Question

对于不好的标题名称，我深表歉意。

我正在尝试自动执行电子邮件任务。我需要发送与特定过程关联的电子邮件。我不知道如何正确遍历数据框。我不想结束每个任务的电子邮件，只是过程。

我需要分组吗？我真的很困惑如何做到这一点。

   Process ID Task Execution Date Execution Time Start Date Start Time  End Date End Time   Status               Emails Process Status
0           A    1       8/7/2019        1:00 PM   8/7/2019    1:00 PM  8/7/2019  1:05 PM  Success                  NaN  Successful
1           A    2       8/7/2019        1:05 PM   8/7/2019    1:05 PM  8/7/2019  1:10 PM  Success                  NaN  Successful
2           A    3       8/7/2019        1:10 PM   8/7/2019    1:10 PM  8/7/2019  1:15 PM  Success  ['user1@gmail.com']  Successful
3           B    1       8/7/2019        2:00 PM   8/7/2019    2:00 PM  8/7/2019  2:05 PM  Success                  NaN     FAILED
4           B    2       8/7/2019        2:05 PM   8/7/2019    2:05 PM  8/7/2019  2:10 PM  Success  ['user2@gmail.com']     FAILED
5           B    3       8/7/2019        2:10 PM   8/7/2019    2:10 PM  8/7/2019  2:15 PM   FAILED                  NaN     FAILED

for process in df['Process ID'].unique():
    print(df['Execution Date'])

        msg = print(Process ID + ‘was’ + DAG Status.  Process ID + ‘was completed on’ + End Date (on task 3) + at End Time (on task 3)).  

        server = smtplib.SMTP('sever.com')
        server.sendmail(
        'my_email@gmail.com',
        df.['Emails'],
        msg)
        server.quit()

所需的输出将是一封发送状态说明的电子邮件。

A was Succesful. A was completed on 8/7/2019 at 1:15pm.

B was FAILED. B was completed on 8/7/2019 at 2:15pm

Answer 1

我如何解释您的问题，实际上是您希望从混乱的数据中获取每个进程ID带有标志all(result_set in [True, False])的级别。然后，您想要获取这些数据并对其进行处理。

要达到该汇总级别，请执行以下操作：

df['process_failed'] = df['Process Status'] == 'FAILED'  # consider case insensitive
result_set = df.groupby('Process ID')['process_failed'].any().rename('any_process_id_failed')

我之所以选择生成一个中间变量，是因为GroupBy访问器不能很好地使用字符串（缺少一些应用调用）。

从此输出中，循环遍历它们中的每一个，以执行制作电子邮件所需的任何操作。熊猫不是电子邮件客户端，除非真正长期使用Zawinski法则，因此这不在您的问题范围内。

如何遍历数据框，仅遍历唯一值？

1 个答案: