如何将其格式化为大熊猫中的日期？

Question

26JAN2015:14:42:03

How do I format that data properly in pandas as a date? I have two columns in a raw file which have that format and I need them to be in date so I can subtract their values to measure the time in between.

Also, for a quick sanity check. When I am dealing with Dates (normally from Excel or .csv files), I am using code like this:

df['Start']= pd.to_datetime(df['Start'], coerce = True)

df['Date'] = df['Start'].apply(lambda x:x.date().strftime('%Y-%m-%d'))

df['TimeDelta'] = ((df['Start'] - df['End']).astype('timedelta64[s]'))/86400

First I do a pd.to_datetime to change the object data to a date format and then I use lambda commands to switch the formats to ISO standard typically. I also subtract two dates columns to get the time between and divide by 86400 seconds to turn it into days. Are these the most efficient commands to do this with?

Answer 1

Call to_datetime and pass the format string:

In [114]:

df = pd.DataFrame({'date':['26Jan2015:14:42:03']})
df['date'] = pd.to_datetime(df['date'], format='%d%b%Y:%H:%M:%S')
df.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 1 entries, 0 to 0
Data columns (total 1 columns):
date    1 non-null datetime64[ns]
dtypes: datetime64[ns](1)
memory usage: 16.0 bytes
In [115]:

df
Out[115]:
                 date
0 2015-01-26 14:42:03

Answer 2

One more variant is by using regex

import re
dat = "26JAN2015:14:42:03"
dat = re.match("(\d+)(\D+)(\d+):(\d+):(\d+):(\d+)", dat)

print dat.groups()

>>> ('26', 'JAN', '2015', '14', '42', '03')