如何为存储的帧制作 networkx 动画？

Question

我想根据我已经生成并存储的指标为我的图表着色。这些将是我的帧 [每个帧的长度 = 图的节点数]，用于 matplotlib 动画，应该传递给图的 node_color 属性。怎么把它变成动画？这是一个小图的最小非工作示例：

# number of nodes
size = 10.

# generate graph
G=nx.complete_graph(size)

# generating input frames here, since my data is too big
# its important that the frames go as input and is not generated 
# on the fly
frame = np.random.random_integers(0, 5, (size, size)) # random ndarray between 0 and 5, length and number of frames = number of nodes in the graph

# draw the topology of the graph, what changes during animation
# is just the color
pos = nx.spring_layout(G)
nodes = nx.draw_networkx_nodes(G,pos)
edges = nx.draw_networkx_edges(G,pos)
plt.axis('off')

# pass frames to funcanimation via update function
# this is where I get stuck, since I cannot break
# out of the loop, neither can I read every array of
# the ndarray without looping over it explicitly
def update(i):
    for i in range(len(frame)):
        # instead of giving frame as input, if I randomly generate it, then it works
        nc =   frame[i] # np.random.randint(2, size=200)
        nodes.set_array(nc)
    return nodes,

# output animation; its important I save it 
ani = FuncAnimation(fig8, update, interval=50,  blit=True)
ani.save('crap.gif', writer='imagemagick',  savefig_kwargs={'facecolor':'white'}, fps=1)

由于上述问题，动画最终只显示最后一帧或第一帧。

请注意，如果我在更新函数中使用 nx.draw()，我会收到以下错误：RuntimeError: the animation function must return a sequence of Artist objects。

我有一种强烈的感觉，我在这里错过了一些明显的简单方法，而且这个问题是微不足道的。但是由于我不经常使用动画功能，所以我不太了解它。

Answer 1

你快到了：update 只想要时间 t 的状态。如果我正确阅读了您的代码，那么您将遍历 update 中的所有 t。所以只需摆脱该循环并将frames 参数添加到FuncAnimation 以便它知道要检查的时间。

import numpy as np
import matplotlib.pyplot as plt
import networkx as nx
from matplotlib.animation import FuncAnimation

# number of nodes
size = 10

# generate graph
G=nx.complete_graph(size)

# generating input frames here, since my data is too big
# its important that the frames go as input and is not generated
# on the fly
frame = np.random.random_integers(0, 5, (size, size)) # random ndarray between 0 and 5, length and number of frames = number of nodes in the graph

# draw the topology of the graph, what changes during animation
# is just the color
pos = nx.spring_layout(G)
nodes = nx.draw_networkx_nodes(G,pos)
edges = nx.draw_networkx_edges(G,pos)
plt.axis('off')

# pass frames to funcanimation via update function
# this is where I get stuck, since I cannot break
# out of the loop, neither can I read every array of
# the ndarray without looping over it explicitly
def update(i):
    # for i in range(len(frame)):
    # instead of giving frame as input, if I randomly generate it, then it works
    nc = frame[i] # np.random.randint(2, size=200)
    nodes.set_array(nc)
    return nodes,

# output animation; its important I save it
fig = plt.gcf()
ani = FuncAnimation(fig, update, interval=50, frames=range(size), blit=True)
ani.save('crap.gif', writer='imagemagick',  savefig_kwargs={'facecolor':'white'}, fps=1)