pandas-dev/pandas · critical · ImportError

Unable to import required dependency

Error message

Unable to import required dependency {_dependency}. Please see the traceback for details.

What it means

Raised at `import pandas` time when one of the hard dependencies (`numpy` or `python-dateutil`) cannot be imported. pandas runs `__import__` on each name in `_hard_dependencies` during package init and re-raises as a clearer ImportError chained (`raise ... from _e`) to the original failure so the underlying cause stays visible in the traceback.

Solutions

1. Install the missing dependency: `python -m pip install numpy python-dateutil` (or just `pip install pandas` which pulls both).
2. Verify the import directly: `python -c "import numpy, dateutil"` to see which one fails and why.
3. Recreate the virtual environment cleanly and reinstall pandas from scratch to rule out a half-installed state.
4. If behind a proxy or offline, ensure numpy and python-dateutil wheels are available in the local index.

Example fix

// before
python -c "import pandas"
# ImportError: Unable to import required dependency numpy. ...

// after
python -m pip install numpy python-dateutil
python -c "import pandas; print(pandas.__version__)"

import importlib.util
def pandas_deps_ok() -> bool:
    return importlib.util.find_spec('numpy') is not None and importlib.util.find_spec('dateutil') is not None
# call before `import pandas` in tooling/scripts
assert pandas_deps_ok(), 'numpy and python-dateutil must be installed'

null

try:
    import pandas
except ImportError as e:
    if 'Unable to import required dependency' in str(e):
        # install missing hard dep then retry once
        ...
    raise

When it happens

Trigger: Executing `import pandas` in an environment where `numpy` or `dateutil` is not installed (or is installed but broken/unimportable). The loop at pandas/__init__.py:9-16 catches the bare ImportError and wraps it.

Common situations: Fresh virtualenv created without installing requirements; a partial/broken `pip install pandas` (wheel metadata present but deps missing); wrong interpreter selected in an IDE/CI; a corrupt numpy install from a killed build; conda/pip mix-up where dateutil was uninstalled.

Related errors

• pyarrow>= is required for PyArrow backed…
• pyarrow>= is required for PyArrow backed…
• pyarrow>= is required for PyArrow backed StringArray.
• Cannot apply ufunc to mixed DataFrame and Series inputs.
• Cannot modify read-only array

AI-assisted analysis of pandas-dev/pandas@3b7651241d (2026-08-11). Data as JSON: /api/errors/71a228704d5d2fd4. Report an issue: GitHub.

Appendix: source

Thrown at pandas/__init__.py:13

from __future__ import annotations

__docformat__ = "restructuredtext"

# Let users know if they're missing any of our hard dependencies
# except tzdata (see https://github.com/pandas-dev/pandas/issues/63264)
_hard_dependencies = ("numpy", "dateutil")

for _dependency in _hard_dependencies:
    try:
        __import__(_dependency)
    except ImportError as _e:  # pragma: no cover
        raise ImportError(
            f"Unable to import required dependency {_dependency}. "
            "Please see the traceback for details."
        ) from _e

del _hard_dependencies, _dependency

try:
    # numpy compat
    from pandas.compat import (
        is_numpy_dev as _is_numpy_dev,  # pyright: ignore[reportUnusedImport] # noqa: F401
    )
except ImportError as _err:  # pragma: no cover
    _module = _err.name
    raise ImportError(
        f"C extension: {_module} not built. If you want to import "
        "pandas from the source directory, you may need to run "
        "'python -m pip install -ve . --no-build-isolation -Ceditable-verbose=true' "
        "to build the C extensions first."

View on GitHub (pinned to 3b7651241d)