我想删除多余的元组但保留外观的顺序。我看了类似的问题。这个问题Find unique rows in numpy.array看起来非常有希望但不知何故它对我不起作用。
我可以在这个答案中使用pandas(https://stackoverflow.com/a/14089586/566035),但我不喜欢使用pandas,这样py2exe生成的可执行文件就会很小。
import numpy as np
data = [('a','z'), ('a','z'), ('a','z'), ('1','z'), ('e','z'), ('c','z')]
#What I want is:
array([['a', 'z'],
['1', 'z'],
['e', 'z'],
['c', 'z']],
dtype='|S1')
#What I have tried:
# (1) numpy.unique, order not preserved
np.unique(data)
array([['a', 'z'],
['c', 'z'],
['1', 'z'],
['e', 'z']],
dtype='|S1')
# (2) python set, order not preserved
set(data)
set([('1', 'z'), ('a', 'z'), ('c', 'z'), ('e', 'z')])
# (3) answer here : https://stackoverflow.com/a/16973510/566035, order not preserved
a = np.array(data)
b = np.ascontiguousarray(a).view(np.dtype((np.void, a.dtype.itemsize * a.shape[1])))
_, idx = np.unique(b, return_index=True)
a[idx]
array([['1', 'z'],
['a', 'z'],
['c', 'z'],
['e', 'z']],
dtype='|S1')
答案 0 :(得分:2)
这在效率方面不是很好,但是可读代码非常简单,可以用于较小的列表:
sorted(set(data), key=data.index)
答案 1 :(得分:1)
糟糕!我自己找到了答案......
seen = set()
np.array([x for x in data if x not in seen and not seen.add(x)])
# output
array([['a', 'z'],
['1', 'z'],
['e', 'z'],
['c', 'z']],
dtype='|S1')