Add data_vars=None as default.

Author: dcherian

xarray/structure/combine.py

@@ -463,15 +463,15 @@ def combine_nested(
     data_vars : {"minimal", "different", "all" or list of str}, optional
         These data variables will be concatenated together:
           * "minimal": Only data variables in which the dimension already
-            appears are included. If concatenating over a dimension _not_
-            present in any of the objects, then all data variables will
-            be concatenated along that new dimension.
+            appears are included.
           * "different": Data variables which are not equal (ignoring
             attributes) across all datasets are also concatenated (as well as
             all for which dimension already appears). Beware: this option may
             load the data payload of data variables into memory if they are not
             already loaded.
           * "all": All data variables will be concatenated.
+          * None: Means ``"all"`` if ``dim`` is not present in any of the ``objs``,
+            and ``"minimal"`` if ``dim`` is present in any of ``objs``.
           * list of dims: The listed data variables will be concatenated, in
             addition to the "minimal" data variables.
 

xarray/structure/concat.py

@@ -37,7 +37,7 @@
         JoinOptions,
     )
 
-    T_DataVars = Union[ConcatOptions, Iterable[Hashable]]
+    T_DataVars = Union[ConcatOptions, Iterable[Hashable], None]
 
 
 # TODO: replace dim: Any by 1D array_likes
@@ -98,18 +98,18 @@ def concat(
         unchanged. If dimension is provided as a Variable, DataArray or Index, its name
         is used as the dimension to concatenate along and the values are added
         as a coordinate.
-    data_vars : {"minimal", "different", "all"} or list of Hashable, optional
+    data_vars : {"minimal", "different", "all", None} or list of Hashable, optional
         These data variables will be concatenated together:
           * "minimal": Only data variables in which the dimension already
-            appears are included. If concatenating over a dimension _not_
-            present in any of the objects, then all data variables will
-            be concatenated along that new dimension.
+            appears are included.
           * "different": Data variables which are not equal (ignoring
             attributes) across all datasets are also concatenated (as well as
             all for which dimension already appears). Beware: this option may
             load the data payload of data variables into memory if they are not
             already loaded.
           * "all": All data variables will be concatenated.
+          * None: Means ``"all"`` if ``dim`` is not present in any of the ``objs``,
+            and ``"minimal"`` if ``dim`` is present in any of ``objs``.
           * list of dims: The listed data variables will be concatenated, in
             addition to the "minimal" data variables.
 
@@ -352,7 +352,7 @@ def _calc_concat_over(
     # variables to be concatenated
     concat_over = set()
     # variables checked for equality
-    equals: dict[Hashable, bool] = {}
+    equals: dict[Hashable, bool | None] = {}
     # skip merging these variables.
     #   if concatenating over a dimension 'x' that is associated with an index over 2 variables,
     #   'x' and 'y', then we assert join="equals" on `y` and don't need to merge it.
@@ -365,6 +365,11 @@ def _calc_concat_over(
     else:
         concat_over_existing_dim = False
 
+    if data_vars is None or (
+        isinstance(data_vars, CombineKwargDefault) and data_vars._value is None
+    ):
+        data_vars = "minimal" if concat_over_existing_dim else "all"
+
     concat_dim_lengths = []
     for ds in datasets:
         if concat_over_existing_dim and dim not in ds.dims and dim in ds:
@@ -380,6 +385,7 @@ def process_subset_opt(opt, subset: Literal["coords", "data_vars"]) -> None:
         compat_str = (
             compat._value if isinstance(compat, CombineKwargDefault) else compat
         )
+        assert compat_str is not None
         if isinstance(opt, str | CombineKwargDefault):
             if opt == "different":
                 if isinstance(compat, CombineKwargDefault) and compat != "override":
@@ -467,20 +473,13 @@ def process_subset_opt(opt, subset: Literal["coords", "data_vars"]) -> None:
                     )
                 )
             elif opt == "minimal":
-                # with "minimal" concatenation over a new dimension;
-                # we act like "all" for data_vars
-                if not concat_over_existing_dim and subset == "data_vars":
-                    concat_over.update(
-                        set().union(
-                            *[set(getattr(d, subset)) - set(d.dims) for d in datasets]
-                        )
-                    )
+                pass
             else:
                 raise ValueError(f"unexpected value for {subset}: {opt}")
 
             if (
                 isinstance(opt, CombineKwargDefault)
-                and opt != "minimal"
+                and opt._value is not None
                 and original != concat_over
                 and concat_over_existing_dim
             ):

xarray/structure/merge.py

@@ -148,6 +148,7 @@ def unique_variable(
         compat_str = (
             compat._value if isinstance(compat, CombineKwargDefault) else compat
         )
+        assert compat_str is not None
         # first check without comparing values i.e. no computes
         for var in variables[1:]:
             equals = getattr(out, compat_str)(var, equiv=lazy_array_equiv)

xarray/tests/test_concat.py

@@ -628,7 +628,7 @@ def test_concat_constant_index(self, data_vars) -> None:
     def test_concat_constant_index_minimal(self) -> None:
         ds1 = Dataset({"foo": 1.5}, {"y": 1})
         ds2 = Dataset({"foo": 2.5}, {"y": 1})
-        actual = concat([ds1, ds2], "new_dim", data_vars="minimal", compat="equals")
+        actual = concat([ds1, ds2], "new_dim", data_vars=None, compat="equals")
         expected = Dataset(
             {"foo": ("new_dim", [1.5, 2.5])},
             coords={"y": 1},

xarray/util/deprecation_helpers.py

@@ -35,10 +35,10 @@
 import warnings
 from collections.abc import Callable
 from functools import wraps
-from typing import Any, TypeVar
+from typing import Any, Self, TypeVar
 
 from xarray.core.options import OPTIONS
-from xarray.core.utils import ReprObject, emit_user_level_warning
+from xarray.core.utils import emit_user_level_warning
 
 T = TypeVar("T", bound=Callable)
 
@@ -155,26 +155,26 @@ class CombineKwargDefault:
     """
 
     _old: str
-    _new: str
+    _new: str | None
     _name: str
 
-    def __init__(self, *, name: str, old: str, new: str):
+    def __init__(self, *, name: str, old: str, new: str | None):
         self._name = name
         self._old = old
         self._new = new
 
-    def __repr__(self) -> str:
+    def __repr__(self) -> str | None:
         return self._value
 
-    def __eq__(self, other: ReprObject | Any) -> bool:
+    def __eq__(self, other: Self | Any) -> bool:
         return (
             self._value == other._value
-            if isinstance(other, ReprObject)
+            if isinstance(other, type(self))
             else self._value == other
         )
 
     @property
-    def _value(self) -> str:
+    def _value(self) -> str | None:
         return self._new if OPTIONS["use_new_combine_kwarg_defaults"] else self._old
 
     def __hash__(self) -> int:
@@ -212,7 +212,7 @@ def error_message(self) -> str:
         )
 
 
-_DATA_VARS_DEFAULT = CombineKwargDefault(name="data_vars", old="all", new="minimal")
+_DATA_VARS_DEFAULT = CombineKwargDefault(name="data_vars", old="all", new=None)
 _COORDS_DEFAULT = CombineKwargDefault(name="coords", old="different", new="minimal")
 _COMPAT_CONCAT_DEFAULT = CombineKwargDefault(
     name="compat", old="equals", new="override"