2

我正在 Python 上编写一个函数，该函数将文本文件的名称（作为字符串）作为输入。该函数应首先确定每个单词在文件中出现的次数。稍后，我将制作一个条形图，显示文件中最常见的十个单词的频率，每个条形旁边是第二个条形，其高度是 Zipf 定律预测的频率。我已经有一些图表代码，但我需要帮助来查找文本文件中最常见的单词。

def zipf_graph(text_file):
    import string
    file = open(text_file, encoding = 'utf8')
    text = file.read()
    file.close()

    #the following strips and removes punctuation and makes the words lowercase
    punc = string.punctuation + '’”—⎬⎪“⎫'
    new_text = text
    for char in punc:
        new_text = new_text.replace(char,'')
        new_text = new_text.lower()
    text_split = new_text.split()

我被困在这里，我试图在列表中找到最常见的字符串，但我不知道从哪里开始，以下是我尝试过的：

    words = text_split
    most_common = max(words, key = words.count)
    # print(most_common)

我还想添加以下代码，因为它被建议帮助

    # Sorting a list by frequency
    # Assumes you have your elements as (word, frequency) tuples
    # (Useful for the zipf function)
    words = [('the', 1), ('and', 1), ('test',2)]
    sorted(words, key = lambda x: x[1], reverse = True)

    # "Sorting" a dictionary by frequency
    # Assumes you have your elements as word:frequency
    # (Useful for the zipf function)
    words = dict()
    words['the'] = 1
    words['and'] = 1
    words['test'] = 2

    # This returns a list of just the most common words without their frequencies
    most_common_words = sorted(words, key = words.get, reverse = True)
    # print(most_common_words)

    # We can go back to the dictionary to get the frequencies
    for word in most_common_words:
        print(word, words[word])

zipf_graph('fortune.txt') #name of the file I chose to use
4

2 回答 2

5

我建议你使用Counterfrom collections。

from collections import Counter

text_split = ["a", "b", "c", "a", "c", "d", "a", "d", "b"]
word_and_freq = Counter(text_split)
top = word_and_freq.most_common(2)

print(top)

有趣的是，这会返回您想要的格式。

[("a", 3), ("b", 2)]
于 2021-03-02T17:14:25.450 回答
3

您可以使用 nltk 库：

import nltk
words = ['words', 'in', 'the', 'file']
fd = nltk.FreqDist(words)
fd.most_common(10)

将以以下格式给出值：

[('file', 1), ('words', 1), ('in', 1), ('the', 1)]
于 2021-03-02T17:11:27.903 回答