Prep for new combine kwargs: improve combine docstrings, tidy up tests

xarray/backends/api.py

@@ -1463,7 +1463,7 @@ def open_mfdataset(
         "netcdf4" over "h5netcdf" over "scipy" (customizable via
         ``netcdf_engine_order`` in ``xarray.set_options()``). A custom backend
         class (a subclass of ``BackendEntrypoint``) can also be used.
-    data_vars : {"minimal", "different", "all"} or list of str, default: "all"
+    data_vars : {"minimal", "different", "all", None} or list of str, default: "all"
         These data variables will be concatenated together:
           * "minimal": Only data variables in which the dimension already
             appears are included.
@@ -1473,9 +1473,12 @@ class (a subclass of ``BackendEntrypoint``) can also be used.
             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 ``concat_dim`` is not present in any of
+            the ``objs``, and ``"minimal"`` if ``concat_dim`` is present
+            in any of ``objs``.
           * list of str: The listed data variables will be concatenated, in
             addition to the "minimal" data variables.
-    coords : {"minimal", "different", "all"} or list of str, optional
+    coords : {"minimal", "different", "all"} or list of str, default: "different"
         These coordinate variables will be concatenated together:
          * "minimal": Only coordinates in which the dimension already appears
            are included.

xarray/structure/combine.py

@@ -509,7 +509,7 @@ def combine_nested(
         Must be the same length as the depth of the list passed to
         ``datasets``.
     compat : {"identical", "equals", "broadcast_equals", \
-              "no_conflicts", "override"}, optional
+              "no_conflicts", "override"}, default: "no_conflicts"
         String indicating how to compare variables of the same name for
         potential merge conflicts:
 
@@ -522,7 +522,7 @@ def combine_nested(
           must be equal. The returned dataset then contains the combination
           of all non-null values.
         - "override": skip comparing and pick variable from first dataset
-    data_vars : {"minimal", "different", "all" or list of str}, optional
+    data_vars : {"minimal", "different", "all", None} or list of str, default: "all"
         These data variables will be concatenated together:
           * "minimal": Only data variables in which the dimension already
             appears are included.
@@ -532,15 +532,16 @@ def combine_nested(
             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``.
+          * None: Means ``"all"`` if ``concat_dim`` is not present in any of
+            the ``objs``, and ``"minimal"`` if ``concat_dim`` is present
+            in any of ``objs``.
           * list of dims: The listed data variables will be concatenated, in
             addition to the "minimal" data variables.
 
-    coords : {"minimal", "different", "all" or list of str}, optional
+    coords : {"minimal", "different", "all"} or list of str, default: "different"
         These coordinate variables will be concatenated together:
-          * "minimal": Only coordinates in which the dimension already appears
-            are included. If concatenating over a dimension _not_
+          * "minimal": Only coordinates 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.
           * "different": Coordinates which are not equal (ignoring attributes)
@@ -557,7 +558,7 @@ def combine_nested(
         Value to use for newly missing values. If a dict-like, maps
         variable names to fill values. Use a data array's name to
         refer to its values.
-    join : {"outer", "inner", "left", "right", "exact"}, optional
+    join : {"outer", "inner", "left", "right", "exact"}, default: "outer"
         String indicating how to combine differing indexes
         (excluding concat_dim) in objects
 
@@ -836,7 +837,8 @@ def combine_by_coords(
     data_objects : Iterable of Datasets or DataArrays
         Data objects to combine.
 
-    compat : {"identical", "equals", "broadcast_equals", "no_conflicts", "override"}, optional
+    compat : {"identical", "equals", "broadcast_equals", "no_conflicts", "override"}, \
+             default: "no_conflicts"
         String indicating how to compare variables of the same name for
         potential conflicts:
 
@@ -850,9 +852,8 @@ def combine_by_coords(
           of all non-null values.
         - "override": skip comparing and pick variable from first dataset
 
-    data_vars : {"minimal", "different", "all" or list of str}, optional
+    data_vars : {"minimal", "different", "all", None} or list of str, default: "all"
         These data variables will be concatenated together:
-
         - "minimal": Only data variables in which the dimension already
           appears are included.
         - "different": Data variables which are not equal (ignoring
@@ -861,18 +862,32 @@ def combine_by_coords(
           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 ``concat_dim`` is not present in any of
+          the ``objs``, and ``"minimal"`` if ``concat_dim`` is present
+          in any of ``objs``.
         - list of str: The listed data variables will be concatenated, in
           addition to the "minimal" data variables.
-
-        If objects are DataArrays, `data_vars` must be "all".
-    coords : {"minimal", "different", "all"} or list of str, optional
-        As per the "data_vars" kwarg, but for coordinate variables.
+    coords : {"minimal", "different", "all"} or list of str, default: "different"
+        These coordinate variables will be concatenated together:
+        - "minimal": Only coordinates 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.
+        - "different": Coordinates 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 coordinate variables into memory if they are not already
+          loaded.
+        - "all": All coordinate variables will be concatenated, except
+          those corresponding to other dimensions.
+        - list of Hashable: The listed coordinate variables will be concatenated,
+          in addition to the "minimal" coordinates.
     fill_value : scalar or dict-like, optional
         Value to use for newly missing values. If a dict-like, maps
         variable names to fill values. Use a data array's name to
         refer to its values. If None, raises a ValueError if
         the passed Datasets do not create a complete hypercube.
-    join : {"outer", "inner", "left", "right", "exact"}, optional
+    join : {"outer", "inner", "left", "right", "exact"}, default: "outer"
         String indicating how to combine differing indexes in objects
 
         - "outer": use the union of object indexes

xarray/structure/concat.py

@@ -114,7 +114,7 @@ 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", None} or list of Hashable, optional
+    data_vars : {"minimal", "different", "all", None} or list of Hashable, default: "all"
         These data variables will be concatenated together:
           * "minimal": Only data variables in which the dimension already
             appears are included.
@@ -129,8 +129,8 @@ def concat(
           * list of dims: The listed data variables will be concatenated, in
             addition to the "minimal" data variables.
 
-        If objects are DataArrays, data_vars must be "all".
-    coords : {"minimal", "different", "all"} or list of Hashable, optional
+        If objects are DataArrays, data_vars must be "all" or None.
+    coords : {"minimal", "different", "all"} or list of Hashable, default: "different"
         These coordinate variables will be concatenated together:
           * "minimal": Only coordinates in which the dimension already appears
             are included.
@@ -143,7 +143,8 @@ def concat(
             those corresponding to other dimensions.
           * list of Hashable: The listed coordinate variables will be concatenated,
             in addition to the "minimal" coordinates.
-    compat : {"identical", "equals", "broadcast_equals", "no_conflicts", "override"}, optional
+    compat : {"identical", "equals", "broadcast_equals", "no_conflicts", "override"}, \
+             default: "equals"
         String indicating how to compare non-concatenated variables of the same name for
         potential conflicts. This is passed down to merge.
 
@@ -164,7 +165,7 @@ def concat(
         Value to use for newly missing values. If a dict-like, maps
         variable names to fill values. Use a data array's name to
         refer to its values.
-    join : {"outer", "inner", "left", "right", "exact"}, optional
+    join : {"outer", "inner", "left", "right", "exact"}, default: "outer"
         String indicating how to combine differing indexes
         (excluding dim) in objects
 

xarray/tests/test_concat.py

@@ -979,8 +979,12 @@ def test_concat_do_not_promote(self) -> None:
             Dataset({"y": ("t", [1])}, {"x": 1, "t": [0]}),
             Dataset({"y": ("t", [2])}, {"x": 2, "t": [0]}),
         ]
-        with pytest.raises(ValueError):
-            concat(objs, "t", coords="minimal")
+        with set_options(use_new_combine_kwarg_defaults=False):
+            with pytest.raises(ValueError):
+                concat(objs, "t", coords="minimal")
+        with set_options(use_new_combine_kwarg_defaults=True):
+            with pytest.raises(ValueError):
+                concat(objs, "t", compat="equals")
 
     def test_concat_dim_is_variable(self) -> None:
         objs = [Dataset({"x": 0}), Dataset({"x": 1})]
@@ -1664,8 +1668,17 @@ def test_concat_datatree_along_existing_dim_defaults(self):
             FutureWarning, match="will change from data_vars='all' to data_vars=None"
         ):
             actual = concat([dt1, dt2], dim="x")
+
         assert actual.identical(expected)
 
+        with set_options(use_new_combine_kwarg_defaults=True):
+            expected = DataTree.from_dict(
+                data={"/a": ("x", [1, 2]), "/b": 3}, coords={"/x": [0, 1]}
+            )
+            actual = concat([dt1, dt2], dim="x")
+
+            assert actual.identical(expected)
+
     def test_concat_datatree_isomorphic_error(self):
         dt1 = DataTree.from_dict(data={"/data": ("x", [1]), "/a": None})
         dt2 = DataTree.from_dict(data={"/data": ("x", [2]), "/b": None})

xarray/tests/test_dask.py

@@ -471,7 +471,16 @@ def test_concat_loads_variables(self):
         assert isinstance(out["d"].data, dask.array.Array)
         assert isinstance(out["c"].data, dask.array.Array)
 
-        out = xr.concat([ds1, ds2, ds3], dim="n", data_vars=[], coords=[])
+        with xr.set_options(use_new_combine_kwarg_defaults=True):
+            out = xr.concat([ds1, ds2, ds3], dim="n", data_vars=[], coords=[])
+            # no extra kernel calls
+            assert kernel_call_count == 6
+            assert isinstance(out["d"].data, dask.array.Array)
+            assert isinstance(out["c"].data, dask.array.Array)
+
+        out = xr.concat(
+            [ds1, ds2, ds3], dim="n", data_vars=[], coords=[], compat="equals"
+        )
         # variables are loaded once as we are validating that they're identical
         assert kernel_call_count == 12
         assert isinstance(out["d"].data, np.ndarray)

xarray/tests/test_merge.py

@@ -528,7 +528,7 @@ def test_merge_coordinates(self):
     def test_merge_error(self):
         ds = xr.Dataset({"x": 0})
         with pytest.raises(xr.MergeError):
-            xr.merge([ds, ds + 1])
+            xr.merge([ds, ds + 1], compat="no_conflicts")
 
     def test_merge_alignment_error(self):
         ds = xr.Dataset(coords={"x": [1, 2]})
@@ -624,7 +624,7 @@ def test_merge(self):
         assert_identical(data, actual)
 
         with pytest.raises(ValueError, match="conflicting values for variable"):
-            ds1.merge(ds2.rename({"var3": "var1"}))
+            ds1.merge(ds2.rename({"var3": "var1"}), compat="no_conflicts")
         with pytest.raises(ValueError, match=r"should be coordinates or not"):
             data.reset_coords().merge(data)
         with pytest.raises(ValueError, match=r"should be coordinates or not"):
@@ -948,7 +948,7 @@ def test_merge_error_includes_path(self) -> None:
                 "Raised whilst mapping function over node(s) with path 'a'"
             ),
         ):
-            xr.merge([tree1, tree2], join="exact")
+            xr.merge([tree1, tree2], join="exact", compat="no_conflicts")
 
     def test_fill_value_errors(self) -> None:
         trees = [xr.DataTree(), xr.DataTree()]

xarray/tests/test_units.py

@@ -785,7 +785,7 @@ def test_combine_by_coords(variant, unit, error, dtype):
 
     if error is not None:
         with pytest.raises(error):
-            xr.combine_by_coords([ds, other])
+            xr.combine_by_coords([ds, other], coords="different", compat="no_conflicts")
 
         return
 
@@ -825,12 +825,6 @@ def test_combine_by_coords(variant, unit, error, dtype):
         "coords",
     ),
 )
-@pytest.mark.filterwarnings(
-    "ignore:.*the default value for join will change:FutureWarning"
-)
-@pytest.mark.filterwarnings(
-    "ignore:.*the default value for compat will change:FutureWarning"
-)
 def test_combine_nested(variant, unit, error, dtype):
     original_unit = unit_registry.m
 
@@ -890,7 +884,7 @@ def test_combine_nested(variant, unit, error, dtype):
         },
     )
 
-    func = function(xr.combine_nested, concat_dim=["x", "y"])
+    func = function(xr.combine_nested, concat_dim=["x", "y"], join="outer")
     if error is not None:
         with pytest.raises(error):
             func([[ds1, ds2], [ds3, ds4]])
@@ -1071,12 +1065,6 @@ def test_concat_dataset(variant, unit, error, dtype):
         "coords",
     ),
 )
-@pytest.mark.filterwarnings(
-    "ignore:.*the default value for join will change:FutureWarning"
-)
-@pytest.mark.filterwarnings(
-    "ignore:.*the default value for compat will change:FutureWarning"
-)
 def test_merge_dataarray(variant, unit, error, dtype):
     original_unit = unit_registry.m
 
@@ -1128,9 +1116,10 @@ def test_merge_dataarray(variant, unit, error, dtype):
         dims=("y", "z"),
     )
 
+    func = function(xr.merge, compat="no_conflicts", join="outer")
     if error is not None:
         with pytest.raises(error):
-            xr.merge([arr1, arr2, arr3])
+            func([arr1, arr2, arr3])
 
         return
 
@@ -1146,13 +1135,13 @@ def test_merge_dataarray(variant, unit, error, dtype):
     convert_and_strip = lambda arr: strip_units(convert_units(arr, units))
 
     expected = attach_units(
-        xr.merge(
+        func(
             [convert_and_strip(arr1), convert_and_strip(arr2), convert_and_strip(arr3)]
         ),
         units,
     )
 
-    actual = xr.merge([arr1, arr2, arr3])
+    actual = func([arr1, arr2, arr3])
 
     assert_units_equal(expected, actual)
     assert_allclose(expected, actual)
@@ -1181,12 +1170,6 @@ def test_merge_dataarray(variant, unit, error, dtype):
         "coords",
     ),
 )
-@pytest.mark.filterwarnings(
-    "ignore:.*the default value for join will change:FutureWarning"
-)
-@pytest.mark.filterwarnings(
-    "ignore:.*the default value for compat will change:FutureWarning"
-)
 def test_merge_dataset(variant, unit, error, dtype):
     original_unit = unit_registry.m
 
@@ -1235,7 +1218,7 @@ def test_merge_dataset(variant, unit, error, dtype):
         },
     )
 
-    func = function(xr.merge)
+    func = function(xr.merge, compat="no_conflicts", join="outer")
     if error is not None:
         with pytest.raises(error):
             func([ds1, ds2, ds3])
@@ -5607,9 +5590,6 @@ def test_content_manipulation(self, func, variant, dtype):
             "coords",
         ),
     )
-    @pytest.mark.filterwarnings(
-        "ignore:.*the default value for join will change:FutureWarning"
-    )
     @pytest.mark.filterwarnings(
         "ignore:.*the default value for compat will change:FutureWarning"
     )
@@ -5651,13 +5631,15 @@ def test_merge(self, variant, unit, error, dtype):
 
         if error is not None:
             with pytest.raises(error):
-                left.merge(right)
+                left.merge(right, compat="no_conflicts", join="outer")
 
             return
 
         converted = convert_units(right, units)
-        expected = attach_units(strip_units(left).merge(strip_units(converted)), units)
-        actual = left.merge(right)
+        expected = attach_units(
+            strip_units(left).merge(strip_units(converted), join="outer"), units
+        )
+        actual = left.merge(right, join="outer")
 
         assert_units_equal(expected, actual)
         assert_equal(expected, actual)