Matplotlib：尝试删除轴，但 plt.axis('off') 将所有图形变为透明，为什么？

Question

我尝试使用 plt.axis('off') 删除轴，但是当我添加此代码时，它会将所有图形变为透明，代码运行没有问题但轴没有 plt.axis('off')。

这是代码

import matplotlib as mpl

import matplotlib.pyplot as plt

import numpy as np

def plot_net(line, dot, name = "test.png"):

    mpl.rcParams['agg.path.chunksize'] = 10000

    fig = plt.figure(figsize = (5, 5), frameon = False)

    axiss = fig.add_axes([0, 0, 1, 1])

    axiss.set_aspect('equal', adjustable = 'datalim')

    plt.axis('off')

    axiss.scatter(dot[:, 0], dot[:, 1], color = 'red', s = 4)

    axiss.plot(line[:, 0], line[:, 1], 'r.', ls = '-', color = '#0063ba', markersize = 2)

    plt.savefig(name, bbox_inches = 'tight', pad_inches = 0, dpi = 200)

    plt.close()

a = np.random.rand(4, 2)

b = np.random.rand(4, 2)

plot_net(a, b, name = "test2.png")

没有 plt.axis('off') 的结果 enter image description here

plt.axis('off') 的结果 enter image description here

Answer 1

尝试将可见性设置为 false 而不是使用“off”：import matplotlib as mplimport matplotlib.pyplot as pltimport numpy as npdef plot_net(line, dot, name = "test.png"):    mpl.rcParams['agg.path.chunksize'] = 10000    fig = plt.figure(figsize = (5, 5), frameon = False)    axiss = fig.add_axes([0, 0, 1, 1])    axiss.set_aspect('equal', adjustable = 'datalim')    axiss.get_yaxis().set_visible(False)    axiss.get_xaxis().set_visible(False)    axiss.scatter(dot[:, 0], dot[:, 1], color = 'red', s = 4)    axiss.plot(line[:, 0], line[:, 1], 'r.', ls = '-', color = '#0063ba', markersize = 2)    plt.savefig(name, bbox_inches = 'tight', pad_inches = 0, dpi = 200)    plt.close()a = np.random.rand(4, 2)b = np.random.rand(4, 2)plot_net(a, b, name = "c:/temp/test2.png")

Answer 2

我刚刚在https://repl.it/repls/DentalRunnyBlockchain#test.png上试过你的代码它工作正常所以 axis off 做你想实现的

Answer 3

看起来它与您对“轴”的定义有关。如果你绕过它它会起作用：import matplotlib as mplimport matplotlib.pyplot as pltimport numpy as npdef plot_net(line, dot, name = "test.png"):    mpl.rcParams['agg.path.chunksize'] = 10000    fig = plt.figure(figsize = (5, 5), frameon = False)    plt.axes().set_aspect('equal', adjustable = 'datalim')    plt.scatter(dot[:, 0], dot[:, 1], color = 'red', s = 4)    plt.plot(line[:, 0], line[:, 1], 'r.', ls = '-', color = '#0063ba', markersize = 2)    plt.axis('off')    plt.savefig(name, bbox_inches = 'tight', pad_inches = 0, dpi = 200)    plt.close()a = np.random.rand(4, 2)b = np.random.rand(4, 2)plot_net(a, b)