kra.label

class kra.label.LabelExpr(expr: Expr)

Bases: object

Expression namespace for label encoding in polars expressions.

Example

>>> import polars as pl
>>> import kra  # noqa: F401, needed for registration
>>> df = pl.DataFrame({"animal": ["cat", "dog", "cat"]})
>>> df.with_columns(pl.col("animal").label.encode().alias("encoded"))
shape: (3, 2)
┌────────┬─────────┐
│ animal ┆ encoded │
├────────┼─────────┤
│ cat    ┆ 0       │
│ dog    ┆ 1       │
│ cat    ┆ 0       │
└────────┴─────────┘

encode()

Encode string values as integer labels using categorical encoding.

Returns:

An expression that encodes string values as integer labels.

Return type:

pl.Expr

Examples

>>> import polars as pl
>>> import kra  # noqa: F401
>>> df = pl.DataFrame({"animal": ["cat", "dog", "cat"]})
>>> df.with_columns(pl.col("animal").label.encode().alias("encoded"))
shape: (3, 2)
┌────────┬─────────┐
│ animal ┆ encoded │
├────────┼─────────┤
│ cat    ┆ 0       │
│ dog    ┆ 1       │
│ cat    ┆ 0       │
└────────┴─────────┘

class kra.label.LabelSeries(series: Series)

Bases: object

Series namespace for label encoding in polars Series.

Example

>>> import polars as pl
>>> import kra  # noqa: F401, needed for registration
>>> s = pl.Series(["cat", "dog", "cat"])
>>> s.label.encode()
shape: (3,)
Series: '' [u32]
[
    0
    1
    0
]

encode()

Encode string values as integer labels using categorical encoding.

Returns:

A Series with integer labels for each unique string value.

Return type:

pl.Series

Examples

>>> import polars as pl
>>> import kra  # noqa: F401
>>> s = pl.Series(["cat", "dog", "cat"])
>>> s.label.encode()
shape: (3,)
Series: '' [u32]
[
    0
    1
    0
]