In [1]:

from datetime import datetime, date
import pandas as pd

In [2]:

pd.__version__

Out[2]:

'1.3.2'

inline date elements¶

In [3]:

from datetime import date
import pandas as pd

df = pd.DataFrame({
    "name":          ["alice",        "bob",            "charlie"],
    "date_of_birth": [date(1999,4,1), date(2001,12,15), date(1985,4,24)] 
})

pandas timestamp now¶

In [4]:

df = pd.DataFrame({
    "name":["alice","bob","charlie", "david"],
    "age":[12,43,22,34]
})
df

Out[4]:

	name	age
0	alice	12
1	bob	43
2	charlie	22
3	david	34

In [5]:

df["time"] = pd.Timestamp(datetime.now())
df

Out[5]:

	name	age	time
0	alice	12	2022-05-05 00:46:43.273572
1	bob	43	2022-05-05 00:46:43.273572
2	charlie	22	2022-05-05 00:46:43.273572
3	david	34	2022-05-05 00:46:43.273572

timestamp to string¶

In [6]:

df = pd.DataFrame({
    "name":["alice","bob","charlie", "david"],
    "age":[12,43,22,34]
})

df["timestamp_col"] = pd.Timestamp(datetime.now())
df

Out[6]:

	name	age	timestamp_col
0	alice	12	2022-05-05 00:46:43.285148
1	bob	43	2022-05-05 00:46:43.285148
2	charlie	22	2022-05-05 00:46:43.285148
3	david	34	2022-05-05 00:46:43.285148

In [7]:

df["formatted_col"] = df["timestamp_col"].map(lambda ts: ts.strftime("%d-%m-%Y"))

df

Out[7]:

	name	age	timestamp_col	formatted_col
0	alice	12	2022-05-05 00:46:43.285148	05-05-2022
1	bob	43	2022-05-05 00:46:43.285148	05-05-2022
2	charlie	22	2022-05-05 00:46:43.285148	05-05-2022
3	david	34	2022-05-05 00:46:43.285148	05-05-2022

group by year¶

In [8]:

import pandas as pd

df = pd.DataFrame({
    'name': ['alice','bob','charlie', 'david'],
    'date_of_birth': ['2001-05-27','1999-02-16','1998-09-25', '1999-01-01']
})

df['date_of_birth'] = pd.to_datetime(df['date_of_birth'])

# step 1: create a 'year' column
df['year_of_birth'] = df['date_of_birth'].map(lambda dt: dt.strftime('%Y'))

# step 2: group by the created columns
grouped_df = df.groupby('year_of_birth').size()

grouped_df.to_frame('count').reset_index()

Out[8]:

	year_of_birth	count
0	1998	1
1	1999	2
2	2001	1

group by start of week¶

In [9]:

from datetime import timedelta

import pandas as pd

df = pd.DataFrame({
    'item':          ['a',          'b',         'c',         'd',          'e',          'f'],
    'purchase_date': ['2001-01-15', '2001-01-18','2001-01-21','2001-01-24', '2001-01-27', '2001-01-30']
})

# convert values to datetime type
df['purchase_date'] = pd.to_datetime(df['purchase_date'])


df['purchase_start_of_week'] = df['purchase_date'].map(lambda dt: dt - timedelta(days=dt.weekday()))

grouped_df = df.groupby('purchase_start_of_week').size()

grouped_df.to_frame('count').reset_index()

Out[9]:

	purchase_start_of_week	count
0	2001-01-15	3
1	2001-01-22	2
2	2001-01-29	1