fix: add_variables ignoring coords for DataArray bounds (#614)

* Fix add_variables silently ignoring coords for DataArray bounds

When DataArray bounds were passed to add_variables with explicit coords,
the coords parameter was silently ignored because as_dataarray skips
conversion for DataArray inputs. Now validates DataArray bounds against
coords: raises ValueError on mismatched or extra dimensions, and
broadcasts missing dimensions via expand_dims.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Another test

* Add additional test coverage for DataArray bounds validation

Test MultiIndex coords (validation skip), xarray Coordinates object,
dims-only DataArrays, and upper bound mismatch detection.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add TODO noting as_dataarray fails for scalars with dict coords

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Fix as_dataarray for scalars with dict coords

Infer dims from dict keys when dims is None and the input is a scalar.
Previously this raised xarray's CoordinateValidationError because
xarray can't broadcast a 0-dim value to coords without explicit dims.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Replace individual tests with parameterized test suite

Consolidate add_variables tests into TestAddVariablesBoundsWithCoords
class with parameterized tests covering all bound types (scalar,
np.number, numpy, pandas, list, DataArray, DataArray-no-coords) x
both coord formats (sequence, dict). Also fixes as_dataarray for
scalars with dict coords by inferring dims from dict keys.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Assert broadcast test checks actual values, not NaN

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add mixed bound type combination and edge case tests

Test DataArray+numpy, DataArray+scalar, DataArray+DataArray combos for
lower/upper. Also test both bounds covering different dim subsets with
broadcast, and that only the mismatched bound raises ValueError.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add 0-dim bound types and fix numpy_to_dataarray with dict coords

Add numpy-0d and dataarray-0d to the parameterized bound type tests.
Fix numpy_to_dataarray to infer dims from dict keys for 0-dim arrays,
matching the scalar fix in as_dataarray.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add tests for inferred coords, multi-dim, string/datetime coords

Cover three gaps: coords inferred from bounds (no coords arg) for
DataArray and pandas, multi-dimensional coord specifications with
both scalar and DataArray bounds, and real-world coordinate types
(string regions, datetime index) including mismatch detection.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Add test for bounds with different dimension order

Verify lower(time, space) and upper(space, time) align correctly
via xarray broadcast.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Reindex DataArray bounds with reordered coordinates

When a DataArray bound has the same coordinate values as coords but in
a different order, reindex to match instead of raising ValueError.
Still raises when the values actually differ (not just reordered).

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Fix mypy errors: remove dead code branch, add type annotations

Remove unreachable hasattr(coords, "dims") branch in _coords_to_dict
(xarray Coordinates are Mappings, caught by isinstance check above).
Add Any type annotations to parameterized test arguments.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Move TestAddVariablesBoundsWithCoords to test_variable.py

Per review feedback, these tests belong in test_variable.py where they
overlap with existing variable tests.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

Author: FBumann

linopy/common.py

@@ -191,6 +191,8 @@ def numpy_to_dataarray(
     """
     # fallback case for zero dim arrays
     if arr.ndim == 0:
+        if dims is None and is_dict_like(coords):
+            dims = list(coords.keys())
         return DataArray(arr.item(), coords=coords, dims=dims, **kwargs)
 
     if isinstance(dims, Iterable | Sequence):
@@ -242,8 +244,12 @@ def as_dataarray(
     elif isinstance(arr, pl.Series):
         arr = numpy_to_dataarray(arr.to_numpy(), coords=coords, dims=dims, **kwargs)
     elif isinstance(arr, np.number):
+        if dims is None and is_dict_like(coords) and np.ndim(arr) == 0:
+            dims = list(coords.keys())
         arr = DataArray(float(arr), coords=coords, dims=dims, **kwargs)
     elif isinstance(arr, int | float | str | bool | list):
+        if dims is None and is_dict_like(coords) and np.ndim(arr) == 0:
+            dims = list(coords.keys())
         arr = DataArray(arr, coords=coords, dims=dims, **kwargs)
 
     elif not isinstance(arr, DataArray):

linopy/model.py

@@ -100,6 +100,73 @@
 logger = logging.getLogger(__name__)
 
 
+def _coords_to_dict(
+    coords: Sequence[Sequence | pd.Index | DataArray] | Mapping,
+) -> dict[str, Any]:
+    """Normalize coords to a dict mapping dim names to coordinate values."""
+    if isinstance(coords, Mapping):
+        return dict(coords)
+    # Sequence of indexes
+    result: dict[str, Any] = {}
+    for c in coords:
+        if isinstance(c, pd.Index) and c.name:
+            result[c.name] = c
+    return result
+
+
+def _validate_dataarray_bounds(arr: Any, coords: Any) -> Any:
+    """
+    Validate and expand DataArray bounds against explicit coords.
+
+    If ``arr`` is not a DataArray, return it unchanged (``as_dataarray``
+    will handle conversion). For DataArray inputs:
+
+    - Raises ``ValueError`` if the array has dimensions not in coords.
+    - Raises ``ValueError`` if shared dimension coordinates don't match.
+    - Expands missing dimensions via ``expand_dims``.
+    """
+    if not isinstance(arr, DataArray):
+        return arr
+
+    expected = _coords_to_dict(coords)
+    if not expected:
+        return arr
+
+    extra = set(arr.dims) - set(expected)
+    if extra:
+        raise ValueError(f"DataArray has extra dimensions not in coords: {extra}")
+
+    for dim, coord_values in expected.items():
+        if dim not in arr.dims:
+            continue
+        if isinstance(arr.indexes.get(dim), pd.MultiIndex):
+            continue
+        expected_idx = (
+            coord_values
+            if isinstance(coord_values, pd.Index)
+            else pd.Index(coord_values)
+        )
+        actual_idx = arr.coords[dim].to_index()
+        if not actual_idx.equals(expected_idx):
+            # Same values, different order → reindex to match expected order
+            if len(actual_idx) == len(expected_idx) and set(actual_idx) == set(
+                expected_idx
+            ):
+                arr = arr.reindex({dim: expected_idx})
+            else:
+                raise ValueError(
+                    f"Coordinates for dimension '{dim}' do not match: "
+                    f"expected {expected_idx.tolist()}, got {actual_idx.tolist()}"
+                )
+
+    # Expand missing dimensions
+    expand = {k: v for k, v in expected.items() if k not in arr.dims}
+    if expand:
+        arr = arr.expand_dims(expand)
+
+    return arr
+
+
 class Model:
     """
     Linear optimization model.
@@ -611,6 +678,10 @@ def add_variables(
                     "Semi-continuous variables require a positive scalar lower bound."
                 )
 
+        if coords is not None:
+            lower = _validate_dataarray_bounds(lower, coords)
+            upper = _validate_dataarray_bounds(upper, coords)
+
         data = Dataset(
             {
                 "lower": as_dataarray(lower, coords, **kwargs),