matplotlib-users Mailing List for matplotlib

Hi all,

I just wanted to let you all know that Hans Petter Langtangen,
well-known author of books on scientific Python and long-time champion
of these tools at the University of Oslo for many years, has arranged
for a donation of $10,000 to support matplotlib's development.

Hans Petter is the Director of the Center for Biomedical Computing at
Simula (http://home.simula.no/~hpl), where a number of projects use
Python as key elements of their research, the Fenics platform being
among the most well-known (http://fenicsproject.org).

We have now confirmed that these funds have been transferred to the
NumFOCUS donations account, where Michael and the rest of the team can
make use of them.

I wanted to publicly thank Hans Petter and Simula Labs for
persistently jumping through the necessary hoops to make this
possible, and to Leah, Travis and Anthony at NumFOCUS for managing the
receiving side of things.

--
Fernando Perez (@fperez_org; http://fperez.org)
fperez.net-at-gmail: mailing lists only (I ignore this when swamped!)
fernando.perez-at-berkeley: contact me here for any direct mail

Benjamin Root wrote:
>The init() function only happens once.  So, each call to 
>ax.contourf() just
>simply adds more contours on top of the previous (you just don't 
>see them
>because you don't have masked regions or transparency).  I would 
>suggest
>doing an ax.cla() in the animate() function before doing 
>ax.contourf().

It's like you said.
Thanks for assistance :)

Cheers

On Wed, May 29, 2013 at 3:03 PM, zetah <ot...@hu...> wrote:

> Please consider this small script:
>
>
The init() function only happens once.  So, each call to ax.contourf() just
simply adds more contours on top of the previous (you just don't see them
because you don't have masked regions or transparency).  I would suggest
doing an ax.cla() in the animate() function before doing ax.contourf().

Cheers!
Ben Root

Please consider this small script:

============================================================
import numpy as np
import matplotlib.pyplot as plt
from matplotlib import animation
from time import time

def init():
    return ax.cla()

def animate(i):
    global t
    r = np.random.random(10)
    c = np.sin(2 * np.pi * r) * np.vstack(np.cos(2 * np.pi * r))
    print(time() - t)
    t = time()
    return ax.contourf(c)


if __name__ == '__main__':
    t = time()
    fig = plt.figure()
    ax = fig.add_subplot(111)
    anim = animation.FuncAnimation(fig, animate, init_func=init,
                                   frames=100,
                                   interval=15,
                                   blit=True)
    anim.save('test.mp4', fps=15, extra_args=['-vcodec', 'libx264'])
============================================================


The output from this script is as follows:

============================================================
0.21799993515
0.375
0.28200006485
0.31200003624
0.26500010490
0.31299996376
0.35899996757
0.32800006866
0.34399986267
0.375
0.40600013732
0.39100003242
0.42200016975
0.46899986267
0.5
0.51600003242
0.5
0.54699993133
0.57800006866
0.57799983024
0.59400010108
0.60899996757
0.64100003242
0.65599989891
0.73399996757
0.71900010108
0.73399996757
0.76600003242
0.79699993133
0.81200003624
1.04699993134
0.86000013351
0.89100003242
0.90600013732
0.92199993133
0.92199993133
1.0
1.10899996758
1.01600003242
1.03099989891
1.06200003624
1.11000013351
1.45299983025
1.21900010109
1.34299993515
1.31299996376
1.35900020599
1.39099979401
1.3900001049
1.43799996376
1.42199993134
1.45300006866
1.48399996758
1.5
1.5
1.54700016975
1.54699993134
1.625
1.625
1.625
1.65700006485
1.68700003624
1.67199993134
1.75
1.73399996758
1.73399996758
1.78200006485
1.81200003624
1.82800006866
1.8599998951
1.84299993515
1.93700003624
1.93799996376
1.98399996758
1.96900010109
2.04699993134
2.03099989891
2.03100013733
2.06299996376
2.06200003624
2.10899996758
2.125
2.17199993134
2.14100003242
2.18700003624
2.21900010109
2.2349998951
2.2650001049
2.28100013733
2.28099989891
2.29600000381
2.32899999619
2.375
2.35899996758
2.42199993134
2.46900010109
2.46799993515
2.48500013351
2.5
2.51600003242
============================================================

So every loops is slower and slower. Now, my original script is more complicated, and when it reaches 100th frame the single loop takes half a minute to finish compared to half a second after the first loop. Can someone suggest what is wrong with the above code, and how to avoid this time aggregation on every loop?

Thanks