polars.Series.to_dummies#
- Series.to_dummies(*, separator: str = '_', drop_first: bool = False) DataFrame [source]#
Get dummy/indicator variables.
- Parameters:
- separator
Separator/delimiter used when generating column names.
- drop_first
Remove the first category from the variable being encoded.
Examples
>>> s = pl.Series("a", [1, 2, 3]) >>> s.to_dummies() shape: (3, 3) ┌─────┬─────┬─────┐ │ a_1 ┆ a_2 ┆ a_3 │ │ --- ┆ --- ┆ --- │ │ u8 ┆ u8 ┆ u8 │ ╞═════╪═════╪═════╡ │ 1 ┆ 0 ┆ 0 │ │ 0 ┆ 1 ┆ 0 │ │ 0 ┆ 0 ┆ 1 │ └─────┴─────┴─────┘
>>> s.to_dummies(drop_first=True) shape: (3, 2) ┌─────┬─────┐ │ a_2 ┆ a_3 │ │ --- ┆ --- │ │ u8 ┆ u8 │ ╞═════╪═════╡ │ 0 ┆ 0 │ │ 1 ┆ 0 │ │ 0 ┆ 1 │ └─────┴─────┘