如何在Jupyter笔记本中以透明背景显示matplotlib图形？

Question

虽然可以在 Jupyter 控制台中渲染具有透明背景的内联图形，但我无法在 Jupyter Notebook 中将图形背景设置为透明。

为什么？有什么想法吗？

请参阅下面的最小可重现示例。

---

创建透明 matplotlib 图形的代码

%matplotlib inline
import matplotlib.pyplot as plt
fig = plt.figure(facecolor="#00FF0000")
ax = fig.add_axes([0, 0, 1, 1], facecolor="#FF000000")
plt.plot(range(10), [x**2 for x in range(10)])

它可以在 Jupyter 控制台中运行。

jupyter qtconsole --style monokai

但它在 Jupyter 笔记本中不起作用（在 VS Code 中测试）

Answer 1

虽然不透明，但在 Jupyter 中添加此行可以设置颜色以更好地匹配您在 qtconsole 中看到的内容：

fig.patch.set_facecolor([0.08, 0.08 , 0.08, 1]); # based on https://stackoverflow.com/q/61891800/8508004;

OP 的示例代码（几乎*）将其合并到 JupyterLab 4.1 中，并打开主题“JupyterLab Dark”：

%matplotlib inline
import matplotlib.pyplot as plt
fig = plt.figure(facecolor="#00FF0000")
ax = fig.add_axes([0, 0, 1, 1], facecolor="#FF000000", frame_on=True)
plt.plot(range(10), [x**2 for x in range(10)])
fig.patch.set_facecolor([0.08, 0.08 , 0.08, 1]); # based on https://stackoverflow.com/q/61891800/8508004;
#fig.patch.set_alpha(1.0); # based on https://stackoverflow.com/a/4708018/8508004

可以通过定义书脊、刻度线和标签的颜色来进一步修改，以使其他部分更好地显示，如下所示：

%matplotlib inline
import matplotlib.pyplot as plt
fig = plt.figure(facecolor="#00FF0000")
ax = fig.add_axes([0, 0, 1, 1], facecolor="#FF000000", frame_on=True)
plt.plot(range(10), [x**2 for x in range(10)])
fig.patch.set_facecolor([0.08, 0.08 , 0.08, 1]); # based on https://stackoverflow.com/q/61891800/8508004;
#fig.patch.set_alpha(1.0); # based on https://stackoverflow.com/a/4708018/8508004
ax.spines['bottom'].set_color('#dddddd') # based on https://stackoverflow.com/a/51850717/8508004
ax.spines['top'].set_color('#dddddd')  # based on https://stackoverflow.com/a/51850717/8508004
ax.spines['right'].set_color('#dddddd')  # based on https://stackoverflow.com/a/51850717/8508004
ax.spines['left'].set_color('#dddddd')  # based on https://stackoverflow.com/a/51850717/8508004
ax.tick_params(axis='x', colors='#dddddd')  # based on https://stackoverflow.com/a/51850717/8508004
ax.tick_params(axis='y', colors='#dddddd')  # based on https://stackoverflow.com/a/51850717/8508004
ax.yaxis.label.set_color('#dddddd')  # based on https://stackoverflow.com/a/51850717/8508004
ax.xaxis.label.set_color('#dddddd')  # based on https://stackoverflow.com/a/51850717/8508004

---

*注意，我离开了

frame_on = True

，因为我认为

False

与 OP 在屏幕截图中显示的内容不匹配。您的里程可能会有所不同，所以也许您想像OP一样尝试

frame_on=False

。