从柜台绘图将维持秩序

Question

我试图从我从维基百科复制的文章中绘制前50个单词的单词频率。我看过How to plot the number of times each element is in a list Python: Frequency of occurrences和Using Counter() in Python to build histogram?这似乎是一个很有希望的结果，直到我意识到解决方案不能保持Counter()的顺序。有没有办法在绘图时保留Counter()的下降？

我用来播放数据的代码：

# Standard Library
import collections
from collections import Counter
import itertools 
import re

# Third Party Library
import matplotlib.pyplot as plt
import nltk
import numpy as np

file = '...\\NLP\\Word_Embedding\\Basketball.txt'
text = open(file, 'r').read()
text = re.sub(r'([\"\'.])([\)\[,.;])', r'\1 \2', text)

vocab = text.split()
vocab = [words.lower() for words in vocab]
print('There are a total of {} words in the corpus'.format(len(vocab)))
tokens = list(set(vocab))
print('There are {} unique words in the corpus'.format(len(tokens)))

vocab_labels, vocab_values = zip(*Counter(vocab).items())
vocab_freq = Counter(vocab)

indexes = np.arange(len(vocab_labels[:10]))
width = 1

# plt.bar(indexes, vocab_values[:10], width) # Random 10 items from list
# plt.xticks(indexes + width * 0.5, vocab_labels[:10])
# plt.show()

链接到Basketball.txt文件

Answer 1

您可以根据vocab_values对vocab_freq进行排序，然后使用[::-1]进行反向：

import collections
from collections import Counter
import itertools
import re

# Third Party Library
import matplotlib.pyplot as plt
import nltk
import numpy as np

file = '.\Basketball.txt'
text = open(file, 'r').read()
text = re.sub(r'([\"\'.])([\)\[,.;])', r'\1 \2', text)

vocab = text.split()
vocab = [words.lower() for words in vocab]
print('There are a total of {} words in the corpus'.format(len(vocab)))
tokens = list(set(vocab))
print('There are {} unique words in the corpus'.format(len(tokens)))

vocab_labels, vocab_values = zip(*Counter(vocab).items())
vocab_freq = Counter(vocab)

sorted_values = sorted(vocab_values)[::-1]
sorted_labels = [x for (y,x) in sorted(zip(vocab_values,vocab_labels))][::-1]
indexes = np.arange(len(sorted_labels[:10]))
width = 1

plt.bar(indexes, sorted_values[:10] ) # Random 10 items from list
plt.xticks(indexes + width * 0.5, sorted_labels[:10])
plt.show()

结果：