The animation engine for agents: blazingly fast math and 3D videos with Manim's Python API, on a Rust GPU core.
Python
import numpy as np
import manimgx as m
def grad_f(x: float, y: float) -> np.ndarray:
return np.array([0.5 * x, 0.5 * y])
class TeacherScene(m.Scene):
def construct(self) -> None:
plane = m.NumberPlane(
x_range=[-3, 3, 1],
y_range=[-3, 3, 1],
x_length=8,
y_length=8,
)
self.add(plane)
title = (
m.Tex(
"Gradient field of ",
"$f(x,y) = \\tfrac{1}{4}(x^2 + y^2)$",
)
.scale(0.85)
.to_edge(m.UP, buff=0.3)
)
title[1].set_color(m.YELLOW)
self.play(m.Write(title))
arrows = m.VGroup()
for x in np.arange(-2.0, 2.01, 0.7):
for y in np.arange(-2.0, 2.01, 0.7):
g = grad_f(x, y)
start = plane.coords_to_point(x, y)
end = plane.coords_to_point(x + 0.6 * g[0], y + 0.6 * g[1])
if np.linalg.norm(end - start) > 0.05:
arrows.add(
m.Arrow(
start,
end,
color=m.YELLOW,
buff=0,
stroke_width=2.5,
max_tip_length_to_length_ratio=0.3,
)
)
self.play(
m.LaggedStart(*[m.GrowArrow(a) for a in arrows], lag_ratio=0.01),
run_time=2.5,
)
self.wait(2.0)