通过从多个列中删除匹配的电子邮件域来过滤 Pandas 数据框

Posted 2023-03-12

【中文标题】通过从多个列中删除匹配的电子邮件域来过滤 Pandas 数据框

【英文标题】：Filter Pandas Dataframe by Removing matching email domains from multiple columns

【发布时间】：2020-02-10 00:04:09

【问题描述】：

我有这个数据框：我想过滤掉“Email__c”列中的域与“Internal_Email”或“Alt_Email”列中的域匹配的行。 例如，在第一行中，“[e-mail]”中的子字符串/域“compx”与“[e-mail]”和“[e-mail]”中的子字符串/域匹配，所以我希望该行被过滤掉。 下面数据框中的所有行都应该被过滤掉。

Company        Email__c      Action      Internal_Email            Alt_Email
CompX     [e-mail]      View        [e-mail]        [e-mail]
Doit Inc  [e-mail]     View        [e-mail]          [e-mail]
PIA       [e-mail]       Sell        [e-mail]          [e-mail]
Techy     [e-mail]    Buy       [e-mail]        [e-mail]
Techy     [e-mail]    Buy       [e-mail]        [e-mail]

【参考方案1】：

试试这个

s1 = df.Email__c.str.findall(r'@(\w+).')
s2 = df.Internal_Email.str.findall(r'@(\w+).')
s3 = df.Alt_Email.str.findall(r'@(\w+).')

df[s1.ne(s2) & s1.ne(s3)]

【参考方案2】：

试试：

df.loc[~((df["Email__c"].apply(lambda x: x.split("@")[1]) == df["Internal_Email"].apply(lambda x: x.split("@")[1])) | (df["Email__c"].apply(lambda x: x.split("@")[1]) == df["Alt_Email"].apply(lambda x: x.split("@")[1])))]

【讨论】：

整个字符串不必匹配，只匹配域，或者@后面的子字符串。

【参考方案3】：

我想你想要这样的东西：

df = pd.read_csv('data.csv', sep=';'). # i just save your data in csv and read it then
df

出来：

    Company Email__c    Action  Internal_Email  Alt_Email
0   CompX   [e-mail]  View    [e-mail]  [e-mail]
1   Doit Inc    [e-mail] View    [e-mail]    [e-mail]
2   PIA [e-mail]   Sell    [e-mail]   [e-mail]
3   Techy   [e-mail]    Buy [e-mail]    [e-mail]
4   Techy   [e-mail]    Buy [e-mail]    [e-mail]

加工条件：

df['email_c_domain'] = [x.split('@')[1] for x in df['Email__c']]   # make additional column with only domain
df['filter_out_1'] = [x.split('@')[1] for x in df['Internal_Email']] # make additional column with only domain 
df['filter_out_2'] = [x.split('@')[1] for x in df['Alt_Email']] # make additional column with only domain 
df['match_1'] = (df['email_c_domain'] == (df['filter_out_1'])) # match domains from email_c to Internal_email
df['match_2'] = (df['email_c_domain'] == (df['filter_out_2']))  # match domains from email_c to Alt_email
df['filtered_out'] = df['match_1'] | df['match_2'] # chose if one of match will true

现在 df 看起来像这样：

    Company Email__c    Action  Internal_Email  Alt_Email   email_c_domain  filter_out_1    filter_out_2    match_1 match_2 filtered_out
0   CompX   [e-mail]  View    [e-mail]  [e-mail] compx.com   compx.com   compx.com   True    True    True
1   Doit Inc    [e-mail] View    [e-mail]    [e-mail]   doit.com.au doit.com    doit.com    False   False   False
2   PIA [e-mail]   Sell    [e-mail]   [e-mail]    pia.com pia1.com    pia.com False   True    True
3   Techy   [e-mail]    Buy [e-mail]    [e-mail]    techy.com.br    techy.com.br    techy1.com.br   True    False   True
4   Techy   [e-mail]    Buy [e-mail]    [e-mail] techy.com.br    techy.com.br    techy.com.br    True    True    True

现在让我们过滤它：

df[df['filtered_out'] == False]

输出是：

     Company        Email__c    Action  Internal_Email  Alt_Email   email_c_domain  filter_out_1    filter_out_2    match_1 match_2 filtered_out
1   Doit Inc    [e-mail] View    [e-mail]    [e-mail]   doit.com.au doit.com    doit.com    False   False   False