pandas-dev/pandas · error · TypeError

to_numpy() got an unexpected keyword argument

Error message

to_numpy() got an unexpected keyword argument '{bad_keys}'

What it means

Raised by IndexOpsMixin.to_numpy on non-ExtensionDtype Series/Index when extra **kwargs are present. ExtensionArrays receive kwargs and may accept them (e.g. na_value); for NumPy-backed data only dtype, copy, na_value are valid, so any leftover kwarg is rejected with a TypeError naming the first offending key. This guards against silent typos like na_vlaue=np.nan being dropped.

Solutions

1. Remove the offending kwarg; the message names it.
2. Check the spelled parameter name against to_numpy signature (dtype, copy, na_value).
3. If you need EA-specific kwargs, ensure the Series actually has an ExtensionDtype (e.g. convert with .astype('Int64')).

Example fix

// before
arr = ser.to_numpy(dtype='int64', na_vlaue=0)
// after
arr = ser.to_numpy(dtype='int64', na_value=0)

import inspect
allowed = {'dtype', 'copy', 'na_value'}
extra = set(kwargs) - allowed
if extra and not isinstance(ser.dtype, pd.api.types.ExtensionDtype):
    raise TypeError(f'unsupported kwargs: {extra}')
arr = ser.to_numpy(**kwargs)

When it happens

Trigger: Calling ser.to_numpy(dtype='int64', na_value=0, foo=1) on an int Series; misspelling na_value (na_vlaue, na_val); passing a kwarg valid only for EA dtypes (e.g. dtype='datetime64[ns]' on an ExtensionDtype that already routed into the EA branch is fine, but extra kwargs on a plain ndarray-backed Series are not).

Common situations: Typing na_value incorrectly; copy-pasting kwargs from an EA example (e.g. Arrow-backed) into code that runs on NumPy-backed Series; version differences where a kwarg was removed or never existed.

Related errors

• Expected Hashable, got
• is both the pipe target and a keyword argument
• to_dict() only accepts initialized defaultdicts
• unsupported type
• Accumulation not supported for

AI-assisted analysis of pandas-dev/pandas@3b7651241d (2026-08-11). Data as JSON: /api/errors/4f0e1f0e48b43dd6. Report an issue: GitHub.

Appendix: source

Thrown at pandas/core/base.py:700

        >>> ser.to_numpy(dtype=object)
        array([Timestamp('2000-01-01 00:00:00+0100', tz='CET'),
               Timestamp('2000-01-02 00:00:00+0100', tz='CET')],
              dtype=object)

        Or ``dtype='datetime64[ns]'`` to return an ndarray of native
        datetime64 values. The values are converted to UTC and the timezone
        info is dropped.

        >>> ser.to_numpy(dtype="datetime64[ns]")
        ... # doctest: +ELLIPSIS
        array(['1999-12-31T23:00:00.000000000', '2000-01-01T23:00:00...'],
              dtype='datetime64[ns]')
        """
        if isinstance(self.dtype, ExtensionDtype):
            return self.array.to_numpy(dtype, copy=copy, na_value=na_value, **kwargs)
        elif kwargs:
            bad_keys = next(iter(kwargs.keys()))
            raise TypeError(
                f"to_numpy() got an unexpected keyword argument '{bad_keys}'"
            )

        fillna = (
            na_value is not lib.no_default
            # no need to fillna with np.nan if we already have a float dtype
            and not (na_value is np.nan and np.issubdtype(self.dtype, np.floating))
        )

        values = self._values
        if fillna and self.hasnans:
            if not can_hold_element(values, na_value):
                # if we can't hold the na_value asarray either makes a copy or we
                # error before modifying values. The asarray later on thus won't make
                # another copy
                values = np.asarray(values, dtype=dtype)
            else:
                values = values.copy()

View on GitHub (pinned to 3b7651241d)