pandas-dev/pandas · error · TypeError

Cannot construct from scalar data. Pass a sequence instead.

Error message

Cannot construct {type(self).__name__} from scalar data. Pass a sequence instead.

What it means

Raised as TypeError by SparseArray.__new__/__init__ when is_scalar(data) is True. SparseArray models a 1-D sequence of values plus a fill value; a single scalar has no length to sparsify, so the constructor refuses it.

Solutions

1. Wrap the scalar in a list: pd.arrays.SparseArray([value]).
2. Repeat to a known length: pd.arrays.SparseArray([value] * n).
3. Build a length-1 Series with dtype='Sparse[...]' instead.

Example fix

# before
pd.arrays.SparseArray(0)
# after
pd.arrays.SparseArray([0])

import numpy as np
from pandas.api.types import is_scalar

def is_sequence_data(data) -> bool:
    return not is_scalar(data) and hasattr(data, '__len__') or isinstance(data, np.ndarray)

from collections.abc import Sized, Iterable
import numpy as np

def sparse_array_safe(data) -> bool:
    return isinstance(data, (list, tuple, range, np.ndarray, pd.Series))

try:
    arr = pd.arrays.SparseArray(data)
except TypeError as e:
    if 'scalar data' in str(e):
        arr = pd.arrays.SparseArray([data])
    else:
        raise

When it happens

Trigger: pd.arrays.SparseArray(0); pd.arrays.SparseArray(np.float64(1.0)); passing a Python int/float/str as data.

Common situations: Default-argument fallthrough that yields a scalar instead of a list; reading a single cell from a DataFrame and trying to wrap it; helper that returns a scalar where a sequence was expected.

Related errors

• NumpyExtensionArray must be 1-dimensional.
• Can only use the '.sparse' accessor with Sparse data.
• Cannot create a from a MultiIndex.
• cannot evaluate scalar only bool ops
• Cannot modify read-only array

AI-assisted analysis of pandas-dev/pandas@3b7651241d (2026-08-11). Data as JSON: /api/errors/59629e1304609e71. Report an issue: GitHub.

Appendix: source

Thrown at pandas/core/arrays/sparse/array.py:430

            # TODO: make kind=None, and use data.kind?
            data = data.sp_values

        # Handle use-provided dtype
        if isinstance(dtype, str):
            # Two options: dtype='int', regular numpy dtype
            # or dtype='Sparse[int]', a sparse dtype
            try:
                dtype = SparseDtype.construct_from_string(dtype)
            except TypeError:
                dtype = pandas_dtype(dtype)

        if isinstance(dtype, SparseDtype):
            if fill_value is None:
                fill_value = dtype.fill_value
            dtype = dtype.subtype

        if is_scalar(data):
            raise TypeError(
                f"Cannot construct {type(self).__name__} from scalar data. "
                "Pass a sequence instead."
            )

        if dtype is not None:
            dtype = pandas_dtype(dtype)

        # TODO: disentangle the fill_value dtype inference from
        # dtype inference
        if data is None:
            # TODO: What should the empty dtype be? Object or float?

            # error: Argument "dtype" to "array" has incompatible type
            # "Union[ExtensionDtype, dtype[Any], None]"; expected "Union[dtype[Any],
            # None, type, _SupportsDType, str, Union[Tuple[Any, int], Tuple[Any,
            # Union[int, Sequence[int]]], List[Any], _DTypeDict, Tuple[Any, Any]]]"
            data = np.array([], dtype=dtype)  # type: ignore[arg-type]

View on GitHub (pinned to 3b7651241d)