Python 类 __call__ 函数未矢量化

Question

我正在尝试使用__call__ dunder 为自定义类绘制函数

import math
import numpy as np
import matplotlib.pyplot as plt

class F:
    def __init__(self, n, m):
        self._n = n
        self._m = m

    def __call__(self, x):
        return math.sin(self._n * x) * math.cos(self._m * x)

u = F(1,1)
v = F(1,1)

x_array = np.linspace(0, 2*np.pi, 20)

u_array = u(x_array)
v_array = v(x_array)

plt.plot(u_array, v_array)
plt.show()

但是，执行代码时，我收到此错误

Traceback (most recent call last):
  File "F.py", line 18, in <module>
    u_array = u(x_array)
  File "F.py", line 11, in __call__
    return math.sin(self._n * x) * math.cos(self._m * x)
TypeError: only size-1 arrays can be converted to Python scalars

似乎调用u_array = u(x_array)将整个 x_array 发送到函数，而不是以矢量化的方式进行。

请帮忙

Answer 1

使用numpy向量化函数而不是来自math函数：

def __call__(self, x):
    return np.sin(self._n * x) * np.cos(self._m * x)