cols

skrub.selectors.cols(*columns)

Select columns by name.

The selected columns are returned in the order they are listed in columns, not the order they appear in the dataframe. If any of the requested columns are missing from the dataframe, an exception is raised.

See also

all

Select all columns

glob

Select columns by UNIX-like name pattern

regex

Select columns by regular expression pattern

Examples

>>> from skrub import selectors as s
>>> import pandas as pd
>>> df = pd.DataFrame(
...     {
...         "height_mm": [297.0, 420.0],
...         "width_mm": [210.0, 297.0],
...         "kind": ["A4", "A3"],
...         "ID": [4, 3],
...     }
... )
>>> df
   height_mm  width_mm kind  ID
0      297.0     210.0   A4   4
1      420.0     297.0   A3   3

>>> s.select(df, s.cols('height_mm', 'ID'))
   height_mm  ID
0      297.0   4
1      420.0   3

When this selector is used on its own, an error is raised if some columns are missing:

>>> s.select(df, s.cols('width_mm', 'depth_mm'))
Traceback (most recent call last):
    ...
ValueError: The following columns are requested for selection but missing
from dataframe: ['depth_mm']

However, no error is raised when this selector is combined with other selectors:

>>> s.select(df, s.all() & s.cols('width_mm', 'depth_mm'))
   width_mm
0     210.0
1     297.0

In all skrub functions that accept a selector, a list of column names can be passed and cols will be used to turn it into a selector.

>>> s.select(df, ['kind', 'ID'])
  kind  ID
0   A4   4
1   A3   3
>>> s.make_selector(['kind', 'ID'])
cols('kind', 'ID')
>>> s.all() & ['kind', 'ID']
(all() & cols('kind', 'ID'))