词干数据框和更改列值

问题描述

我正在尝试对数据框列值

进行词干，并希望在新列 data['category'] 中以词干为基础附加值

我收到错误消息。请帮忙解决：

ValueError: 值的长度与索引的长度不匹配

words= ['wedding','property','house','university','education','car']
for word in words:
    print(english_stemmer.stem(word))

from nltk.stem import snowballstemmer 
english_stemmer = snowballstemmer('english') 

    
    
queries = data['purpose']

data['category'] = []
 
def purpose_category():

    for query in queries:
        for word in query.split(' '):
            stemmed_word = english_stemmer.stem(word)
            if stemmed_word == 'wed':
                return 'wedding'
            elif stemmed_word == 'properti':
                return 'property'
            elif stemmed_word == 'car':
                return 'car'
            elif stemmed_word =='hous':
                return 'house'
            elif (stemmed_word == 'univers') | (stemmed_word == 'educ'):
                return 'education'
            else:
                return 'real estate'
 
 
for row in data:
    category_value= data['purpose'].apply(purpose_category)
    data['category'].append(category_value)

解决方法

用正则表达式解决这个问题：

dict_map = {
    'wed': 'wedding','properti': 'property','car': 'car','hous': 'house','univers': 'education','educ': 'education'
}
regexp_pattern = '|'.join(dict_map)
data['category'] = (data['purpose']
   .str.extract(regexp_pattern)
   .map(dict_map).fillna('real estate'))

,

如果有帮助，这就是我解决我的方法。看来我们都在做同一个项目。

import pandas as pd
from nltk.stem import SnowballStemmer
english_stemmer = SnowballStemmer('english')
credit_worthiness = pd.read_csv('some_file_here.csv')

def purpose_category_eng(queries):
    for word in queries.split(' '):
        stemmed_word = english_stemmer.stem(word)
        if stemmed_word == 'hous':
            return 'housing'
        if stemmed_word == 'car':
            return 'car purchase'
        if stemmed_word == 'educ' or stemmed_word == 'univers':
            return 'education'
        if stemmed_word == 'wed':
            return 'wedding'
        if stemmed_word == 'estat':
            return 'real estate'
        if stemmed_word == 'properti':
            return 'property'
    return 'unknown'

credit_worthiness['purpose_category'] = credit_worthiness['purpose'].apply(purpose_category_eng)