Array value assignment via `.values` and `.data` fails with Pandas 3.0.0 (assignment destination is read-only)

[tomvothecoder]

What happened?

With Pandas 3.0.0, value assignment through DataArray.values and DataArray.data now fails with: ValueError: assignment destination is read-only

This appears to be due to the underlying NumPy array having its WRITEABLE flag set to False with Pandas 3.0. This affects common mutation patterns used in Xarray workflows, even when working with simple in-memory datasets.

The MVCE works with Pandas 2.x, but breaks with Pandas 3.0.0.

• Pandas 2.x environment: conda create -n xr_pd2 -c conda-forge xarray netcdf4 "pandas<3.0.0" ipykernel
• Pandas 3.0.0 environment: conda create -n xr_pd3 -c conda-forge xarray netcdf4 "pandas>=3.0.0" ipykernel

What did you expect to happen?

I expected value assignment via .values and .data to continue working consistently across Pandas 2.x and 3.x.

Minimal Complete Verifiable Example

# /// script
# requires-python = ">=3.11"
# dependencies = [
#   "xarray[complete]@git+https://github.com/pydata/xarray.git@main",
# ]
# ///
#
# This script automatically imports the development branch of xarray to check for issues.
# Please delete this header if you have _not_ tested this script with `uv run`!

import xarray as xr
xr.show_versions()

import numpy as np

# Create a dummy xarray dataset
ds = xr.Dataset(
    {
        "lon": ("lon", [180, 359, 0, 1, 179, 0]),
        "lon_bnds": (("lon", "bnds"), np.random.rand(6, 2)),
    }
)

# .values[] -> ValueError: assignment destination is read-only
ds["lon"].values[-1] = 360

# .data[] -> ValueError: assignment destination is read-only
ds["lon"].data[-1] = 360

arr = ds["lon"].values
print(arr.flags["WRITEABLE"])
# False

Steps to reproduce

No response

MVCE confirmation

• Minimal example — the example is as focused as reasonably possible to demonstrate the underlying issue in xarray.
• Complete example — the example is self-contained, including all data and the text of any traceback.
• Verifiable example — the example copy & pastes into an IPython prompt or Binder notebook, returning the result.
• New issue — a search of GitHub Issues suggests this is not a duplicate.
• Recent environment — the issue occurs with the latest version of xarray and its dependencies.

Relevant log output

Anything else we need to know?

No response

Environment

Details

>>> xr.show_versions()

INSTALLED VERSIONS

commit: None
python: 3.14.2 | packaged by conda-forge | (main, Jan 26 2026, 20:19:28) [Clang 20.1.8 ]
python-bits: 64
OS: Darwin
OS-release: 24.6.0
machine: arm64
processor: arm
byteorder: little
LC_ALL: None
LANG: en_US.UTF-8
LOCALE: ('en_US', 'UTF-8')
libhdf5: 1.14.6
libnetcdf: 4.9.3

xarray: 2026.1.0
pandas: 3.0.0
numpy: 2.4.2
scipy: None
netCDF4: 1.7.4
pydap: None
h5netcdf: None
h5py: None
zarr: None
cftime: 1.6.5
nc_time_axis: None
iris: None
bottleneck: None
dask: None
distributed: None
matplotlib: None
cartopy: None
seaborn: None
numbagg: None
fsspec: None
cupy: None
pint: None
sparse: None
flox: None
numpy_groupies: None
setuptools: None
pip: 26.0
conda: None
pytest: None
mypy: None
IPython: 9.10.0
sphinx: None

[jsignell]

Hmm. I am wondering if this is actually a bug. Does it only affect IndexVariable fields? With the example you provided if you do instead:

ds["lon"][-1] = 360

you get:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In[8], line 1
----> 1 ds["lon"][-1] = 0

File ~/xarray/xarray/core/dataarray.py:906, in DataArray.__setitem__(self, key, value)
    901 # DataArray key -> Variable key
    902 key = {
    903     k: v.variable if isinstance(v, DataArray) else v
    904     for k, v in self._item_key_to_dict(key).items()
    905 }
--> 906 self.variable[key] = value

File ~/xarray/xarray/core/variable.py:2821, in IndexVariable.__setitem__(self, key, value)
   2820 def __setitem__(self, key, value):
-> 2821     raise TypeError(f"{type(self).__name__} values cannot be modified")

TypeError: IndexVariable values cannot be modified

If instead you try to assign to a data variable value it works fine:

ds.lon_bnds.values[-1] = 2

[tomvothecoder]

Does it only affect IndexVariable fields?

I updated my minimum example below and confirmed this only affects IndexVariable fields.

I'm not familiar with the internals of Xarray, but it looks like Pandas 3.0.0 now flags the underlying NumPy array of IntervalVariables as read-only. This doesn't necessarily seem like a bug, but rather an unexpected breaking change compared to Pandas 2.x.

I actually think it might be correct to raise ValueError: assignment destination is read-only, given that Xarray also raises TypeError: IndexVariable values cannot be modified in your example.

# %%
import xarray as xr
import numpy as np

# Create dummy coordinates
lon = [180, 359, 0, 1, 179, 0]
lat = [10, 20]
time = np.arange("2000-01", "2000-03", dtype="datetime64[M]")

# Create a dummy xarray dataset with 'lon', 'lat', 'time' as dimensions for 'tas'
ds = xr.Dataset(
    {
        "tas": (
            ("time", "lat", "lon"),
            np.random.rand(len(time), len(lat), len(lon)),
        ),
        "lon_bnds": (("lon", "bnds"), np.random.rand(len(lon), 2)),
    },
    coords={
        "lon": ("lon", lon),
        "lat": ("lat", lat),
        "time": ("time", time),
    },
)

# Assignment fails with IndexVariable
# ----------------------------------------
# .values[] - `ValueError: assignment destination is read-only`
ds["lon"].values[-1] = 360

# .data[] - `ValueError: assignment destination is read-only`
ds["lon"].data[-1] = 360

# %%
print(ds.lon.values.flags)
#   C_CONTIGUOUS : True
#   F_CONTIGUOUS : True
#   OWNDATA : False
#   WRITEABLE : False
#   ALIGNED : True
#   WRITEBACKIFCOPY : False

# %%
# Assignment works with data variables
# ----------------------------------------
ds["tas"].values[-1] = 1

# %%
print(ds.tas.values.flags)
#   C_CONTIGUOUS : True
#   F_CONTIGUOUS : False
#   OWNDATA : True
#   WRITEABLE : True
#   ALIGNED : True
#   WRITEBACKIFCOPY : False

I now use .assign_coords() instead of .values/.data to update IndexVariables.

[jsignell]

Right so I think the behavior is actually more correct with Pandas 3. I am going to go ahead and close this since I don't think there is anything to do from xarray's perspective.