python 3.x -如何将熊猫时间戳转换为自Epoch以来的纳米

Question

自Epoch以来，将Pandas Timestamp转换为nano的最有效方法是什么?

import pandas as pd
ns = 1470924597871000000   
timestamp = pd.to_datetime(ns, unit="ns")

那么如何

timestamp => 1470924597871000000  ns???

Answer 1

对于我来说，它可以很好地使用参数unit，但令人惊讶的是没有参数:

import pandas as pd
ns = 1470924597871000000  
timestamp1 = pd.to_datetime(ns)
print (timestamp1)
2016-08-11 14:09:57.871000
timestamp = pd.to_datetime(ns, unit='ns')
print (timestamp)
2016-08-11 14:09:57.871000

如果需要从Timestamp转换为epoch:

print (timestamp.value)
1470924597871000000

Answer 2

您可以通过它的值访问它:

import pandas as pd
ns = 1470924597871000000   
timestamp = pd.to_datetime(ns)
timestamp.value
Out: 1470924597871000000

Answer 3

传递unit='ns'指定单元类型:

In [46]:
ns = 1470924597871000000   
timestamp = pd.to_datetime(ns, unit='ns')
timestamp
Out[46]:
Timestamp('2016-08-11 14:09:57.871000')

，但这似乎工作得很好，无论如何没有传递这个，除非你是在别的东西之后?

您可以使用timestamp()来获取时间戳值:

In [50]:
timestamp.timestamp()
Out[50]:
1470920997.871