perspective-dev/perspective · error · ValueError

Unknown aggregate

Error message

Unknown aggregate '{agg_name}'

What it means

get_polars_agg_expr looks up the aggregate name in AGG_MAP to build a Polars expression; names outside the map (typos like "mean" vs "avg", or client-sent invalid config from a view request) cannot be translated, so the rollup/group-by pipeline raises ValueError before any computation runs.

Solutions

1. Use an aggregate name present in AGG_MAP (e.g. count, sum, avg, first, last)
2. Check the client's aggregates config for typos or aggregates unsupported by the Polars virtual server
3. Add the missing aggregate to AGG_MAP if it is a legitimately supported operation

When it happens

Trigger: Thrown at rust/perspective-python/perspective/virtual_servers/polars.py:364 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of perspective-dev/perspective@11c8238c0c (2026-09-09). Data as JSON: /api/errors/74c84d4eb77954d9. Report an issue: GitHub.

Appendix: source

Thrown at rust/perspective-python/perspective/virtual_servers/polars.py:364

            mask = mask & (col_expr <= value)

    return df.filter(mask)


def get_polars_agg_expr(col, agg_name, filter_expr=None):
    """Convert an aggregate name to a Polars expression."""
    if isinstance(agg_name, list):
        agg_name = agg_name[0]
    if isinstance(agg_name, dict):
        agg_name = "first"
    expr = pl.col(col)
    if filter_expr is not None:
        expr = expr.filter(filter_expr)
    if agg_name in AGG_MAP:
        return AGG_MAP[agg_name](expr)

    msg = f"Unknown aggregate '{agg_name}'"
    raise ValueError(msg)


def default_aggregate(col_name, df):
    """Return the default aggregate for a column based on its type."""
    dtype = df[col_name].dtype
    psp_type = polars_type_to_psp(dtype)
    if psp_type in ("integer", "float"):
        return "sum"
    return "count"


def build_rollup(df, group_by, columns, aggregates, col_alias):
    """Emulate GROUP BY ROLLUP using multiple group_by operations."""
    n = len(group_by)
    frames = []
    data_columns = [c for c in columns if c not in group_by]

    for level in range(n + 1):

View on GitHub (pinned to 11c8238c0c)