What causes "{values.dtype} cannot be converted to {name}" in pandas-dev/pandas?

Pass an explicit numeric dtype: pd.array(values, dtype='Int64').

pandas-dev/pandas · error · TypeError

cannot be converted to

Error message

{values.dtype} cannot be converted to {name}

What it means

Raised by _coerce_to_data_and_mask when the input values have object/string dtype, lib.infer_dtype reports 'boolean', and no explicit target dtype was provided. Constructing a nullable numeric array from an object array of booleans without a dtype is ambiguous (boolean is not numeric), so pandas refuses rather than guessing.

Solutions

Pass an explicit numeric dtype: pd.array(values, dtype='Int64').
Convert the booleans to integers first: np.asarray(values, dtype=int).
If booleans are intended, use dtype='boolean' instead of a numeric NumericArray.

Example fix

// before
_coerce_to_data_and_mask(np.array([True, False], dtype=object), None, ...)   # raises
// after
_coerce_to_data_and_mask(np.array([True, False], dtype=object), 'Int64', ...)

Defensive patterns

Strategy: validation

Validate before calling

import numpy as np
from pandas.core.lib import infer_dtype
if values.dtype == object and infer_dtype(values, skipna=True) == 'boolean' and dtype is None:
    dtype = 'Int64'   # or 'boolean', per intent

Type guard

def object_array_is_safe_for_numeric(values, dtype) -> bool:
    from pandas.core.lib import infer_dtype
    return not (values.dtype == object and infer_dtype(values, skipna=True) == 'boolean' and dtype is None)

Try / catch

try:
    _coerce_to_data_and_mask(values, dtype, ...)
except TypeError as e:
    if 'cannot be converted' in str(e):
        _coerce_to_data_and_mask(values, 'Int64', ...)
    else:
        raise

Prevention

Always pass an explicit dtype when constructing nullable numeric arrays from object data.
Convert object boolean arrays to int before passing to numeric constructors.
Run infer_dtype on object columns during ingestion to plan dtype choices.

When it happens

Trigger: Calling _coerce_to_data_and_mask on an object-dtype ndarray (or list-like inferred as object) whose contents are all booleans, with dtype=None; equivalent to pd.array([True, False, True]) without specifying a numeric dtype.

Common situations: Building a nullable Int/Float array from a column that was stored as object dtype booleans; reading mixed data that inferred as boolean; forgetting to pass dtype when the source is a Python list of bools routed through object.

Related errors

AI-assisted analysis of pandas-dev/pandas@3b7651241d (2026-08-11). Data as JSON: /api/errors/4e80a43ac99dedcc. Report an issue: GitHub.

Appendix: source

Thrown at pandas/core/arrays/numeric.py:173

            values = values.astype(dtype.numpy_dtype, copy=False)

        if copy:
            values = values.copy()
            mask = mask.copy()
        return values, mask

    original = values
    if not copy:
        values = np.asarray(values)
    else:
        values = np.array(values, copy=copy)
    inferred_type = None
    if values.dtype == object or is_string_dtype(values.dtype):
        inferred_type = lib.infer_dtype(values, skipna=True)
        if inferred_type == "boolean" and dtype is None:
            # object dtype array of bools
            name = dtype_cls.__name__.strip("_")
            raise TypeError(f"{values.dtype} cannot be converted to {name}")

    elif values.dtype.kind == "b" and checker(dtype):
        # fastpath
        mask = np.zeros(len(values), dtype=np.bool_)
        if not copy:
            values = np.asarray(values, dtype=default_dtype)
        else:
            values = np.array(values, dtype=default_dtype, copy=copy)

    elif values.dtype.kind not in "iuf":
        name = dtype_cls.__name__.strip("_")
        raise TypeError(f"{values.dtype} cannot be converted to {name}")

    if values.ndim != 1:
        raise TypeError("values must be a 1D list-like")

    if mask is None:
        if values.dtype.kind in "iu":

View on GitHub (pinned to 3b7651241d)