Have you ever tried to work on datetime features in a dataset? It may look quite complicated to write datetime in its format, we can write date and time in form of strings but how to convert it in DateTime Stamp.
Converting strings to datetime objects in Python has become a common practice for data scientists especially in time series projects. Performing this is often times difficult due to various date formats - different month lengths, timezone variations etc.
To solve this, Python provides a specific data type called "datetime". But in many datasets, the dates might be represented as strings. This recipe demonstrates how to convert date strings to the datetime format.
datetime.strptime is the primary routine for parsing strings into datetimes. datetime.strptime(date_string, format)
Once you have your value in datetime objects, you can then extract specific components of the date such as the month, day, or year, all of which are available as the object's attributes.
References: https://docs.python.org/3/library/datetime.html#datetime.datetime.strptime http://strftime.org/
So this is the recipe on how we can change string to DateTime in Python. In this we will do this by using three different functions.
from datetime import datetime
from dateutil.parser import parse
import pandas as pd
We have imported datetime, parse and pandas. These three modules will be required.
We have first defined an object called date_start in which we have stored an string in format %Y-%m-%d. Then we have tried to print it as a DateTime Stamp by using function datetime.strptime.
date_start = '2020-01-01'
print(datetime.strptime(date_start, '%Y-%m-%d'))
We have created a list of date in the format %m/%d/%y and used parse function on all the values of date_list to convert it in the format of datetime64.
date_list = ['2/7/2027', '6/8/2019', '10/25/2020', '6/29/2018', '2/5/2022']
print([parse(x) for x in date_list])
We have created a dictionary of values and passed in function pd.DataFrame to change it into a DataFrame with columns date and value. Then we have checked the data type in the dataframe (ie object) and to change it to datetime format, we have used pd.to_datetime function.
data = {'date': ['2020-05-01 18:47:05.069722', '2016-01-01 18:47:05.119994',
'2014-02-05 18:47:05.178768', '2018-04-02 18:47:05.230071',
'2018-04-06 18:47:05.230071', '2019-08-02 18:47:05.280592',
'2019-07-01 18:47:05.332662', '2011-03-03 18:47:05.385109',
'2024-04-09 18:47:05.436523', '2015-04-04 18:47:05.486877'],
'value': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1]}
df = pd.DataFrame(data, columns = ['date', 'value'])
print(df.dtypes)
print(pd.to_datetime(df['date']))
print(pd.to_datetime(df['date']).dtypes)
So the final output of all the methods are
2020-01-01 00:00:00 [datetime.datetime(2027, 2, 7, 0, 0), datetime.datetime(2019, 6, 8, 0, 0), datetime.datetime(2020, 10, 25, 0, 0), datetime.datetime(2018, 6, 29, 0, 0), datetime.datetime(2022, 2, 5, 0, 0)] date object value int64 dtype: object 0 2020-05-01 18:47:05.069722 1 2016-01-01 18:47:05.119994 2 2014-02-05 18:47:05.178768 3 2018-04-02 18:47:05.230071 4 2018-04-06 18:47:05.230071 5 2019-08-02 18:47:05.280592 6 2019-07-01 18:47:05.332662 7 2011-03-03 18:47:05.385109 8 2024-04-09 18:47:05.436523 9 2015-04-04 18:47:05.486877 Name: date, dtype: datetime64[ns] datetime64[ns]