Python 每帧行的字频率

我正试图找出如何在每个数据帧行中获得最频繁的单词，比如说前10个最频繁的单词。我有一些代码可以让我在整个DF中使用最频繁的单词，但现在我需要更细粒度

import pandas as pd
import numpy as np
df1 = pd.read_csv('C:/temp/comments.csv',encoding='latin-1',names=['client','comments'])
df1.head(3)

现在我可以得到整个df1中最常见的单词：

y = pd.Series(' '.join(df1['description']).lower().split()).value_counts()[:10]

---

如何获取每个df行的信息？

根据您想要的是数据帧、一系列词典还是词典列表，您可以使用几种不同的方法来实现这一点

from collections import Counter

# dataframe of word counts per row
res = df['comments'].str.split().apply(pd.value_counts)

# series of dictionaries of word counts, each series entry covering one row
res = df['comments'].str.split().apply(Counter)

# list of dictionaries of word counts, each list item covering one row
res = [Counter(x) for x in df['comments'].str.split()]