将 NumPy 数组条目映射到字典值

Question

我目前有一个 NumPy 数组，如下所示：

>>> print(a)
[[ 1.      26.9952  30.6501  10.06952  1.61731]
 [ 2.      26.9952  30.6501  10.06952  1.61731]
 [ 0.      26.9952  30.6501  10.06952  1.61731]
 [ 2.      26.9952  30.6501  10.06952  1.61731]
 [ 3.      26.9952  30.6501  10.06952  1.61731]
 [ 2.      26.9952  30.6501  10.06952  1.61731]
 [ 1.      26.9952  30.6501  10.06952  1.61731]
 [ 2.      26.9952  30.6501  10.06952  1.61731]
 [ 3.      26.9952  30.6501  10.06952  1.61731]
 [ 2.      26.9952  30.6501  10.06952  1.61731]]

我有一本看起来像这样的字典：

map_dict = {0.0: 'a', 1.0: 'b', 2.0: 'c', 3.0: 'd'}

我想要做的就是将所有的值在与NumPy阵列的第一列a到相应的值map_dict。有没有一种有效的方法可以做到这一点？我正在使用的当前方法是通过 for 循环并在我进行时迭代地替换值。

Answer 1

你可以试试np.vectorize：

>>> import numpy as np
>>> a = np.arange(12).reshape(4,3)
>>> a
array([[ 0,  1,  2],
       [ 3,  4,  5],
       [ 6,  7,  8],
       [ 9, 10, 11]])
>>> map_dict = {0.0: 'a', 3.0: 'b', 6.0: 'c', 9.0: 'd'}
>>> a = a.astype(object)
>>> a[:,0] = np.vectorize(map_dict.get)(a[:,0])
>>> a
array([['a', 1, 2],
       ['b', 4, 5],
       ['c', 7, 8],
       ['d', 10, 11]], dtype=object)