Assigning a DataArray with different shape to Dataset variable does not update the variable

[jolackyuan]

What happened?

I encountered an issue when trying to assign a new DataArray to an existing variable in an xarray Dataset. The assignment appears to not update the variable as expected.

Here is a minimal example that reproduces the problem:

What did you expect to happen?

After the assignment, dataSet['open']should have the same data and coordinates as dataArray_new, i.e., shape (3,3) with codes ['000123', '000456', '000111'].

Minimal Complete Verifiable Example

import xarray as xr
import numpy as np


DataTime=[20260102,20260103,20260104]
codes = ['000123', '000456']

dataArray = xr.DataArray(
    np.random.randn(len(DataTime), len(codes)),
    dims=['time', 'code'],
    coords={'code': codes,'time': DataTime}
)
dataArray_new = xr.DataArray(
    np.random.randn(len(DataTime), 3),
    dims=['time', 'code'],
    coords={'code': ['000123', '000456','000111'],'time': DataTime}
)


dataSet = xr.Dataset({'open': dataArray})
dataSet['open']=dataArray_new
print(dataSet['open'])
print()
print(dataArray_new)
print("xarray version",xr.__version__)

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

output:
Python 3.13.7 (tags/v3.13.7:bcee1c3, Aug 14 2025, 14:15:11) [MSC v.1944 64 bit (AMD64)] on win32
Enter "help" below or click "Help" above for more information.

= RESTART: C:/Users/Administrator/AppData/Local/Programs/Python/Python313/delok.py
<xarray.DataArray 'open' (time: 3, code: 2)> Size: 48B
array([[ 0.00311068, -1.68021968],
[ 0.23174369, 1.72308605],
[ 0.05651745, -0.07747445]])
Coordinates:

time (time) int64 24B 20260102 20260103 20260104
code (code) <U6 48B '000123' '000456'
<xarray.DataArray (time: 3, code: 3)> Size: 72B
array([[ 0.00311068, -1.68021968, -1.19742281],
[ 0.23174369, 1.72308605, -1.11061083],
[ 0.05651745, -0.07747445, 0.70037003]])
Coordinates:

time (time) int64 24B 20260102 20260103 20260104
code (code) <U6 72B '000123' '000456' '000111'
xarray version 2026.1.0

Anything else we need to know?

No response

Environment

Details

[kmuehlbauer]

@jolackyuan That's by purpose, xarray does label-based alignment under the hood. It doesn’t just “replace” old values, it aligns dataArray_new to the existing Dataset coordinate labels.
If there are labels in dataArray_new that are not present in the Dataset, those labels are dropped, because they don’t exist in the dataset's index. This is exactly why your '000111' code disappears: the dataset's defined coordinate set only includes ['000123', '000456'], so xarray aligns the new data to that set and drops the extra label.

Please also refer to: https://docs.xarray.dev/en/latest/user-guide/data-structures.html#dictionary-like-methods and https://docs.xarray.dev/en/latest/user-guide/combining.html#update