我有一些数据记录文件，其中可悲的是，记录器选择记录数据与时间码的精度在浮动的秒，即使记录每2秒。由于这个原因，在一些文件可以有几行相同的采样时间。
日志数据开始示例：

12/1/2014 12:02:02 PM
1
Time,Notes,CO2 Pump Set Point,CO2 Pump Control Mode,CO2 Pump Alarm Set Point,CO2 Pump Flow,CO2 Pump Pressure,CO2 Pump Control Profile,CO2 Pump RPM,CO2 Pump Run State,CO2 Pump Alarm State,Cosolvent Pump Set Point,Cosolvent Pump Control Mode,Cosolvent Pump Alarm Set Point,Cosolvent Pump Flow,Cosolvent Pump Pressure,Cosolvent Pump Control Profile,Cosolvent Pump RPM,Cosolvent Pump Run State,Cosolvent Pump Alarm State,Heater Controller Zone 1 Set Point,Heater Controller Zone 2 Set Point,Heater Controller Zone 3 Set Point,Heater Controller Zone 4 Set Point,Heater Controller Zone 5 Set Point,Heater Controller Zone 6 Set Point,Heater Controller Zone 1 Alarm Set Point,Heater Controller Zone 2 Alarm Set Point,Heater Controller Zone 3 Alarm Set Point,Heater Controller Zone 4 Alarm Set Point,Heater Controller Zone 5 Alarm Set Point,Heater Controller Zone 6 Alarm Set Point,Heater Controller Zone 1 Temperature,Heater Controller Zone 2 Temperature,Heater Controller Zone 3 Temperature,Heater Controller Zone 4 Temperature,Heater Controller Zone 5 Temperature,Heater Controller Zone 6 Temperature,Heater Controller Zone 1 on/off Status,Heater Controller Zone 2 on/off Status,Heater Controller Zone 3 on/off Status,Heater Controller Zone 4 on/off Status,Heater Controller Zone 5 on/off Status,Heater Controller Zone 6 on/off Status,Heater Controller Zone 1 alarm Status,Heater Controller Zone 2 alarm Status,Heater Controller Zone 3 alarm Status,Heater Controller Zone 4 alarm Status,Heater Controller Zone 5 alarm Status,Heater Controller Zone 6 alarm Status,Pressure Regulator Set Point,Pressure Regulator Alarm Set Point,Pressure Regulator External Heater Set Point,Pressure Regulator External Heater Alarm Set Point,Pressure Regulator Valve Heater Set Point,Pressure Regulator Pressure,Pressure Regulator External Heater Temp.,Pressure Regulator Valve Heater Temp.,Pressure Regulator Run State,Pressure Regulator Needle Position,Pressure Regulator Alarm State
12:02:03:562,,50.000000,0,400,0.000000,0,2,0,0,0,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
12:02:04:125,,,,,,,,,,,,,,,,,,,,0,0,0,0,,,0,0,0,0,,,0,0,0,0,,,1,1,1,1,,,0,0,0,0,,,,,,,,,,,,,
12:02:04:218,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,0,400,0,100,35,0,0,26,0,3649,0
12:02:05:625,,50.000000,0,400,0.000000,0,2,0,0,0,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
12:02:06:187,,,,,,,,,,,,,,,,,,,,0,0,0,0,,,0,0,0,0,,,0,0,0,0,,,1,1,1,1,,,0,0,0,0,,,,,,,,,,,,,
12:02:06:281,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,0,400,0,100,35,0,0,26,0,3649,0
12:02:07:687,,50.000000,0,400,0.000000,0,2,0,0,0,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
12:02:08:250,,,,,,,,,,,,,,,,,,,,0,0,0,0,,,0,0,0,0,,,0,0,0,0,,,1,1,1,1,,,0,0,0,0,,,,,,,,,,,,,
12:02:08:343,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,0,400,0,100,35,0,0,27,0,3649,0

通过以下步骤获得带有标题名称的数据：

def timeConv(x):
    return DT.datetime.strptime(x, '%H:%M:%S:%f')
dat=genfromtxt(name,skip_header=2,delimiter=',',usecols=(2,6,32,33,34,35,50,55,59),names=True)
time=genfromtxt(name,skip_header=3,delimiter=',',usecols=(0), converters={0:timeConv},dtype=('S'))

给我数据数组中的前三行是void类型：

(50.0, 0.0, nan, nan, nan, nan, nan, nan, nan)
(nan, nan, 0.0, 0.0, 0.0, 0.0, nan, nan, nan)
(nan, nan, nan, nan, nan, nan, 0.0, 0.0, 3649.0)

我想遍历数据中的每一行，并将dt与前一行进行比较，然后连接这些行，但如何最有效地连接这些数据行呢？

for i in range(len(time)-1):
    dt=time[i+1]-time[i]
    if dt.seconds==0:
        WHAT TO PUT HERE??