groupby_operator

Groupby Operator

class tasrif.processing_pipeline.pandas.groupby_operator.GroupbyOperator(selector=None, **kwargs)

Examples

>>> import pandas as pd
>>> import numpy as np
>>> from tasrif.processing_pipeline.pandas import GroupbyOperator
>>>
>>>
>>>
>>> df = pd.DataFrame([
...     [1,'2016-03-12 01:00:00',10],
...     [1,'2016-03-12 04:00:00',250],
...     [1,'2016-03-12 06:00:00',30],
...     [1,'2016-03-12 20:00:00',10],
...     [1,'2016-03-12 23:00:00',23],
...     [2,'2016-03-12 00:05:00',20],
...     [2,'2016-03-12 19:06:00',120],
...     [2,'2016-03-12 21:07:00',100],
...     [2,'2016-03-12 23:08:00',50],
...     [3,'2016-03-12 10:00:00',300]
... ], columns=['Id', 'ActivityTime', 'Calories'])
>>>
>>> df['ActivityTime'] = pd.to_datetime(df['ActivityTime'])
>>>
>>> operator = GroupbyOperator(by='ActivityTime')
>>> df = operator.process(df)[0]
>>>
>>> print(df.get_group(1))
>>> print(df.get_group(2))
>>> print(df.get_group(3))
Id        ActivityTime  Calories
0   1 2016-03-12 01:00:00        10
1   1 2016-03-12 04:00:00       250
2   1 2016-03-12 06:00:00        30
3   1 2016-03-12 20:00:00        10
4   1 2016-03-12 23:00:00        23
...
Id        ActivityTime  Calories
5   2 2016-03-12 00:05:00        20
6   2 2016-03-12 19:06:00       120
7   2 2016-03-12 21:07:00       100
8   2 2016-03-12 23:08:00        50
Id        ActivityTime  Calories
9   3 2016-03-12 10:00:00       300

__init__(selector=None, **kwargs)

Creates a new instance of GroupbyOperator

Parameters

• selector – selects the columns of a groupby object

• **kwargs – Arguments to pandas pd.groupby function

fill_na_operator json_normalize_operator