6.19. DataFrame Extract

  • Series.str.split(regex, expand=True) - Split strings around given separator/delimiter

  • Series.str.extract(regex, expand=True) - Extract capture groups in the regex pat as columns in a DataFrame

  • Series.str.extractall() - Extract capture groups in the regex pat as columns in DataFrame

  • Series.str.findall(regex) - Find all occurrences of pattern or regular expression in the Series/Index

  • Series.str.fullmatch(regex) - Determine if each string entirely matches a regular expression

  • Series.dt.strftime(...) - formatted strings specified by date_format, which supports the same string format as the python standard library

  • Series.dt.date - Returns numpy array of python datetime.date objects

  • Series.dt.time - Returns numpy array of datetime.time objects

  • Series.dt.timez - Returns numpy array of datetime.time objects with timezone information

6.19.1. SetUp

>>> import pandas as pd
>>>
>>> df = pd.DataFrame([
...     {'firstname': 'Alice', 'lastname': 'Apricot', 'email': 'alice@example.com', 'lastlogin': pd.Timestamp('2000-01-01'), 'groups': 'users;staff'},
...     {'firstname': 'Bob', 'lastname': 'Blackthorn', 'email': 'bob@example.com', 'lastlogin': pd.Timestamp('2000-01-02'), 'groups': 'users;staff'},
...     {'firstname': 'Carol', 'lastname': 'Corn', 'email': 'carol@example.com', 'lastlogin': pd.Timestamp('2000-01-03'), 'groups': 'users'},
...     {'firstname': 'Dave', 'lastname': 'Durian', 'email': 'dave@example.org', 'lastlogin': pd.Timestamp('2000-01-04'), 'groups': 'users'},
...     {'firstname': 'Eve', 'lastname': 'Elderberry', 'email': 'eve@example.org', 'lastlogin': pd.Timestamp('2000-01-05'), 'groups': 'users;staff;admins'},
...     {'firstname': 'Mallory', 'lastname': 'Melon', 'email': 'mallory@example.net', 'lastlogin': pd.NaT, 'groups': None}
... ])

>>> df
  firstname    lastname                email  lastlogin              groups
0     Alice     Apricot    alice@example.com 2000-01-01         users;staff
1       Bob  Blackthorn      bob@example.com 2000-01-02         users;staff
2     Carol        Corn    carol@example.com 2000-01-03               users
3      Dave      Durian     dave@example.org 2000-01-04               users
4       Eve  Elderberry      eve@example.org 2000-01-05  users;staff;admins
5   Mallory       Melon  mallory@example.net        NaT                 NaN

6.19.2. Split

>>> df['email'].str.split('@')
0      [alice, example.com]
1        [bob, example.com]
2      [carol, example.com]
3       [dave, example.org]
4        [eve, example.org]
5    [mallory, example.net]
Name: email, dtype: object

>>> df['email'].str.split('@', expand=True)
         0            1
0    alice  example.com
1      bob  example.com
2    carol  example.com
3     dave  example.org
4      eve  example.org
5  mallory  example.net

>>> new = pd.DataFrame()
>>> new[['username','domain']] = df['email'].str.split('@', expand=True)
>>>
>>> new
  username       domain
0    alice  example.com
1      bob  example.com
2    carol  example.com
3     dave  example.org
4      eve  example.org
5  mallory  example.net

6.19.3. Extract

>>> df
  firstname    lastname                email  lastlogin              groups
0     Alice     Apricot    alice@example.com 2000-01-01         users;staff
1       Bob  Blackthorn      bob@example.com 2000-01-02         users;staff
2     Carol        Corn    carol@example.com 2000-01-03               users
3      Dave      Durian     dave@example.org 2000-01-04               users
4       Eve  Elderberry      eve@example.org 2000-01-05  users;staff;admins
5   Mallory       Melon  mallory@example.net        NaT                 NaN
>>>
>>> df['email'].str.extract(r'([a-z]+)@example.*')
         0
0    alice
1      bob
2    carol
3     dave
4      eve
5  mallory

6.19.4. Findall

>>> df['groups'].str.findall('[a-z]+')
0            [users, staff]
1            [users, staff]
2                   [users]
3                   [users]
4    [users, staff, admins]
5                       NaN
Name: groups, dtype: object