Question

我有一个很大的制表符指定的csv文件：第一个制表符用于情感词，第二个制表符用于八种基本情感，加上值positive和negative，最后一个制表符是布尔值，如果第二个制表符值适合第一个。

文件摘录：

snarl   anger   1
snarl   anticipation    0
snarl   disgust 1
snarl   fear    0
snarl   joy 0
snarl   negative    1
snarl   positive    0
snarl   sadness 0
snarl   surprise    0
snarl   trust   0
snarling    anger   1
snarling    anticipation    0
snarling    disgust 0
snarling    fear    0
snarling    joy 0
snarling    negative    1
snarling    positive    0
snarling    sadness 0
snarling    surprise    0
snarling    trust   0

到目前为止，

我的代码要这样做：

import csv
from pprint import pprint
from itertools import groupby

l = list(csv.reader(open('NRC-Emotion-Lexicon-Wordlevel-v0.92.txt')))
f = lambda x: x[-1] #manipulate number to see different results
{k:[tuple(x[0:1]) for x in v] for k,v in groupby(sorted(l[1:], key=f), f)}

pprint(l)

我当前的输出看起来不太好：

['asylum\tanger\t0'],
 ['asylum\tanticipation\t0'],
 ['asylum\tdisgust\t0'],
 ['asylum\tfear\t1'],
 ['asylum\tjoy\t0'],
 ['asylum\tnegative\t1'],
 ['asylum\tpositive\t0'],
 ['asylum\tsadness\t0'],
 ['asylum\tsurprise\t0'],
 ['asylum\ttrust\t0'],

我的问题是：如何为每个重复的情感单词创建一个带有唯一键的列表字典（将10个重复的重复次数减少为1），并且仅在其中包含第二个制表符元素字典键的列表，当它们的布尔值为1时？

任何帮助将不胜感激！

编辑：作为答复之一，所需输出的示例如下所示：

{'snarl': ['anger', 'disgust'], #included in list due to having '1', ignoring 'positve' and 'negative'
 'snarling': ['anger'], #etc...
}

编辑2：

文件的第一行和最后一行为空，正如我在每个注释的答案中所述。

Answer 1

这是一种方法。使用defaultdict

例如：

import csv
from collections import defaultdict

d = defaultdict(list)
with open(filename) as infile:
    reader = csv.reader(infile, delimiter="\t")
    for row in reader:
        if row[2] == '1':
            d[row[0]].append(row[1])
print(d)

根据评论进行编辑

from collections import defaultdict

d = defaultdict(list)
with open(filename) as infile:
    for row in infile:
        if row.strip():
            val = row.split()
            if val[2] == '1':
                d[val[0]].append(val[1])
print(d)

Answer 2

您可以在迭代collections.defaultdict对象的同时使用csv.reader和更新列表字典。

您的条件会添加到if语句中，请小心通过int将数字转换为整数。

import csv
from collections import defaultdict
from io import StringIO

x = StringIO("""snarl   anger   1
snarl   anticipation    0
...
snarling    surprise    0
snarling    trust   0""")

d = defaultdict(list)

# replace x with open('file.csv', 'r')
with x as fin:
    reader = filter(None, csv.reader(x, delimiter=' ', skipinitialspace=True))
    # or, reader = filter(None, csv.reader(x, delimiter='\t'))
    for word, emotion, num in reader:
        if int(num):
            d[word].append(emotion)

结果：

print(d)

defaultdict(list,
            {'snarl': ['anger', 'disgust', 'negative'],
             'snarling': ['anger', 'negative']})

Answer 3

我想你几乎接近答案了。但是，当您调用csv.reader时，没有指定定界符（这意味着它默认以逗号作为定界符）

>>> from itertools import groupby
>>> l = map(str.split, open('NRC-Emotion-Lexicon-Wordlevel-v0.92.txt').readlines())
>>> f = lambda x: x[1]
>>> {k:set(e[0] for e in v) for k,v in groupby(sorted(filter(bool, l), key=f), f)}
{'anger': {'snarling', 'snarl'}, 'anticipation': {'snarling', 'snarl'}, 'disgust': {'snarling', 'snarl'}, 'fear': {'snarling', 'snarl'}, 'joy': {'snarling', 'snarl'}, 'negative': {'snarling', 'snarl'}, 'positive': {'snarling', 'snarl'}, 'sadness': {'snarling', 'snarl'}, 'surprise': {'snarling', 'snarl'}, 'trust': {'snarling', 'snarl'}}

Answer 4

这就是我要做的。如果愿意，也可以使用collections.defaultdict（而不是setdefault）：

import csv

with open('NRC-Emotion-Lexicon-Wordlevel-v0.92.txt', newline='') as file:
    l = [row[:-1] for row in csv.reader(file, delimiter='\t')
            if row and row[-1] == '1']  # Not empty and last elem is true.

d = {}
for e_word, basic in l:
    d.setdefault(e_word, []).append(basic)

print('dictionary d:\n', d)

输出：

dictionary d:
 {'snarl': ['anger', 'disgust', 'negative'], 'snarling': ['anger', 'negative']}

如何使用布尔值从制表符分隔的csv文件中创建带有唯一提及的列表的字典

4 个答案: