convert_to_datetime

Operator to convert a column feature from string to datetime

class tasrif.processing_pipeline.pandas.convert_to_datetime.ConvertToDatetimeOperator(feature_names, **kwargs)

Converts a set of (string) features to datetime using Pandas to_datetime

Examples

>>> import pandas as pd
>>> from tasrif.processing_pipeline.pandas import ConvertToDatetimeOperator
>>>
>>> df0 = pd.DataFrame([[1, "2020-05-01 00:00:00", 1], [1, "2020-05-01 01:00:00", 1],
>>>                     [1, "2020-05-01 03:00:00", 2], [2, "2020-05-02 00:00:00", 1],
>>>                     [2, "2020-05-02 01:00:00", 1]],
>>>                     columns=['logId', 'timestamp', 'sleep_level'])
>>>
>>> operator = ConvertToDatetime(feature_names=["timestamp"], utc=True)
>>> df0 = operator.process(df0)
>>>
>>> print(df0)
.   logId   timestamp   sleep_level
0   1   2020-05-01 00:00:00+00:00   1
1   1   2020-05-01 01:00:00+00:00   1
2   1   2020-05-01 03:00:00+00:00   2
3   2   2020-05-02 00:00:00+00:00   1
4   2   2020-05-02 01:00:00+00:00   1

__init__(feature_names, **kwargs)

Convert a set of columns features from string to datetime

Parameters

• feature_names (str) – Name of the string columns that represent datetime objects

• **kwargs – key word arguments passed to pandas to_datetime method

concat_operator corr_operator