dtoolkit.accessor.series.expand#

dtoolkit.accessor.series.expand(s: Series, /, suffix: list[Hashable] = None, delimiter: str = '_', flatten: bool = False) → DataFrame[source]#

Transform each element of a list-like to a column.

Parameters:

suffixlist of Hashable, optional: New columns of return DataFrame.
delimiterstr, default “_”: The delimiter between name and suffix.
flattenbool, default False: Flatten all like-list elements or not. It would cost more time.

Returns:

DataFrame: The structure of new column name is {column name}{delimiter}{suffix}.

Raises:

ValueError

If s.name is None.
If len(suffix) less than the max size of s’s elements.

See also

pandas.Series.explode: Transform each element of a list-like to a row.
pandas.DataFrame.explode: Transform each element of a list-like to a row.
dtoolkit.accessor.series.expand: Transform each element of a list-like to a column.
dtoolkit.accessor.dataframe.expand: Transform each element of a list-like to a column.

Examples

>>> import dtoolkit
>>> import pandas as pd

Expand the list-like element.

>>> s = pd.Series([[1, 2, 3], 'foo', [], [3, 4]], name="item")
>>> s.expand()
   item_0  item_1  item_2
0       1     2.0     3.0
1     foo     NaN     NaN
2    None     NaN     NaN
3       3     4.0     NaN

Expand sub-element type is list-like.

>>> s = pd.Series([("a", "b"), [1, [2, 3]]], name="item")
>>> s.expand(flatten=True)
   item_0  item_1  item_2
0       a       b     NaN
1       1       2     3.0

Set the columns of name.

>>> s = pd.Series([("a", 1), ["b", 2]], name="item")
>>> s.expand(suffix=["index", "value"], delimiter="-")
   item-index  item-value
0           a           1
1           b           2

Also could handle different lengths of element and suffix list.

>>> s = pd.Series([(1, 2), [1, 2, 3]], name="item")
>>> s.expand()
   item_0  item_1  item_2
0       1       2     NaN
1       1       2     3.0
>>> s.expand(suffix=["a", "b", "c", "d"])
   item_a  item_b  item_c
0       1       2     NaN
1       1       2     3.0