使用numpy测试一个数组中的每个元素是否存在于另一个数组中的最有效方法

Question

I would like to know the most efficient way of determining whether elements from a 1-d numpy array exist in another array.

Specifically, I have two numpy 1-d arrays. The first is an unsorted array of integers. The second is a sorted array of target values.

Sample input:

[45982, 124, 12, 1092, 45982, 1, 985, 299, 10092] # array 1

[1, 12, 299] # array 2

Expected output (ie array 1 elements in array 2):

[False, False, True, False, False, True, False, True, False]

The actual arrays will be much longer: array 1 is likely to contain >5,000,000 elements, array 2 is likely to contain from 500,000 to 1,000,000 elements.

Answer 1

maybe np.in1d :

>>> xs = np.array([45982, 124, 12, 1092, 45982, 1, 985, 299, 10092])
>>> ys = np.array([1, 12, 299])
>>> np.in1d(xs, ys)
array([False, False,  True, False, False,  True, False,  True, False], dtype=bool)