pola-rs/polars · warning · DeprecationWarning

using 'agg_list=True' is deprecated and will be removed in 2

Error message

using 'agg_list=True' is deprecated and will be removed in 2.0

Consider using {self}.implode() instead

What it means

Expr.map_elements(agg_list=True) raises a DeprecationWarning as a hard exception (DeprecationWarning subclasses Exception, so this propagates with a traceback; the self.implode() line after the raise is dead code). The agg_list flag pre-aggregates groups into lists before applying the function, and polars 1.x removed that convenience in favor of calling .implode() yourself. Despite being a warning class, nothing is printed — it must be caught as an exception.

Source

Thrown at py-polars/src/polars/expr/expr.py:5242

        ...     )
        ... )
        shape: (4, 3)
        ┌─────┬─────┬───────────┐
        │ a   ┆ b   ┆ a_times_b │
        │ --- ┆ --- ┆ ---       │
        │ i64 ┆ i64 ┆ i64       │
        ╞═════╪═════╪═══════════╡
        │ 5   ┆ 4   ┆ 20        │
        │ 1   ┆ 2   ┆ 2         │
        │ 0   ┆ 3   ┆ 0         │
        │ 3   ┆ 4   ┆ 12        │
        └─────┴─────┴───────────┘
        """
        if agg_list:
            msg = f"""using 'agg_list=True' is deprecated and will be removed in 2.0

Consider using {self}.implode() instead"""
            raise DeprecationWarning(msg)
            self = self.implode()

        def _wrap(sl: Sequence[pl.Series], *args: Any, **kwargs: Any) -> pl.Series:
            return function(sl[0], *args, **kwargs)

        return F.map_batches(
            [self],
            _wrap,
            return_dtype,
            is_elementwise=is_elementwise,
            returns_scalar=returns_scalar,
        )

    def map_elements(
        self,
        function: Callable[[Any], Any],
        return_dtype: PolarsDataType | pl.DataTypeExpr | None = None,
        *,

View on GitHub (pinned to df599052da)

Solutions

1. Replace agg_list=True with an explicit .implode() before mapping: expr.implode().map_elements(f) (inside aggregation contexts) or drop it if you do not want lists.
2. If the function already receives a Series per group without the flag, simply delete agg_list=True.
3. For new code needing grouped lists, use expr implode within group_by().agg().

Example fix

# before
(pl.DataFrame({'g': ['a', 'a'], 'v': [1, 2]})
 .group_by('g').agg(pl.col('v').map_elements(lambda s: s.sum(), agg_list=True)))

# after
(pl.DataFrame({'g': ['a', 'a'], 'v': [1, 2]})
 .group_by('g').agg(pl.col('v').implode().map_elements(lambda s: s.sum())))

try:
    out = expr.map_elements(f, agg_list=True)
except DeprecationWarning:
    out = expr.implode().map_elements(f)

When it happens

Trigger: Any call pl.col('a').map_elements(f, agg_list=True) or return_dtype=..., agg_list=True inside group_by/over contexts, typically legacy code from polars 0.x.

Common situations: Upgrading a codebase across the polars 1.0 boundary; old StackOverflow answers and tutorials written against 0.19 or earlier; internal libs that shipped agg_list=True for per-group list processing.

Related errors

• strategy {strategy!r} is not supported
• cannot turn {qualified_type_name(input)!r} into selector
• Cannot pass a dictionary as a single positional argument.\nI
• unexpected or unrecognised op name ({opname})\n\nPlease repo
• expressions should be passed to the `by_predicate` parameter

AI-assisted analysis of pola-rs/polars@df599052da (2026-08-16). Data as JSON: /api/errors/ca49619d94f005c5. Report an issue: GitHub.