python - 转换 Pandas DF 分组，然后创建字典列表

Question

我有这个熊猫数据框

group_id        person_nid  email
apac-others     1014890     [e-mail]
amer-others     71560392    [e-mail]
apac-others     715603926   [e-mail]
emea-admins     3315024     [e-mail]
amer-finance    100103077   [e-mail]
amer-others     1001203872  [e-mail]
emea-others     100134585   [e-mail]

我想把它变成类似的东西：

{
"apac-others": [
    {"person_nid":"1014890", "email":"[e-mail]"},
    {"person_nid":"715603926", "email":"[e-mail]"},
],
"amer-others": [
    {"person_nid":"71560392", "email":"[e-mail]"},
    {"person_nid":"1001203872", "email":"[e-mail]"},
],
"emea-admins": [
    {"person_nid":"3315024", "email":"[e-mail]"}
],
"amer-finance": [
    {"person_nid":"100103077", "email":"[e-mail]"}
]

}

我的意思是按第一个字段分组，然后将其他字段包含在一个字典中。

标签： pythonpandasdataframe

Answer 1

这是预期的输出吗？

>>> df.set_index('group_id').T.apply(dict).reset_index().groupby('group_id', sort=False).agg(list).to_dict()[0]
{'apac-others': [{'person_nid': 1014890, 'email': '[e-mail]'}, {'person_nid': 715603926, 'email': '[e-mail]'}],
'amer-others': [{'person_nid': 71560392, 'email': '[e-mail]'}, {'person_nid': 1001203872, 'email': '[e-mail]'}],
'emea-admins': [{'person_nid': 3315024, 'email': '[e-mail]'}],
'amer-finance': [{'person_nid': 100103077, 'email': '[e-mail]'}],
'emea-others': [{'person_nid': 100134585, 'email': '[e-mail]'}]}