Добавить новый столбец данных с классификацией

Question

У меня есть следующий фрейм данных:

      df = pd.DataFrame({'Depth' : [13000, 13012, 13015, 13020, 13055, 17000,
                                    18000, 19000, 20000, 33000, 33050, 33300,
                                    33400, 24542, 30000, 30009, 30000, 30050,
                                    30051, 20000, 18000, 9000],

                         'Date': ['2019-10-10 01:14:14', '2019-10-10 03:14:14',                 
                                  '2019-10-10 06:14:14', '2019-10-10 09:14:14',
                                  '2019-10-10 14:14:14', '2019-10-10 14:14:14',                 
                                  '2019-10-10 14:14:14', '2019-10-10 14:14:14',
                                  '2019-10-10 14:14:14', '2019-10-10 14:14:14',                 
                                  '2019-10-10 14:15:14', '2019-10-10 14:16:14',
                                  '2019-10-10 14:17:14', '2019-10-10 14:18:14',                 
                                  '2019-10-10 14:19:14', '2019-10-10 14:20:14',
                                  '2019-10-10 14:21:14', '2019-10-10 14:22:14',                 
                                  '2019-10-10 14:23:14', '2019-10-10 14:24:14',
                                  '2019-10-10 14:25:14', '2019-10-10 14:26:14']
              })

Мне нужно было добавить новый столбец в фрейм данных.

Требуемый вывод:

         Depth        Date                Classification
         13000  2019-10-10 01:14:14       nan #not accepted because the readings are too far in time    
         13012  2019-10-10 03:14:14       nan #not accepted because the readings are too far in time 
         13015  2019-10-10 06:14:14       nan #not accepted because the readings are too far in time 
         13020  2019-10-10 09:14:14       nan #not accepted because the readings are too far in time 
         13055  2019-10-10 14:14:14       nan #not accepted because the readings are too far in time 
         17000  2019-10-10 14:14:14       nan #not accepted because the depth values ​​are far
         18000  2019-10-10 14:14:14       nan #not accepted because the depth values ​​are far
         19000  2019-10-10 14:14:14       nan #not accepted because the depth values ​​are far
         20000  2019-10-10 14:14:14       nan #not accepted because the depth values ​​are far
         33000  2019-10-10 14:14:14        Accept
         33050  2019-10-10 14:15:14        Accept
         33300  2019-10-10 14:16:14        Accept
         33400  2019-10-10 14:17:14        Accept
         24542  2019-10-10 14:18:14          nan
         30000  2019-10-10 14:19:14        Accept
         30009  2019-10-10 14:20:14        Accept
         30000  2019-10-10 14:21:14        Accept
         30050  2019-10-10 14:22:14        Accept
         30051  2019-10-10 14:23:14        Accept
         20000  2019-10-10 14:24:14          nan
         18000  2019-10-10 14:25:14          nan
          9000  2019-10-10 14:26:14          nan

Примечание : показания датчика глубины принимаются, когда разница глубин меньше 1500, а дата показаний близка (разница приблизительно в 10 минут).

Я попытался реализовать следующий код (этот код неверен):

       df['var_time'] = df['Date'].diff()
       df['var_delta_sensor'] = df['depth'].diff().abs()

       time_delta=datetime.timedelta(minutes=10) #10 minutes   

       df['Classification'] = np.where((( df['var_time'] < time_delta) & 
                                         (df['var_delta_sensor'] <1500)), 'Accept', np.NaN) 

Мой неверный вывод:

      depth           Date         var_time    var_delta_sensor   Classification
      13000   2019-10-10 01:14:14   NaT                NaN         nan
      13012   2019-10-10 03:14:14   02:00:00           12.0        nan
      13015   2019-10-10 06:14:14   03:00:00           3.0         nan
      13020   2019-10-10 09:14:14   03:00:00           5.0         nan
      13055   2019-10-10 14:14:14   05:00:00           35.0        nan
      17000   2019-10-10 14:14:14   00:00:00           3945.0      nan
      18000   2019-10-10 14:14:14   00:00:00           1000.0    Accept
      19000   2019-10-10 14:14:14   00:00:00           1000.0    Accept
      20000   2019-10-10 14:14:14   00:00:00           1000.0    Accept
      33000   2019-10-10 14:14:14   00:00:00           13000.0    nan
      33050   2019-10-10 14:15:14   00:01:00           50.0      Accept
      33300   2019-10-10 14:16:14   00:01:00           250.0    Accept
      33400   2019-10-10 14:17:14   00:01:00           100.0    Accept
      24542   2019-10-10 14:18:14   00:01:00           8858.0   nan
      30000   2019-10-10 14:19:14   00:01:00           5458.0   nan
      30009   2019-10-10 14:20:14   00:01:00           9.0       Accept
      30000   2019-10-10 14:21:14   00:01:00           9.0       Accept
      30050   2019-10-10 14:22:14   00:01:00           50.0      Accept
      30051   2019-10-10 14:23:14   00:01:00           1.0       Accept
      20000   2019-10-10 14:24:14   00:01:00          10051.0   nan
      18000   2019-10-10 14:25:14   00:01:00          2000.0    nan
       9000   2019-10-10 14:26:14   00:01:00          9000.0    nan