any_date

skrub.selectors.any_date()

Select columns that have a Date or Datetime data type.

See also

skrub.Cleaner

Parse and clean date columns into proper datetime types.

skrub.ToDatetime

Convert string columns to datetime types.

skrub.DatetimeEncoder

Encode datetime columns into numeric features for machine learning.

Notes

Only datetime columns are selected. Time-only, period, and duration types are not selected. Selection is based on the column’s dtype: for example string columns containing date-like values are not selected.

Selected columns depend on the dataframe library and its supported dtypes: in pandas, this selector selects columns with dtype datetime64[ns], while in polars, it selects both Date and Datetime dtypes.

Examples

>>> import datetime
>>> from skrub import selectors as s
>>> import pandas as pd

>>> df = pd.DataFrame(
...     dict(
...         dt=[datetime.datetime(2020, 3, 2, 10, 30)],
...         tzdt=[
...             datetime.datetime(2020, 3, 2, 10, 30, tzinfo=datetime.timezone.utc)
...         ],
...         str_=["2020-03-02 10:30:00"],
...     )
... )
>>> df
                   dt                      tzdt                 str_
0 2020-03-02 10:30:00 2020-03-02 10:30:00+00:00  2020-03-02 10:30:00

>>> df.dtypes
dt           datetime64[...]
tzdt    datetime64[..., UTC]
str_                     ...
dtype: object

Select all date/datetime columns:

>>> s.select(df, s.any_date())
                       dt                      tzdt
0 2020-03-02 10:30:00 2020-03-02 10:30:00+00:00

Note that string columns with date-like values are not selected (use filtering for that):

>>> s.select(df, s.any_date() | s.string())
                       dt                      tzdt                 str_
0 2020-03-02 10:30:00 2020-03-02 10:30:00+00:00  2020-03-02 10:30:00