[PYTHON] Animation with matplotlib

Rotating animation with matplotlib

Introduction

matplotlib can not only create graphs, but also animations. The main animation functions are Artist Animation and Func Animation, but here, Func Animation is used to display a rotating animation as shown below.

pic.twitter.com/96f0d4GMRQ

— sabopy.com (@Sabopy_com) January 31, 2020

Commentary

Module import

import.py

import matplotlib.pyplot as plt
import numpy as np
from matplotlib.patches import Circle
import matplotlib.animation as animation
from IPython.display import HTML

Creating diagrams and displaying circles

makefig.py

fig, ax = plt.subplots(figsize=(6,6))

#plot circle
c1 = Circle((0, 0), 5,color='0.75')
ax.add_patch(c1)
c2 = Circle((0, 0), 4.5,color='0.7')
ax.add_patch(c2)
c3 = Circle((0, 0), 4,color='0.65')
ax.add_patch(c3)
c4 = Circle((0, 0), 3.5,color='0.6')
ax.add_patch(c4)
c5 = Circle((0, 0), 3,color='0.55')
ax.add_patch(c5)
c6 = Circle((0, 0), 2.5,color='0.5')
ax.add_patch(c6)
c7 = Circle((0, 0), 2,color='0.45')
ax.add_patch(c7)
c8 = Circle((0, 0), 1.5,color='0.4')
ax.add_patch(c8)
c9 = Circle((0, 0), 1,color='0.35')
ax.add_patch(c9)

Create a fig with plt.subplots and display 9 circles with Circle.

Parameters for moving the circle

param.py

def f(a,t):
    return np.cos(a*t)

def g(b,t,sig):
    return np.sin(b*t+sig)

t = np.linspace(0,2*np.pi,200)
sig = np.pi/4

x=f(3,t)
y=g(2,t,sig)
x1 = np.hstack([x,np.ones(8)])
y1 = np.hstack([y,np.ones(8)*np.sin(sig)])
x2 = np.hstack([np.ones(1),x,np.ones(7)])
y2 = np.hstack([np.ones(1)*np.sin(sig),y,np.ones(7)*np.sin(sig)])
x3 = np.hstack([np.ones(2),x,np.ones(6)])
y3 = np.hstack([np.ones(2)*np.sin(sig),y,np.ones(6)*np.sin(sig)])
x4 = np.hstack([np.ones(3),x,np.ones(5)])
y4 = np.hstack([np.ones(3)*np.sin(sig),y,np.ones(5)*np.sin(sig)])
x5 = np.hstack([np.ones(4),x,np.ones(4)])
y5 = np.hstack([np.ones(4)*np.sin(sig),y,np.ones(4)*np.sin(sig)])
x6 = np.hstack([np.ones(5),x,np.ones(3)])
y6 = np.hstack([np.ones(5)*np.sin(sig),y,np.ones(3)*np.sin(sig)])
x7 = np.hstack([np.ones(6),x,np.ones(2)])
y7 = np.hstack([np.ones(6)*np.sin(sig),y,np.ones(2)*np.sin(sig)])
x8 = np.hstack([np.ones(7),x,np.ones(1)])
y8 = np.hstack([np.ones(7)*np.sin(sig),y,np.ones(1)*np.sin(sig)])
x9 = np.hstack([np.ones(8),x])
y9 = np.hstack([np.ones(8)*np.sin(sig),y])

The circle moves along the Lissajous curve. Use np.hstack etc. to shift the movement of each circle little by little.

Display range setting

xlimylim.py

xmin, xmax = xlim = x.min()-5, x.max()+5
ymin, ymax = ylim = y.min()-5, y.max()+5
ax.set(xlim=xlim, ylim=ylim, autoscale_on=False)
ax.set_aspect('equal')
ax.axis('off')

Animation settings, saves, markings

xlimylim.py

def init():   
    return c1, c2, c3, c4, c5, c6, c7,c8,c9,

def update(num):
    c1.set_center([x1[num],y1[num]])
    c2.set_center([x2[num],y2[num]])
    c3.set_center([x3[num],y3[num]])
    c4.set_center([x4[num],y4[num]])
    c5.set_center([x5[num],y5[num]])
    c6.set_center([x6[num],y6[num]])
    c7.set_center([x7[num],y7[num]])
    c8.set_center([x8[num],y8[num]])
    c9.set_center([x9[num],y9[num]])
    
    return c1, c2, c3, c4, c5, c6, c7,c8,c9

ani = animation.FuncAnimation(fig, update, init_func=init,interval = 20, frames = 208)
ani.save('Tunnel_animation.mp4', writer="ffmpeg",dpi=100)
HTML(ani.to_html5_video())

init is the first process to be executed when the animation is executed, and since we do nothing here, we just return the circle. Since update is a function that is executed for animation, we move each circle with set_center.

In animation.FuncAnimation (), you can specify the switching interval of the figure with interval. The unit is ms. In frames, you can set the number of times the update function is executed.

You can save the animation at the specified resolution with ani.save ('filename.mp4', writer = "ffmpeg", dpi = 100). If ffmpeg is not included

ffmpeg.py

conda install -c conda-forge ffmpeg

Let's put it in. You can display the animation in jupyterlab or notebook with HTML (ani.to_html5_video ()).

reference

https://discourse.matplotlib.org/t/tunnel-animation/20733

Promotion

I have written articles about various matplotlib animations on Blog. Please see if you are interested.

https://sabopy.com/category/py/matplotlib-animation/