Question

我有一个如下数据框：

root@Seli:/etc/bind# cat /etc/resolv.conf
search kmlt.lan
domain kmlt.lan
search dmz.kmlt.lan
domain dmz.kmlt.lan

nameserver 192.168.0.200

现在，我需要将“ ind_code”中的所有类似单词归为一组，并且我的DF应该如下所示：

所有带有“ pay”字样的“ ind_code”（无论大小写，在前，最后还是在中间）都应替换为“ Payment” 同样，所有带有“ rev”的单词都应替换为Rev

Answer 1

如果您只想替换具有相似模式的单词，则可以使用此代码段

df.loc[df.ind_code.str.contains(r'[Pp][Aa][Yy]'),'ind_code']='Payment'
df.loc[df.ind_code.str.contains(r'[Rr][Ee][Vv]'),'ind_code']='Rev'

编辑

df.ind_code.str.extract(r'([PpRr][AaEe][YyVv])').str.lower().map({'pay':'Payment','rev':'Rev'})

出局：

    group_code  ind_code
0   111 Payment
1   111 Payment
2   111 Payment
3   111 Payment
4   111 Payment
5   111 Payment
6   111 Payment
7   222 Rev
8   222 Rev
9   222 Rev
10  222 Rev
11  222 Rev
12  222 Rev

Answer 2

def replace_(row):
    if 'pay' in row.lower():
        return 'Payment'
    if 'rev' in row.lower():
        return 'Rev'
    return row
df.ind_code = df.ind_code.apply(lambda row : replace_(row))
print(df)
output:

   group_code ind_code
0         111  Payment
1         111  Payment
2         111  Payment
3         111  Payment
4         111  Payment
5         111  Payment
6         111  Payment
7         222      Rev
8         222      Rev
9         222      Rev
10        222      Rev
11        222      Rev
12        222      Rev

Answer 3

您可以使用正则表达式来做到这一点！

import pandas as pd
import numpy as np
import re

df = pd.DataFrame({"group_code": ['111', '111', '111', '111', '111', '111', 
                                  '111','222','222','222','222','222','222'], 
                     "ind_code": ['Credit pay', 'PAYMENT', 'loan payment', 'bill payment', 
                                  'pays', 'PayMent', 'Payer','Rev12',
                                  'Rev11','13 rev','Rev13','Rev .!','REV 17']})

conditions = [df['ind_code'].str.contains('(pay)', case=False) , df['ind_code'].str.contains('(rev)', case=False)]
choices = ['pay', 'rev']
df['result'] = np.select(conditions, choices, default='unclear')
df

https://regexr.com/

https://pandas.pydata.org/pandas-docs/stable/generated/pandas.Series.str.contains.html

相似词替换为新词

3 个答案: