python - 删除timeindex中的整数

Question

这是数据帧的一部分。正如您所看到的，时间索引中有一些整数。它不应该是时间戳。所以我想删除它。那么我们如何删除以整数作为时间索引的记录呢？

        rent_time                    rent_price_per_square_meter
0      2016-11-28 09:01:58                     0.400000
1      2016-11-28 09:02:35                     0.400000
2      2016-11-28 09:02:43                     0.400000
3      2016-11-28 09:03:21                     0.400000
4      2016-11-28 09:03:21                     0.400000
5      2016-11-28 09:03:34                     0.400000
6      2016-11-28 09:03:34                     0.400000
7      2017-06-17 02:49:33                     0.933333
8      2017-03-19 01:30:03                     0.490196
9      2017-03-10 06:39:03                    11.111111
10     2017-03-09 14:40:03                    16.666667
11     908797                                 11.000000
12     2017-06-08 03:27:52                    22.000000
13     2017-06-30 03:03:11                    22.000000
14     2017-02-20 11:04:48                    12.000000
15     2017-03-05 13:53:39                     6.842105
16     2017-03-06 14:00:01                     6.842105
17     2017-03-15 02:38:54                    20.000000
18     2017-03-15 02:19:07                    13.043478
19     2017-02-24 15:10:00                    25.000000
20     2017-06-26 02:17:31                    13.043478
21     82368                                  11.111111
22     2017-06-30 07:53:55                     4.109589
23     2017-07-17 02:42:43                    20.000000
24     2017-06-30 07:38:00                     5.254237
25     2017-06-30 07:49:00                     4.920635
26     2017-06-30 05:26:26                     4.189189

Answer 1

您可以使用boolean indexing与 to_datetime参数 errors=coerce 表示没有 datetime 值时返回 NaN，然后添加 notnull返回所有日期时间:

df1 = df[pd.to_datetime(df['rent_time'], errors='coerce').notnull()]
print (df1)
              rent_time  rent_price_per_square_meter
0   2016-11-28 09:01:58                     0.400000
1   2016-11-28 09:02:35                     0.400000
2   2016-11-28 09:02:43                     0.400000
3   2016-11-28 09:03:21                     0.400000
4   2016-11-28 09:03:21                     0.400000
5   2016-11-28 09:03:34                     0.400000
6   2016-11-28 09:03:34                     0.400000
7   2017-06-17 02:49:33                     0.933333
8   2017-03-19 01:30:03                     0.490196
9   2017-03-10 06:39:03                    11.111111
10  2017-03-09 14:40:03                    16.666667
12  2017-06-08 03:27:52                    22.000000
13  2017-06-30 03:03:11                    22.000000
14  2017-02-20 11:04:48                    12.000000
15  2017-03-05 13:53:39                     6.842105
16  2017-03-06 14:00:01                     6.842105
17  2017-03-15 02:38:54                    20.000000
18  2017-03-15 02:19:07                    13.043478
19  2017-02-24 15:10:00                    25.000000
20  2017-06-26 02:17:31                    13.043478
22  2017-06-30 07:53:55                     4.109589
23  2017-07-17 02:42:43                    20.000000
24  2017-06-30 07:38:00                     5.254237
25  2017-06-30 07:49:00                     4.920635
26  2017-06-30 05:26:26                     4.189189

编辑:

如果需要的话，进行下一步数据处理DatetimeIndex:

df['rent_time'] = pd.to_datetime(df['rent_time'], errors='coerce')
df = df.dropna(subset=['rent_time']).set_index('rent_time')
print (df)
                    rent_price_per_square_meter
rent_time                                       
2016-11-28 09:01:58                     0.400000
2016-11-28 09:02:35                     0.400000
2016-11-28 09:02:43                     0.400000
2016-11-28 09:03:21                     0.400000
2016-11-28 09:03:21                     0.400000
2016-11-28 09:03:34                     0.400000
2016-11-28 09:03:34                     0.400000
2017-06-17 02:49:33                     0.933333
2017-03-19 01:30:03                     0.490196
2017-03-10 06:39:03                    11.111111
2017-03-09 14:40:03                    16.666667
2017-06-08 03:27:52                    22.000000
2017-06-30 03:03:11                    22.000000
2017-02-20 11:04:48                    12.000000
2017-03-05 13:53:39                     6.842105
2017-03-06 14:00:01                     6.842105
2017-03-15 02:38:54                    20.000000
2017-03-15 02:19:07                    13.043478
2017-02-24 15:10:00                    25.000000
2017-06-26 02:17:31                    13.043478
2017-06-30 07:53:55                     4.109589
2017-07-17 02:42:43                    20.000000
2017-06-30 07:38:00                     5.254237
2017-06-30 07:49:00                     4.920635
2017-06-30 05:26:26                     4.189189