How to create datetime index from string in python?

Question

There are three files with names: file_2018-01-01_01_temp.tif, file_2018-01-01_02_temp.tif and file_2018-01-01_03_temp.tif. I want to list them names as ['2018010101', '2018010102', '2018010103'] in python. The below code create an incorrect list.

import pandas as pd
from glob import glob
from os import path

pattern = '*.tif'
filenames = [path.basename(x) for x in glob(pattern)]
pd.DatetimeIndex([pd.Timestamp(f[5:9]) for f in filenames])

Result: DatetimeIndex(['2018-01-01', '2018-01-01', '2018-01-01']

jezrael · Accepted Answer · 2019-06-05 06:28:55Z

3

I think simpliest is indexing with replace in list comprehension:

a = [f[5:18].replace('_','').replace('-','') for f in filenames]
print (a)
['2018010101', '2018010102', '2018010103']

Similar with Series.str.replace:

a = pd.Index([f[5:18] for f in filenames]).str.replace('\-|_', '')
print (a)
Index(['2018010101', '2018010102', '2018010103'], dtype='object')

Or convert values to DatetimeIndex and then use DatetimeIndex.strftime:

a = pd.to_datetime([f[5:18] for f in filenames], format='%Y-%m-%d_%H').strftime('%Y%m%d%H')
print (a)
Index(['2018010101', '2018010102', '2018010103'], dtype='object')

EDIT:

dtype is in object, but it must be in dtype='datetime64[ns]

If need datetimes, then formating has to be default, not possible change it:

d = pd.to_datetime([f[5:18] for f in filenames], format='%Y-%m-%d_%H')
print (d)
DatetimeIndex(['2018-01-01 01:00:00', '2018-01-01 02:00:00',
               '2018-01-01 03:00:00'],
              dtype='datetime64[ns]', freq=None)

edited Jun 5, 2019 at 6:28

answered Jun 5, 2019 at 6:23

jezrael

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

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

1 Comment

HMadadi Over a year ago

Hi @jezrael, dtype is in object, but it must be in dtype='datetime64[ns].

Collectives™ on Stack Overflow

How to create datetime index from string in python?

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