在Matplotlib 3D绘图中获取查看/摄像机角度？

Question

当我用鼠标旋转Matplotlib 3D绘图时,如何保存视角/摄像机位置,并在下次运行脚本时使用这些值以编程方式设置视角？

Answer 1

TL; DR

的视角被存储在名称下图的轴对象elev和azim,和视图可与被设置plt.gca().view_init(elev, azim).

详细说明答案和例子

导入库并生成数据以绘制:

import matplotlib as mpl  # noqa
from mpl_toolkits.mplot3d import Axes3D  # noqa
import matplotlib.pyplot as plt
import numpy as np
mpl.style.use('seaborn')

np.random.seed(11)
n = 200
x = y = np.linspace(-10, 10, n)
z = np.random.randn(n)*3 + 2

现在绘制数据:

fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
ax.scatter(x, y, z)
ax.set_xlabel('x')
ax.set_ylabel('y')
ax.set_zlabel('z')

我们展示情节并调整视角,直到看起来不错.关闭它后,将打印高程和方位角变量.

plt.show()

print('ax.azim {}'.format(ax.azim))
print('ax.elev {}'.format(ax.elev))

获得这些值后,您可以使用编程方式设置视角

ax.view_init(elev, azim)

示例图 - 在调整之前和调整到ax.azim = -164.5和ax.elev = 51.25之后.

此外,这是一个完整的脚本,如果您想尝试一下,可以将其复制到:

#!/usr/bin/env pythonw
import matplotlib as mpl  # noqa
from mpl_toolkits.mplot3d import Axes3D  # noqa
import matplotlib.pyplot as plt
import numpy as np
mpl.style.use('seaborn')


# ****************************************************************************
# *                               Create data                                *
# ****************************************************************************
np.random.seed(11)
n = 200
x = y = np.linspace(-10, 10, n)
z = np.random.randn(n)*3 + 2


# ****************************************************************************
# *                                 Plot data                                *
# ****************************************************************************
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
ax.scatter(x, y, z)
ax.set_xlabel('x')
ax.set_ylabel('y')
ax.set_zlabel('z')

# # If we knew what angles we wanted to set, these lines will set it
# elev = 42.0
# azim = 105.5
# ax.view_init(elev, azim)

# Show the figure, adjust it with the mouse
plt.show()

# Once the figure is closed, the azimutal angle and the elevation will
# be printed. They may be used as input for ax.view_init(elev, azim)
print('ax.azim {}'.format(ax.azim))
print('ax.elev {}'.format(ax.elev))