pandas-dev/pandas · error · TypeError

'{type(self).__name__}' with dtype {self.dtype} does not sup

Error message

'{type(self).__name__}' with dtype {self.dtype} does not support operation '{name}'

What it means

ExtensionArray._reduce (base.py:2480) dispatches by looking up an attribute named `name` (e.g. 'sum', 'mean', 'median') on self; if no such method exists it raises TypeError stating the class+dtype does not support that operation. This is the generic reduction fallback used by Series.min/max/sum/mean/median/std/var/prod/sem/kurt/skew.

Source

Thrown at pandas/core/arrays/base.py:2480

        Series.kurt : Return the kurtosis.
        Series.skew : Return the skewness.

        Examples
        --------
        >>> pd.array([1, 2, 3])._reduce("min")
        np.int64(1)
        >>> pd.array([1, 2, 3])._reduce("max")
        np.int64(3)
        >>> pd.array([1, 2, 3])._reduce("sum")
        np.int64(6)
        >>> pd.array([1, 2, 3])._reduce("mean")
        np.float64(2.0)
        >>> pd.array([1, 2, 3])._reduce("median")
        np.float64(2.0)
        """
        meth = getattr(self, name, None)
        if meth is None:
            raise TypeError(
                f"'{type(self).__name__}' with dtype {self.dtype} "
                f"does not support operation '{name}'"
            )
        if name != "count":
            kwargs["skipna"] = skipna
        result = meth(**kwargs)
        if keepdims:
            if name in ["min", "max"]:
                result = self._from_sequence([result], dtype=self.dtype)
            else:
                result = np.array([result])

        return result

    def count(self):
        """
        Count the number of non-NA values in the array.

View on GitHub (pinned to 71959b8cb9)

Solutions

1. Pass numeric_only=True (or select numeric columns) so the reduction skips unsupported dtypes.
2. Convert the Series to a supported numeric dtype before reducing: s.astype('Float64').mean().
3. Implement the missing reduction method on your ExtensionArray subclass so _reduce can dispatch to it.
4. Choose a reduction that the dtype supports (e.g. count/min/max instead of mean for non-numeric).

Example fix

# before
s = pd.Series(["1", "2"], dtype="string")
s.mean()  # raises

# after
s.astype("Int64").mean()

def safe_reduce(s, name):
    import pandas as pd
    if not pd.api.types.is_numeric_dtype(s) and name not in ("count", "min", "max"):
        raise TypeError(f"{s.dtype} does not support {name}")
    return getattr(s, name)()

def supports_reduction(dtype, name) -> bool:
    import pandas as pd
    numeric_ops = {"sum","mean","median","prod","std","var","sem","kurt","skew"}
    return name in ("count","min","max") or pd.api.types.is_numeric_dtype(dtype) or name not in numeric_ops

try:
    val = s.mean()
except TypeError as e:
    if "does not support operation" in str(e):
        val = s.astype("Float64").mean()
    else:
        raise

When it happens

Trigger: Calling an unsupported reduction on an EA-backed Series, e.g. s.median() on a string-dtype EA, s.prod() on a datetime EA, or s.kurt() on a boolean EA, where the EA lacks a method of that name.

Common situations: Applying df.mean(numeric_only=False) across heterogeneous columns; calling a statistical reduction on a non-numeric Series; using a custom EA that only implemented some reductions.

Related errors

• Encountered an NA value with skipna=False
• Cannot round dtype {self.dtype} as it is non-numeric
• `axis` must be fewer than the number of dimensions ({ndim})
• cannot diff {type(arr).__name__} on axis={axis}
• {type(arr).__name__} has no 'diff' method. Convert to a suit

AI-assisted analysis of pandas-dev/pandas@71959b8cb9 (2026-08-07). Data as JSON: /api/errors/cefedee36ffe1e6d. Report an issue: GitHub.