具有空列表解释的嵌套列表的总和

Question

我正在编写一个gensim教程并遇到了一些我不理解的东西.texts是一个嵌套的字符串列表:

In [37]: texts
Out[37]:
[['human', 'machine', 'interface', 'lab', 'abc', 'computer', 'applications'],
 ['survey', 'user', 'opinion', 'computer', 'system', 'response', 'time'],
 ['eps', 'user', 'interface', 'management', 'system'],
 ['system', 'human', 'system', 'engineering', 'testing', 'eps'],
 ['relation', 'user', 'perceived', 'response', 'time', 'error', 'measurement'],
 ['generation', 'random', 'binary', 'unordered', 'trees'],
 ['intersection', 'graph', 'paths', 'trees'],
 ['graph', 'minors', 'iv', 'widths', 'trees', 'well', 'quasi', 'ordering'],
 ['graph', 'minors', 'survey']]

并sum(texts,[])给出:

Out[38]:
['human',
 'machine',
 'interface',
 'lab',
 'abc',
 'computer',
 'applications',
 'survey',
 'user',
 'opinion',
 'computer',

该列表继续了几行,但我省略了其余的以节省空间.我有两个问题:

1)为什么sum(texts,[])产生结果(即展平嵌套列表)？

2)为什么输出显示奇怪 - 每行一个元素？这个输出有什么特别之处(......或者我怀疑它可能是我的iPython表现得很奇怪).请确认您是否也看到了这一点.

Answer 1

这是因为将列表连在一起会将它们连接起来.

sum([a, b, c, d, ..., z], start)

相当于

start + a + b + c + d + ... + z

所以

sum([['one', 'two'], ['three', 'four']], [])

相当于

[] + ['one', 'two'] + ['three', 'four']

哪个给你

['one', 'two', 'three', 'four']

Note that start, by default, is 0, since by default it works with numbers, so if you were to try

sum([['one', 'two'], ['three', 'four']])

then it would try the equivalent of

0 + ['one', 'two'] + ['three', 'four']

and it would fail because you can't add integers to lists.

The one-per-line thing is just how IPython is deciding to output your long list of strings.