fix: keep dash-named parameters in netcdf round trip

doc/release_notes.rst

@@ -4,6 +4,8 @@ Release Notes
 Upcoming Version
 ----------------
 
+* ``read_netcdf`` no longer drops parameters, parameter attributes or ``MultiIndex`` dimensions whose name contains ``-``. (`#1025 <https://github.com/PyPSA/linopy/pull/1025>`__)
+
 Version 0.10.0
 --------------
 

linopy/io.py

@@ -1218,9 +1218,6 @@ def read_netcdf(path: Path | str, **kwargs: Any) -> Model:
     ds = xr.load_dataset(path, **kwargs)
     m = Model(sparse=bool(ds.attrs.get("sparse", False)))
 
-    def has_prefix(k: str, prefix: str) -> bool:
-        return k.rsplit("-", 1)[0] == prefix
-
     def remove_prefix(k: str, prefix: str) -> str:
         return k[len(prefix) + 1 :]
 
@@ -1231,7 +1228,7 @@ def parse_multiindex_attr(value: str | Iterable[str]) -> list[str]:
         return [str(n) for n in value]
 
     def get_prefix(ds: xr.Dataset, prefix: str) -> xr.Dataset:
-        ds = ds[[k for k in ds if has_prefix(str(k), prefix)]]
+        ds = ds[[k for k in ds if owners[str(k)] == prefix]]
         multiindexes = []
         for dim in ds.dims:
             attr = ds.attrs.get(f"{dim}_multiindex")
@@ -1245,7 +1242,7 @@ def get_prefix(ds: xr.Dataset, prefix: str) -> xr.Dataset:
         ds.attrs = {
             remove_prefix(k, prefix): v
             for k, v in ds.attrs.items()
-            if has_prefix(k, prefix)
+            if owners[k] == prefix
         }
 
         for dim in ds.dims:
@@ -1269,15 +1266,30 @@ def container_names(kind: str) -> list[str]:
             )
         return ordered
 
+    containers = {
+        kind: container_names(kind)
+        for kind in ("variables", "expressions", "constraints")
+    }
+    prefixes = {"objective", "parameters"}.union(*containers.values())
+
+    def owner(k: str) -> str | None:
+        while "-" in k:
+            k = k.rsplit("-", 1)[0]
+            if k in prefixes:
+                return k
+        return None
+
+    owners = {k: owner(k) for k in [*map(str, ds), *ds.attrs]}
+
     variables = {}
-    for k in container_names("variables"):
+    for k in containers["variables"]:
         name = remove_prefix(k, "variables")
         variables[name] = Variable(get_prefix(ds, k), m, name)
 
     m._variables = Variables(variables, m)
 
     expressions: dict[str, LinearExpression | QuadraticExpression] = {}
-    for k in container_names("expressions"):
+    for k in containers["expressions"]:
         name = remove_prefix(k, "expressions")
         expr_ds = get_prefix(ds, k)
         expr_type = expr_ds.attrs.pop(EXPR_TYPE_ATTR, None)
@@ -1295,7 +1307,7 @@ def container_names(kind: str) -> list[str]:
     m._expressions = Expressions(expressions, m)
 
     constraints: dict[str, ConstraintBase] = {}
-    for k in container_names("constraints"):
+    for k in containers["constraints"]:
         name = remove_prefix(k, "constraints")
         con_ds = get_prefix(ds, k)
         if con_ds.attrs.get("_linopy_format") == "csr":

test/test_io.py

@@ -360,6 +360,44 @@ def test_model_to_netcdf_with_dash_names(
     assert_model_equal(m, p)
 
 
+def test_model_to_netcdf_with_dash_named_parameters(
+    model: Model, tmp_path: Path
+) -> None:
+    m = model
+    m.parameters["my-param"] = xr.DataArray(
+        [5, 6], dims=["my-dim"], coords={"my-dim": ["a", "b"]}
+    )
+    m.parameters.attrs["my-attr"] = "kept"
+    fn = tmp_path / "test.nc"
+    m.to_netcdf(fn)
+    p = read_netcdf(fn)
+
+    assert_model_equal(m, p)
+    assert p.parameters.attrs == m.parameters.attrs
+
+
+def test_model_to_netcdf_with_dash_named_multiindex_dim(tmp_path: Path) -> None:
+    from linopy.semantics import is_v1
+
+    if is_v1():
+        pytest.skip("v1 rejects MultiIndex; this model only builds under legacy")
+    index = pd.MultiIndex.from_tuples(
+        [(1, "a"), (1, "b"), (2, "a")], names=["first", "second"]
+    )
+    coords = xr.Coordinates.from_pandas_multiindex(index, "multi-dim")
+    m = Model()
+    x = m.add_variables(coords=coords, name="x")
+    m.add_objective(x.sum())
+    m.parameters["param"] = xr.DataArray([1, 2, 3], coords=coords)
+    fn = tmp_path / "test.nc"
+    m.to_netcdf(fn)
+    p = read_netcdf(fn)
+
+    assert_model_equal(m, p)
+    for index in (p.variables["x"].indexes, p.parameters.indexes):
+        assert isinstance(index["multi-dim"], pd.MultiIndex)
+
+
 def test_model_to_netcdf_with_status_and_condition(
     model_with_dash_names: Model, tmp_path: Path
 ) -> None: