Reputation: 23
I am trying to create a smooth Manim curve by interpolating points using SciPy. Can anyone explain to me why is this Manim curve renders jagged?
I have tried to use a ValueTracker to iterate over the indexes of the smooth curve generated by SciPy. However, the result is not smooth.
When I plot it with matplotlib it's smooth:
import numpy as np
from manim import *
from scipy import interpolate
from scipy.interpolate import interp1d
def generate_random_curves(n):
x = np.array([0, 0.2, 0.4, 0.8, 0.7, 0.2, 0.0])
y = np.array([0, 0.1, 0.2, 0.9, 1, 0.25, 0])
curves = []
for i in range(n):
rnd_x = np.random.uniform(-0.1, 0.1, 7)
rnd_y = np.random.uniform(-0.1, 0.1, 7)
rnd_x[0] = 0
rnd_y[0] = 0
rnd_x[-1] = 0
rnd_y[-1] = 0
perterb_x = x + rnd_x
perterb_y = y + rnd_y
curves.append([perterb_x, perterb_y])
return curves
class Intro(Scene):
def construct(self):
ax = Axes(
x_range=[0, 1, 10], y_range=[0, 1, 10], axis_config={"include_tip": False}
)
labels = ax.get_axis_labels(x_label="d", y_label="a")
t = ValueTracker(0)
tx, ty = generate_random_curves(1)[0]
tck, u = interpolate.splprep([tx, ty], s=0)
eval_pts = np.linspace(0, 1, 1000)
smooth_tx,smooth_ty = interpolate.splev(eval_pts, tck)
initial_point = [ax.coords_to_point(t.get_value(), interpolate.splev(t.get_value(), tck)[1])]
dot = Dot(point=initial_point)
path = VMobject()
path.set_points_as_corners([dot.get_center(), dot.get_center()])
def update_path(path):
previous_path = path.copy()
previous_path.add_points_as_corners([dot.get_center()])
path.become(previous_path)
path.add_updater(update_path)
dot.add_updater(lambda x: x.move_to(ax.c2p(smooth_tx[int(t.get_value())],
smooth_ty[int(t.get_value())])))
self.add(ax, labels, dot, path)
self.wait()
self.play(t.animate.set_value(999))
self.wait(10)
Upvotes: 2
Views: 653
Reputation: 1423
The more you increase the runtime, the smoother the curve will look:
That's because when you render at 15 fps, the updater will use 15 points in total for interpolation.
Upvotes: 1