pandas-dev/pandas · error · AbstractMethodError

This classmethod must be defined in the concrete class {name

Error message

This classmethod must be defined in the concrete class {name}

What it means

Raised (AbstractMethodError, a NotImplementedError subclass) from the base ExtensionArray._from_sequence classmethod when a concrete ExtensionArray subclass has not overridden it. `_from_sequence` is the canonical constructor that turns a 1-D sequence of scalars into an instance; the base implementation only exists to give a clear error. The message uses methodtype='classmethod' so it reads 'This classmethod must be defined in the concrete class <name>'.

Source

Thrown at pandas/core/arrays/base.py:334

        Returns
        -------
        ExtensionArray

        See Also
        --------
        api.extensions.ExtensionArray._from_sequence_of_strings : Construct a new
            ExtensionArray from a sequence of strings.
        api.extensions.ExtensionArray._hash_pandas_object : Hook for
            hash_pandas_object.

        Examples
        --------
        >>> pd.arrays.IntegerArray._from_sequence([4, 5])
        <IntegerArray>
        [4, 5]
        Length: 2, dtype: Int64
        """
        raise AbstractMethodError(cls)

    @classmethod
    def _from_sequence_of_strings(
        cls, strings, *, dtype: ExtensionDtype, copy: bool = False
    ) -> Self:
        """
        Construct a new ExtensionArray from a sequence of strings.

        This method is used to parse string data into the appropriate
        scalar type for the ExtensionArray. It is commonly used when
        reading data from text files via parsers like ``read_csv``.

        Parameters
        ----------
        strings : Sequence
            Each element will be an instance of the scalar type for this
            array, ``cls.dtype.type``.
        dtype : ExtensionDtype

View on GitHub (pinned to 3b7651241d)

Solutions

1. Implement `_from_sequence(cls, scalars, *, dtype=None, copy=False)` in your ExtensionArray subclass and return a new instance.
2. If you are just consuming an existing array type and hit this, the type is incompletely implemented — file/fix it upstream rather than working around it.

Example fix

# before
class MyArray(ExtensionArray):
    ...
# after
class MyArray(ExtensionArray):
    @classmethod
    def _from_sequence(cls, scalars, *, dtype=None, copy=False):
        data = np.array(scalars, dtype=object)
        return cls(data)

from pandas.api.extensions import ExtensionArray
if '_from_sequence' not in vars(type(my_array)):
    raise TypeError(f"{type(my_array).__name__} does not implement _from_sequence")
type(my_array)._from_sequence(scalars)

def implements_from_sequence(arr_cls) -> bool:
    return '_from_sequence' in vars(arr_cls)

try:
    out = MyArray._from_sequence(scalars)
except AbstractMethodError:
    raise AbstractMethodError(f"Implement _from_sequence on {MyArray.__name__}")

When it happens

Trigger: Subclassing ExtensionArray without implementing `_from_sequence`, then calling `MyArray._from_sequence([...])` (directly or via pandas internals like `pandas.array`/casting).

Common situations: Writing a custom extension array and forgetting the mandatory construction classmethods; partial implementations that only override `_from_factorized`.

Related errors

• This method must be defined in the concrete class {name}
• This method must be defined in the concrete class {name}
• {type(self)} does not implement __setitem__.
• 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@3b7651241d (2026-08-11). Data as JSON: /api/errors/c5f2e0d1df1bb6e7. Report an issue: GitHub.