python pandas read_csv如何解析微秒

Question

我有微秒的csv文件作为时间。

Time,Bid

2014-03-03 23:30:30:224323224323,0.8925

2014-03-03 23:30:30:224390224390,0.892525

2014-03-03 23:30:30:224408224408,0.892525

2014-03-03 23:30:30:364299364299,0.892525

如何使用read_csv（）或其他函数将微秒解析为时间索引 read_json可能吗？

谢谢！

Answer 1

继@ Jeff的评论后，您可以执行以下操作：

In [29]:
import pandas as pd
# specifically set the Time column to object dtype
df = pd.read_csv(r'c:\data\temp1.txt', dtype={'Time':object})
df

Out[29]:

                               Time       Bid
0  2014-03-03 23:30:30:224323224323  0.892500
1  2014-03-03 23:30:30:224390224390  0.892525
2  2014-03-03 23:30:30:224408224408  0.892525
3  2014-03-03 23:30:30:364299364299  0.892525

[4 rows x 2 columns]

In [32]:
# trim the erroneous data
df.Time=df.Time.apply(lambda x: x[:-6])
# now apply to_datetime and pass the format string
df.Time = pd.to_datetime(df.Time, format='%Y-%m-%d %H:%M:%S:%f')
df.dtypes

Out[32]:

Time    datetime64[ns]
Bid            float64
dtype: object