Python/Datetime/Pandas - Changing datetime format

Question

I have this Pandas dataframe:

             Value
Date
2015-06-01     6.9
2015-07-01     7.5
2015-08-01     7.6
2015-09-01     7.6
2015-10-01     7.9
2015-11-01     7.5

I'm accessing its last index item with this code:

somedate=dict.index[-1]

It is already in a datetime format. When I print somedate I get the following:

2015-11-01 00:00:00
2015-11-01 00:00:00

(not sure why it comes in two lines, but it is working fine anyway)

I'm using this same date to build another dataframe from Pandas:

            Prediction  MAPE  Score
2015-11-01        7.93  1.83      1

Well, all I need is that instead of 2015-11-01 in this last frame, it shows up only as 2015-11.

Can I work this out in the variable somedate or that's something I need to adjust when building the dataframe?

OBS.: I tried to use datetime.strptime, however it seems only to work on string dates, not on timestamps.

jezrael · Accepted Answer · 2016-01-11 07:07:44Z

1

IIUC you need DatetimeIndex.to_period:

print df

#            Prediction  MAPE  Score
#                                   
#2015-11-01        7.93  1.83      1

print df.index

#DatetimeIndex(['2015-11-01'], dtype='datetime64[ns]', name=u'', freq=None)

df.index = df.index.to_period('M')
print df

#         Prediction  MAPE  Score
#                                
#2015-11        7.93  1.83      1

Or you can use DataFrame.to_period:

df = df.to_period('M')
print df

#         Prediction  MAPE  Score
#                                
#2015-11        7.93  1.83      1

EDIT:

I think dtype of index is not datetime but object, so you can convert it:

import pandas as pd
import io

temp=u"""Prediction;MAPE;Score
2015-11-01;7.93;1.83;1"""
df = pd.read_csv(io.StringIO(temp), sep=";", index_col=None)
df.index.name = ''    
print df

#            Prediction  MAPE  Score
#                                   
#2015-11-01        7.93  1.83      1

print df.index
#Index([u'2015-11-01'], dtype='object', name=u'')

#convert index to datetime
df.index = pd.to_datetime(df.index)

df = df.to_period('M')
print df

#         Prediction  MAPE  Score
#                                
#2015-11        7.93  1.83      1

edited Jan 11, 2016 at 7:07

answered Jan 10, 2016 at 15:05

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

aabujamra Over a year ago

I'm getting this: AttributeError: 'Index' object has no attribute 'to_period'. Any idea?

Collectives™ on Stack Overflow

Python/Datetime/Pandas - Changing datetime format

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related