pandas-dev/pandas · error · IndexError

Indexing with a float is no longer supported. Manually…

Error message

Indexing with a float is no longer supported. Manually convert to an integer key instead.

What it means

Raised by cast_scalar_indexer when a scalar float that satisfies val.is_integer() (e.g. 1.0, 2.0) is used as a positional/label indexer. Pandas deprecated float indexing in GH#34193 because 1.0 == 1 caused silent confusion between label and positional access; now it is a hard IndexError. The check runs only on scalars that pass lib.is_scalar.

Solutions

Convert with int(): ser[int(idx)].
Compute the index with integer arithmetic: i = i // 1 if you want floor.
Validate isinstance(idx, (int, np.integer)) before indexing.

Example fix

// before
i = positions.mean()  # float
val = ser[i]
// after
i = int(round(positions.mean()))
val = ser[i]

Defensive patterns

Strategy: validation

Validate before calling

if isinstance(i, float) and i.is_integer():
    i = int(i)
val = ser[i]

Type guard

def is_integer_indexer(i) -> bool:
    import numpy as np
    return isinstance(i, (int, np.integer))

Prevention

Coerce float indices to int explicitly with int().
Validate isinstance(i, (int, np.integer)) at boundaries.
Avoid numpy float results as direct indices.

When it happens

Trigger: ser[1.0]; df.iloc[2.0]; ser.loc[3.0] on an integer index; idx = some_float; ser[idx] where idx happens to be a whole float.

Common situations: Math that produces float results fed directly as an index (np.mean, division, math.ceil returns int but np.divide returns float); JSON-parsed values that arrive as float; numba/numpy interop where ints are widened to float64.

Related errors

AI-assisted analysis of pandas-dev/pandas@3b7651241d (2026-08-11). Data as JSON: /api/errors/21ac28d2a9593c94. Report an issue: GitHub.

Appendix: source

Thrown at pandas/core/common.py:181

    return False


def cast_scalar_indexer(val: Any) -> Any:
    """
    Disallow indexing with a float key, even if that key is a round number.

    Parameters
    ----------
    val : scalar

    Returns
    -------
    outval : scalar
    """
    # assumes lib.is_scalar(val)
    if lib.is_float(val) and val.is_integer():
        raise IndexError(
            # GH#34193
            "Indexing with a float is no longer supported. Manually convert "
            "to an integer key instead."
        )
    return val


def not_none(*args: object) -> Generator[object]:
    """
    Returns a generator consisting of the arguments that are not None.
    """
    return (arg for arg in args if arg is not None)


def any_none(*args: object) -> bool:
    """
    Returns a boolean indicating if any argument is None.
    """

View on GitHub (pinned to 3b7651241d)