您可能正在寻找np.meshgrid(). 通过示例最容易了解它是如何工作的：

>>> import numpy as np

# Start with three different 1-D arrays, possibly with different sizes.
>>> a = np.random.random(10)
>>> b = np.random.random(15)
>>> c = np.random.random(20)

# Use np.meshgrid() to create 3-D arrays by joining the three 1-D arrays.
>>> A, B, C = np.meshgrid(a, b, c, indexing="ij")
>>> print(A.shape, B.shape, C.shape)
(10, 15, 20) (10, 15, 20) (10, 15, 20)

# It's easy to use these arrays for computations like x_ijk = a_i + b_j * c_k.
>>> X = A + B * C

# Let's do this with loops to check that we get the same result.
>>> Y = np.empty((10, 15, 20))
>>> for i in range(a.size):                    # i = 0, ..., 9
...     for j in range(b.size):                # j = 0, ..., 14
...         for k in range(c.size):            # k = 0, ..., 19
...             Y[i,j,k] = a[i] + b[j] * c[k]  # y_ijk = a_i + b_j * c_k
... 
>>> np.allclose(X,Y)
True

这使您可以在所需的网格上以一种高效、易读的方式计算任何标量值函数只要记住在适用的情况下使用 NumPy 函数（例如，而不是）。$f: (a,b,c) \mapsto \mathbb{R}$np.sin()math.sin()

例如，如果，首先定义网格数组, , 和, 用于获取, , 和, 和 设置。$f(a,b,c) = \sin(a)b^2 + e^{c}$abcnp.meshgrid()ABCF = np.sin(A) * B**2 + np.exp(C)