Styjun

I have a pandas DataFrame df with a few datetime columns. In the Python console a sample of the frame is displayed like this:

>>> df

 action dt_completed dt_created dt_modified 
39 update 2010-11-07 22:04:48.668 2010-06-07 07:23:40.536 2017-07-17 06:12:48.059 
6056 release 2010-03-25 10:44:42.575 2010-03-24 17:21:54.751 2017-07-17 NaN 
5913 publish 2018-12-15 11:12:13.000 2018-12-14 00:30:58.393 2018-12-15 11:12:17.441 
7589 create 2011-09-03 22:55:23.656 2011-08-11 12:48:23.543 2011-09-03 22:55:23.656 

When I explore the DataFrame I'm only interested in seeing the date portion of the datetime value. How can I tell pandas to display a DataFrame on the console only with the date portions of datetime values?

I looking for something like display.precision, but for datetimes not for floats. Or some way to (globally?) overwrite pandas' datetime-to-string conversion function.

Please note: I'm not after changing the values in the datetime columns to dates or after adding new columns with shortened fields, but rather I'm looking to alter the display of the datetime columns in the console only.

Try something like this....

Lets create some date.... (Sorry I am too lazy)

import pandas as pd
from datetime import datetime
data='1':[datetime.now(),datetime.now()],
 '2':[datetime.now(),datetime.now()],
 '3':[datetime.now(),datetime.now()]

df=pd.DataFrame.from_dict(data,orient='index')
df.columns=['date1','date2']

I can see the data ....

df.head()

Something like

 date1 date2
1 2019-03-26 12:28:51.080622 2019-03-26 12:28:51.080627
2 2019-03-26 12:28:51.080628 2019-03-26 12:28:51.080628
3 2019-03-26 12:28:51.080629 2019-03-26 12:28:51.080630

I use a custom style

def mydateformat(date):
 return "--".format(date.year, date.month, date.day)

df2=df.style.format('date1':mydateformat)
df2

Which shows

 date1 date2
1 2019-3-26 2019-03-26 12:41:02.834557
2 2019-3-26 2019-03-26 12:41:02.834558
3 2019-3-26 2019-03-26 12:41:02.834560

To check it has not altered the date1 column

df['diff']=-1
df['diff']=df.date1-df.date2

And to view again ....

df2 

I see - which is the new Column in df with style in df2.

 date1 date2
1 2019-3-26 2019-03-26 12:42:14.417441 -1 days +23:59:59.999995
2 2019-3-26 2019-03-26 12:42:14.417443 -1 days +23:59:59.999999
3 2019-3-26 2019-03-26 12:42:14.417444 -1 days +23:59:59.999999

The slight rounding caused by the timestamp accuracy.... not the roungind on the date1 column.

Hope that helps.

import pandas as pd
from datetime import datetime

data='1':[datetime.now(),datetime.now()],
 '2':[datetime.now(),datetime.now()],
 '3':[datetime.now(),datetime.now()]

df=pd.DataFrame.from_dict(data,orient='index')
df.columns=['date1','date2']
df['date3'] = df['date1'].apply(datetime.date)
df['date4'] = df['date2'].apply(datetime.date)
df