如何通过检查行 x+1-行 x 是否大于某个值来更改值（熊猫）

Question

我有一个看起来像这样的表：Dates,               Minutes1/24/2020 2:58:04 PM, 01/24/2020 3:13:04 PM, 01/27/2020 10:04:09 AM, 31/27/2020 10:19:09 AM, 01/27/2020 10:34:09 AM, 01/27/2020 10:49:10 AM, 11/27/2020 11:04:09 AM, 01/27/2020 11:19:09 AM, 11/27/2020 11:34:09 AM, 11/27/2020 11:49:09 AM, 01/27/2020 12:04:09 PM, 131/27/2020 12:19:09 PM, 01/27/2020 12:34:09 PM, 01/27/2020 12:49:09 PM, 01/27/2020 1:04:09 PM, 111/27/2020 1:19:09 PM, 261/27/2020 1:34:09 PM, 411/27/2020 1:49:09 PM, 01/27/2020 2:04:09 PM, 01/27/2020 2:19:09 PM, 121/27/2020 2:34:09 PM, 0我正在检查当前行与前一行之间的差异是否大于或等于 15，如果是，则将值更改为 15。因此新表如下所示：Dates,               Minutes1/24/2020 2:58:04 PM, 01/24/2020 3:13:04 PM, 01/27/2020 10:04:09 AM, 31/27/2020 10:19:09 AM, 01/27/2020 10:34:09 AM, 01/27/2020 10:49:10 AM, 11/27/2020 11:04:09 AM, 01/27/2020 11:19:09 AM, 11/27/2020 11:34:09 AM, 11/27/2020 11:49:09 AM, 01/27/2020 12:04:09 PM, 131/27/2020 12:19:09 PM, 01/27/2020 12:34:09 PM, 01/27/2020 12:49:09 PM, 01/27/2020 1:04:09 PM, 111/27/2020 1:19:09 PM, 15*1/27/2020 1:34:09 PM, 15*1/27/2020 1:49:09 PM, 01/27/2020 2:04:09 PM, 01/27/2020 2:19:09 PM, 121/27/2020 2:34:09 PM, 0*=值改变

Answer 1

让我们做

df.Minutes = df.Minutes.mask(df.Minutes.diff().ge(15), 15)

Answer 2

.loc使用&的另一种方法shift

df.loc[(df['Minutes'] - df['Minutes'].shift(1)).ge(15),'Minutes'] = 15

print(df)

                    Dates  Minutes

0    1/24/2020 2:58:04 PM        0

1    1/24/2020 3:13:04 PM        0

2   1/27/2020 10:04:09 AM        3

3   1/27/2020 10:19:09 AM        0

4   1/27/2020 10:34:09 AM        0

5   1/27/2020 10:49:10 AM        1

6   1/27/2020 11:04:09 AM        0

7   1/27/2020 11:19:09 AM        1

8   1/27/2020 11:34:09 AM        1

9   1/27/2020 11:49:09 AM        0

10  1/27/2020 12:04:09 PM       13

11  1/27/2020 12:19:09 PM        0

12  1/27/2020 12:34:09 PM        0

13  1/27/2020 12:49:09 PM        0

14   1/27/2020 1:04:09 PM       11

15   1/27/2020 1:19:09 PM       15

16   1/27/2020 1:34:09 PM       15

17   1/27/2020 1:49:09 PM        0

18   1/27/2020 2:04:09 PM        0

19   1/27/2020 2:19:09 PM       12

20   1/27/2020 2:34:09 PM        0