pandas-dev/pandas · error · ValueError

Encountered an NA value with skipna=False

Error message

Encountered an NA value with skipna=False

What it means

`ValueError` from `NDArrayBackedExtensionArray.argmin` when `skipna=False` and the backing ndarray contains NA. pandas refuses to pick a 'minimum' position when an NA is present and the user explicitly opted out of skipping NAs, because the result would be ambiguous. The check uses `self._hasna` (a cheap cache) before dispatching to the cython `nargminmax`.

Solutions

1. Pass `skipna=True` (the default) to ignore NA values: `s.argmin(skipna=True)`.
2. Drop or fill NA before computing: `s.dropna().argmin()` or `s.fillna(...).argmin()`.
3. If you truly need skipna=False, first assert the data has no NA: `if s.isna().any(): raise ...` so the failure is intentional and explained.

Example fix

// before
pos = s.argmin(skipna=False)   # s has NA -> ValueError

// after
pos = s.argmin(skipna=True)
# or
pos = s.dropna().argmin()

if not skipna and s.isna().any():
    raise ValueError('cannot compute argmin(skipna=False) with NA present')
pos = s.argmin(skipna=skipna)

def safe_for_argmin_no_skipna(s) -> bool:
    return not s.isna().any()

try:
    pos = s.argmin(skipna=False)
except ValueError as e:
    if 'skipna=False' in str(e):
        pos = s.dropna().argmin()
    else:
        raise

When it happens

Trigger: `Series.argmin(skipna=False)` or `DataFrame.idxmin(skipna=False)` on a nullable-backed (IntegerArray/Float64Array/ArrowExtensionArray/etc.) Series or frame that contains NA, when the underlying array is an `NDArrayBackedExtensionArray`. Also `np.argmin`-style internal calls routed through this method.

Common situations: Calling `idxmin`/`argmin` with `skipna=False` on real-world data containing nulls; switching from numpy-backed to nullable or Arrow dtypes where `_hasna` becomes true; assert-style code that expects a deterministic minimum on incomplete data.

Related errors

• Cannot cast NaN value to Integer dtype.
• `axis` must be fewer than the number of dimensions
• cannot convert float NaN to integer
• Cannot mask with non-boolean array containing NA / NaN…
• cannot pass mask for BooleanArray input

AI-assisted analysis of pandas-dev/pandas@3b7651241d (2026-08-11). Data as JSON: /api/errors/c70985c75434c066. Report an issue: GitHub.

Appendix: source

Thrown at pandas/core/arrays/_mixins.py:220

    ) -> npt.NDArray[np.uint64]:
        from pandas.core.util.hashing import hash_array

        values = self._ndarray
        return hash_array(
            values, encoding=encoding, hash_key=hash_key, categorize=categorize
        )

    def _cast_pointwise_result(self, values: ArrayLike) -> ArrayLike:
        if not (isinstance(values, np.ndarray) and values.dtype == object):
            values = construct_1d_object_array_from_listlike(values)  # type: ignore[arg-type]
        return lib.maybe_convert_objects(values, convert_non_numeric=True)

    # Signature of "argmin" incompatible with supertype "ExtensionArray"
    def argmin(self, axis: AxisInt = 0, skipna: bool = True):  # type: ignore[override]
        # override base class by adding axis keyword
        validate_bool_kwarg(skipna, "skipna")
        if not skipna and self._hasna:
            raise ValueError("Encountered an NA value with skipna=False")
        return nargminmax(self, "argmin", axis=axis)

    # Signature of "argmax" incompatible with supertype "ExtensionArray"
    def argmax(self, axis: AxisInt = 0, skipna: bool = True):  # type: ignore[override]
        # override base class by adding axis keyword
        validate_bool_kwarg(skipna, "skipna")
        if not skipna and self._hasna:
            raise ValueError("Encountered an NA value with skipna=False")
        return nargminmax(self, "argmax", axis=axis)

    def unique(self) -> Self:
        new_data = unique(self._ndarray)
        return self._from_backing_data(new_data)

    def sort(
        self,
        *,
        ascending: bool = True,

View on GitHub (pinned to 3b7651241d)