我有一個多索引數據框，如下所示： numberlocation category created_on Arab Republic of Egypt ACCESS 2018-06-25 00:00:00 4 ACCOUNT 2018-04-24 04:00:00 3 2018-05-31 04:00:00 3 2018-06-28 00:00:00 3 ACTIVE DIRECTORY 2018-04-01 00:00:00 3... ... ... ...United States of America WINDOWS 10 2018-09-25 04:00:00 8 2018-09-25 08:00:00 13 2018-09-26 08:00:00 12 2018-09-27 08:00:00 8 2018-09-27 12:00:00 9我想將其轉換為以日期列為索引的數據框，但不丟失列號中的計數，新數據框應該是這樣的：created_on number category location2018-06-25 00:00:00 4 ACCESS Arab Republic of Egypt2018-04-24 04:00:00 3 ACCOUNT Arab Republic of Egypt2018-05-31 04:00:00 3 ACCOUNT Arab Republic of Egypt2018-06-28 00:00:00 3 ACCOUNT Arab Republic of Egypt2018-04-01 00:00:00 3 ACTIVE DIRECTORY Arab Republic of Egypt ... ... ... ...2018-09-25 04:00:00 8 WINDOWS 10 United States of America2018-09-25 08:00:00 13 WINDOWS 10 United States of America2018-09-26 08:00:00 12 WINDOWS 10 United States of America2018-09-27 08:00:00 8 WINDOWS 10 United States of America2018-09-27 12:00:00 9 WINDOWS 10 United States of America我怎樣才能做到這一點？
1 回答
繁星點點滴滴
TA貢獻1803條經驗 獲得超3個贊
swaplevel在級別上使用0，2然后reset_index在級別1和上使用2：
df1 = df.swaplevel(0, 2).reset_index(level=[1, 2])
或者另一個想法先使用然后在列上reset_index使用：set_indexcreated_at
df1 = df.reset_index().set_index('created_on')
結果：
print(df1)
category location number
created_on
2018-06-25 00:00:00 ACCESS Arab Republic of Egypt 4
2018-04-24 04:00:00 ACCOUNT Arab Republic of Egypt 3
2018-05-31 04:00:00 ACCOUNT Arab Republic of Egypt 3
2018-06-28 00:00:00 ACCOUNT Arab Republic of Egypt 3
2018-04-01 00:00:00 ACTIVE DIRECTORY Arab Republic of Egypt 3
...
2018-09-25 04:00:00 WINDOWS 10 United States of America 8
2018-09-25 08:00:00 WINDOWS 10 United States of America 13
2018-09-26 08:00:00 WINDOWS 10 United States of America 12
2018-09-27 08:00:00 WINDOWS 10 United States of America 8
2018-09-27 12:00:00 WINDOWS 10 United States of America 9
添加回答
舉報
0/150
提交
取消