4)],
color=m.PURPLE,
stroke_width=2,
fill_color=m.PURPLE,
fill_opacity=0.2,
)
)
for d in dots:
self.add(d)
self.add(polygon)
self.play(t.animate.set_value(2 * math.pi), run_time=6.0, rate_func=m.linear)
self.wait(1.5)