Pandas Series: rolling() function

Last update on September 15 2022 12:54:19 (UTC/GMT +8 hours)

Rolling window calculations in Pandas

The rolling() function is used to provide rolling window calculations.

Syntax:

Series.rolling(self, window, min_periods=None, center=False, win_type=None, on=None, axis=0, closed=None)

Parameters:

Name	Description	Type/Default Value	Required / Optional
window	Size of the moving window. This is the number of observations used for calculating the statistic. Each window will be a fixed size. If its an offset then this will be the time period of each window. Each window will be a variable sized based on the observations included in the time-period. This is only valid for datetimelike indexes.	int, or offset	Required
min_periods	Minimum number of observations in window required to have a value (otherwise result is NA). For a window that is specified by an offset, min_periods will default to 1. Otherwise, min_periods will default to the size of the window.	int Default Value : None	Required
center	Set the labels at the center of the window.	bool Default Value : False	Required
win_type	Provide a window type. If None, all points are evenly weighted. See the notes below for further information.	str Default Value : None	Required
on	For a DataFrame, a datetime-like column on which to calculate the rolling window, rather than the DataFrame’s index. Provided integer column is ignored and excluded from result since an integer index is not used to calculate the rolling window.	str	Optional
axis		int or str Default Value : 0	Required
closed	Make the interval closed on the ‘right’, ‘left’, ‘both’ or ‘neither’ endpoints. For offset-based windows, it defaults to ‘right’. For fixed windows, defaults to ‘both’. Remaining cases not implemented for fixed windows.	str Default Value : None	Required

Returns: a Window or Rolling sub-classed for the particular operation

Example:

Python-Pandas Code:

import numpy as np
import pandas as pd
df = pd.DataFrame({'Q': [0, 2, 4, np.nan, 6]})
df

Output:

    Q
0	0.0
1	2.0
2	4.0
3	NaN
4	6.0

Example - Rolling sum with a window length of 2, using the ‘triang’ window type:

Python-Pandas Code:

import numpy as np
import pandas as pd
df = pd.DataFrame({'Q': [0, 2, 4, np.nan, 6]})
df.rolling(2, win_type='triang').sum()

Output:

    Q
0	NaN
1	1.0
2	3.0
3	NaN
4	NaN

Example - Rolling sum with a window length of 1, min_periods defaults to the window length:

Python-Pandas Code:

import numpy as np
import pandas as pd
df = pd.DataFrame({'Q': [0, 2, 4, np.nan, 6]})
df.rolling(2, win_type='triang').sum()
df.rolling(1).sum()

Output:

    Q
0	0.0
1	2.0
2	4.0
3	NaN
4	6.0

Example - Same as above, but explicitly set the min_periods:

Python-Pandas Code:

import numpy as np
import pandas as pd
df = pd.DataFrame({'Q': [0, 2, 4, np.nan, 6]})
df.rolling(2, min_periods=1).sum()

Output:

Example - A ragged (meaning not-a-regular frequency), time-indexed DataFrame:

Python-Pandas Code:

import numpy as np
import pandas as pd
df = pd.DataFrame({'Q': [0, 2, 4, np.nan, 6]},
                  index = [pd.Timestamp('20190201 09:00:00'),
                           pd.Timestamp('20190201 09:00:02'),
                           pd.Timestamp('20190201 09:00:03'),
                           pd.Timestamp('20190201 09:00:05'),
                           pd.Timestamp('20190201 09:00:06')])
df

Output:

                      Q
2019-02-01 09:00:00	0.0
2019-02-01 09:00:02	2.0
2019-02-01 09:00:03	4.0
2019-02-01 09:00:05	NaN
2019-02-01 09:00:06	6.0

Example - Contrasting to an integer rolling window, this will roll a variable length window corresponding to the time period:

The default for min_periods is 1.

Python-Pandas Code:

import numpy as np
import pandas as pd
df = pd.DataFrame({'Q': [0, 2, 4, np.nan, 6]},
                  index = [pd.Timestamp('20190201 09:00:00'),
                           pd.Timestamp('20190201 09:00:02'),
                           pd.Timestamp('20190201 09:00:03'),
                           pd.Timestamp('20190201 09:00:05'),
                           pd.Timestamp('20190201 09:00:06')])
df.rolling('2s').sum()

Output:

                      Q
2019-02-01 09:00:00	0.0
2019-02-01 09:00:02	2.0
2019-02-01 09:00:03	6.0
2019-02-01 09:00:05	NaN
2019-02-01 09:00:06	6.0

Previous: Splitting the object in Pandas
Next: Expanding transformations in Pandas