williamchenjun / pythonanimation

Quickly animate your plots via matplotlib using this simple class.

Introduction

Note: This is not the final version.

If you're here because you want to animate your plots in Python, you could benefit from using this module. It is a simple class called Animate and it is just as simple as defining the function you want to plot and it is done! It's still just a prototype, but I will be updating this in order to animate more types of plots.

How to install

You can start using this either by manually making a copy of the code, or you can clone this repo in your work directory:

git clone https://github.com/williamchenjun/PythonAnimation.git

How to use

The first thing to do is to import the class from animate.py

from animate import Animate

Simple animation example

Let's assume that we want to plot and animate $$f(x) = \sin(x \cdot \sin(x)).$$

then it is as simple as the running the following code

from animate import Animate
import numpy as np

def f(x):
  return np.sin(x*np.sin(x))

anim = Animate()
anim.set_func(f).set_xrange(0, 10, 200).set_yrange(-1, 1).animate()

which will generate the following

Animation.mp4

The video was saved by passing the save = True argument to animate. I have also increased the fps to 60.

Simple customization

It is just as simple to customize your plots by using the set_plot_attrs method. For example, if we want to change the line color, make it thicker, perhaps remove the right and top axes, we can just run the following script

from animate import Animate
import numpy as np

def f(x):
  return np.sin(x*np.sin(x))

anim = Animate()
anim.ax.spines[['right', 'top']].set_visible(False)
anim.set_func(f).set_xrange(0, 10, 500).set_yrange(-1, 1, padding = 1).set_plot_attrs(0, color = 'orange', lw = 3).animate()

as you can see it is very simple and you can also still access the matplotlib components themselves, such as the Axes and Figure.

Animation.mp4

Multiple plots example

Lastly, it is also very simple to plot multiple plots at the same time!

Note: The plots will automatically be sharing the same x-axis. I will be working on making separate animated subplots.

Let's assume we want to plot $$f(x) = \cos(x), \hspace{1em} g(x) = \sin(x).$$ This can be done simply

from animate import Animate
import numpy as np

def f(x):
  return np.cos(x)

def g(x):
  return np.sin(x)

anim = Animate()
anim.ax.spines[['right', 'top']].set_visible(False)
anim.set_funcs([f, g]).set_xrange(0, 10, 500).set_yrange(-1, 1, padding = 1)
anim.set_plot_attrs(0, color = 'orange', lw = 3)
anim.set_plot_attrs(1, color = 'green', lw = 3, ls = 'dashed')
anim.animate()

You can obviously decide to call each method separately instead of concatenating everything, which might become less readable. As you can see we just use set_funcs instead of set_func, and we can target each function (in order) to set their own attributes:

Animation.mp4

Animating vertical lines and inverses

You can also implement animations where functions "grow" upwards rather than left to right. For example, if we consider the function

$$f(x) = x^2 \implies x = \sqrt{y}$$

we can animate this by specifying the inverse argument in set_func

from animate import Animate
import numpy as np

def f(y):
  return y**2

anim = Animate()
anim.ax.spines[['right', 'top']].set_visible(False)
anim.set_func(f, inverse = True).set_xrange(-1, 5, 500).set_yrange(-2, 2, padding = 1)
anim.set_plot_attrs(0, color = 'violet', lw = 3)
anim.animate()

which generates the following

Animation.mp4

As we can see, we don't need to write the inverse function itself. If we define the inverse parameter, it will automatically invert it. If you write the inverse function, and set inverse to True then it will just output the original non-inverted function.

I have also included an additional module called Lines.py in this repository. It is just a utility class to define vertical and horizontal lines. We can use those in conjuction to the inverse property to animate vertical lines:

from animate import Animate
from Lines import Lines
import numpy as np

def f(x):
  return x

def g(x):
  return x**2

def h(x):
  return np.sin(x)

anim = Animate()
line = Lines()
anim.ax.spines[['right', 'top']].set_visible(False)
anim.set_func(f).set_xrange(0, 5, 500).set_yrange(0, 5)
anim.set_func(line.vline(x = 2), inverse=True, _sleep = 1)
anim.set_func(line.vline(x = 4), inverse=True, _sleep = 2)
anim.set_func(line.hline(y = 2), _sleep = 1)
anim.set_plot_attrs(0, lw = 3, color = 'blueviolet')
anim.set_plot_attrs(1, lw = 3, ls = 'dashed', color = 'mediumturquoise')
anim.set_plot_attrs(2, lw = 3, ls = 'dashed', color = 'limegreen')
anim.set_plot_attrs(3, lw = 3, ls = 'dashed', color = 'lightsalmon')
anim.ax.set_title("Testing vertical and horizontal lines".title(), fontdict = {'weight' : 'bold', 'size' : 15, 'family' : 'Helvetica', 'color' : '#393939'})
anim.animate(save = True, duration = 5)

which outputs the following

Animation.mp4

Note: The code looks very messy, but you could easily define a few functions since they don't really require much changing.

Delay Animation

Note: This is still a work in progress.

If you want your plots to start at a different time, you can do this by specifying the _sleep parameter in the set_func method. This is still not available for the set_funcs method.

It can be achieved easily in the following way:

from animate import Animate
import numpy as np

def f(x):
  return x**3

def g(x):
  return x**2

def h(x):
  return np.sin(x)

anim = Animate()
anim.ax.spines[['right', 'top']].set_visible(False)
anim.set_func(f).set_xrange(-1.5, 1.5, 500).set_yrange(-3, 3, padding = 0.5)
anim.set_func(g, _sleep = 1)
anim.set_func(h, _sleep = 2)
anim.set_plot_attrs(0, color = 'indianred', lw = 3)
anim.set_plot_attrs(1, color = 'dodgerblue', lw = 3, ls = 'dotted')
anim.set_plot_attrs(2, color = 'mediumpurple', lw = 5, ls = 'dashdot')
anim.animate(duration=5, save=True)

which will result in the following animation

Animation.mp4

Note: You can specify the duration of the video output by passing the duration in seconds. The video will end when all animations terminate (once).