lux.utils package¶
Submodules¶
lux.utils.date_utils module¶
-
lux.utils.date_utils.
compute_date_granularity
(date_column: pandas.core.series.Series)[source]¶ Given a temporal column (pandas.core.series.Series), finds out the granularity of dates.
Example
[‘2018-01-01’, ‘2019-01-02’, ‘2018-01-03’] -> “day” [‘2018-01-01’, ‘2019-02-01’, ‘2018-03-01’] -> “month” [‘2018-01-01’, ‘2019-01-01’, ‘2020-01-01’] -> “year”
- Parameters
date_column (pandas.core.series.Series) – Column series with datetime type
- Returns
field – A str specifying the granularity of dates for the inspected temporal column
- Return type
str
-
lux.utils.date_utils.
date_formatter
(time_stamp, ldf)[source]¶ Given a numpy timestamp and ldf, inspects which date granularity is appropriate and reformats timestamp accordingly
Example
For changing granularity the results differ as so. days: ‘2020-01-01’ -> ‘2020-1-1’ months: ‘2020-01-01’ -> ‘2020-1’ years: ‘2020-01-01’ -> ‘2020’
- Parameters
time_stamp (np.datetime64) – timestamp object holding the date information
ldf (lux.core.frame) – LuxDataFrame with a temporal field
- Returns
date_str – A reformatted version of the time_stamp according to granularity
- Return type
str