107107 CombineAttrsOptions ,
108108 CompatOptions ,
109109 DatetimeUnitOptions ,
110- Ellipsis ,
110+ Dims ,
111111 ErrorOptions ,
112112 ErrorOptionsWithWarn ,
113113 InterpOptions ,
@@ -1698,14 +1698,14 @@ def set_coords(self: T_Dataset, names: Hashable | Iterable[Hashable]) -> T_Datas
16981698
16991699 def reset_coords (
17001700 self : T_Dataset ,
1701- names : Hashable | Iterable [ Hashable ] | None = None ,
1701+ names : Dims = None ,
17021702 drop : bool = False ,
17031703 ) -> T_Dataset :
17041704 """Given names of coordinates, reset them to become variables
17051705
17061706 Parameters
17071707 ----------
1708- names : hashable or iterable of hashable , optional
1708+ names : str, Iterable of Hashable or None , optional
17091709 Name(s) of non-index coordinates in this dataset to reset into
17101710 variables. By default, all non-index coordinates are reset.
17111711 drop : bool, default: False
@@ -4457,7 +4457,7 @@ def _get_stack_index(
44574457
44584458 def _stack_once (
44594459 self : T_Dataset ,
4460- dims : Sequence [Hashable | Ellipsis ],
4460+ dims : Sequence [Hashable | ellipsis ],
44614461 new_dim : Hashable ,
44624462 index_cls : type [Index ],
44634463 create_index : bool | None = True ,
@@ -4516,10 +4516,10 @@ def _stack_once(
45164516
45174517 def stack (
45184518 self : T_Dataset ,
4519- dimensions : Mapping [Any , Sequence [Hashable | Ellipsis ]] | None = None ,
4519+ dimensions : Mapping [Any , Sequence [Hashable | ellipsis ]] | None = None ,
45204520 create_index : bool | None = True ,
45214521 index_cls : type [Index ] = PandasMultiIndex ,
4522- ** dimensions_kwargs : Sequence [Hashable | Ellipsis ],
4522+ ** dimensions_kwargs : Sequence [Hashable | ellipsis ],
45234523 ) -> T_Dataset :
45244524 """
45254525 Stack any number of existing dimensions into a single new dimension.
@@ -4770,7 +4770,7 @@ def _unstack_full_reindex(
47704770
47714771 def unstack (
47724772 self : T_Dataset ,
4773- dim : Hashable | Iterable [ Hashable ] | None = None ,
4773+ dim : Dims = None ,
47744774 fill_value : Any = xrdtypes .NA ,
47754775 sparse : bool = False ,
47764776 ) -> T_Dataset :
@@ -4782,7 +4782,7 @@ def unstack(
47824782
47834783 Parameters
47844784 ----------
4785- dim : hashable or iterable of hashable , optional
4785+ dim : str, Iterable of Hashable or None , optional
47864786 Dimension(s) over which to unstack. By default unstacks all
47874787 MultiIndexes.
47884788 fill_value : scalar or dict-like, default: nan
@@ -4860,15 +4860,13 @@ def unstack(
48604860 for v in nonindexes
48614861 )
48624862
4863- for dim in dims :
4863+ for d in dims :
48644864 if needs_full_reindex :
48654865 result = result ._unstack_full_reindex (
4866- dim , stacked_indexes [dim ], fill_value , sparse
4866+ d , stacked_indexes [d ], fill_value , sparse
48674867 )
48684868 else :
4869- result = result ._unstack_once (
4870- dim , stacked_indexes [dim ], fill_value , sparse
4871- )
4869+ result = result ._unstack_once (d , stacked_indexes [d ], fill_value , sparse )
48724870 return result
48734871
48744872 def update (self : T_Dataset , other : CoercibleMapping ) -> T_Dataset :
@@ -5324,15 +5322,15 @@ def drop_isel(self: T_Dataset, indexers=None, **indexers_kwargs) -> T_Dataset:
53245322
53255323 def drop_dims (
53265324 self : T_Dataset ,
5327- drop_dims : Hashable | Iterable [Hashable ],
5325+ drop_dims : str | Iterable [Hashable ],
53285326 * ,
53295327 errors : ErrorOptions = "raise" ,
53305328 ) -> T_Dataset :
53315329 """Drop dimensions and associated variables from this dataset.
53325330
53335331 Parameters
53345332 ----------
5335- drop_dims : hashable or iterable of hashable
5333+ drop_dims : str or Iterable of Hashable
53365334 Dimension or dimensions to drop.
53375335 errors : {"raise", "ignore"}, default: "raise"
53385336 If 'raise', raises a ValueError error if any of the
@@ -5763,7 +5761,7 @@ def combine_first(self: T_Dataset, other: T_Dataset) -> T_Dataset:
57635761 def reduce (
57645762 self : T_Dataset ,
57655763 func : Callable ,
5766- dim : Hashable | Iterable [ Hashable ] = None ,
5764+ dim : Dims | ellipsis = None ,
57675765 * ,
57685766 keep_attrs : bool | None = None ,
57695767 keepdims : bool = False ,
@@ -5778,8 +5776,8 @@ def reduce(
57785776 Function which can be called in the form
57795777 `f(x, axis=axis, **kwargs)` to return the result of reducing an
57805778 np.ndarray over an integer valued axis.
5781- dim : str or sequence of str , optional
5782- Dimension(s) over which to apply `func`. By default `func` is
5779+ dim : str, Iterable of Hashable or None , optional
5780+ Dimension(s) over which to apply `func`. By default `func` is
57835781 applied over all dimensions.
57845782 keep_attrs : bool or None, optional
57855783 If True, the dataset's attributes (`attrs`) will be copied from
@@ -5837,18 +5835,15 @@ def reduce(
58375835 or np .issubdtype (var .dtype , np .number )
58385836 or (var .dtype == np .bool_ )
58395837 ):
5840- reduce_maybe_single : Hashable | None | list [Hashable ]
5841- if len (reduce_dims ) == 1 :
5842- # unpack dimensions for the benefit of functions
5843- # like np.argmin which can't handle tuple arguments
5844- (reduce_maybe_single ,) = reduce_dims
5845- elif len (reduce_dims ) == var .ndim :
5846- # prefer to aggregate over axis=None rather than
5847- # axis=(0, 1) if they will be equivalent, because
5848- # the former is often more efficient
5849- reduce_maybe_single = None
5850- else :
5851- reduce_maybe_single = reduce_dims
5838+ # prefer to aggregate over axis=None rather than
5839+ # axis=(0, 1) if they will be equivalent, because
5840+ # the former is often more efficient
5841+ # keep single-element dims as list, to support Hashables
5842+ reduce_maybe_single = (
5843+ None
5844+ if len (reduce_dims ) == var .ndim and var .ndim != 1
5845+ else reduce_dims
5846+ )
58525847 variables [name ] = var .reduce (
58535848 func ,
58545849 dim = reduce_maybe_single ,
@@ -6957,7 +6952,7 @@ def sortby(
69576952 def quantile (
69586953 self : T_Dataset ,
69596954 q : ArrayLike ,
6960- dim : str | Iterable [ Hashable ] | None = None ,
6955+ dim : Dims = None ,
69616956 method : QUANTILE_METHODS = "linear" ,
69626957 numeric_only : bool = False ,
69636958 keep_attrs : bool = None ,
@@ -8303,7 +8298,9 @@ def argmin(self: T_Dataset, dim: Hashable | None = None, **kwargs) -> T_Dataset:
83038298 # Return int index if single dimension is passed, and is not part of a
83048299 # sequence
83058300 argmin_func = getattr (duck_array_ops , "argmin" )
8306- return self .reduce (argmin_func , dim = dim , ** kwargs )
8301+ return self .reduce (
8302+ argmin_func , dim = None if dim is None else [dim ], ** kwargs
8303+ )
83078304 else :
83088305 raise ValueError (
83098306 "When dim is a sequence or ..., DataArray.argmin() returns a dict. "
@@ -8361,7 +8358,9 @@ def argmax(self: T_Dataset, dim: Hashable | None = None, **kwargs) -> T_Dataset:
83618358 # Return int index if single dimension is passed, and is not part of a
83628359 # sequence
83638360 argmax_func = getattr (duck_array_ops , "argmax" )
8364- return self .reduce (argmax_func , dim = dim , ** kwargs )
8361+ return self .reduce (
8362+ argmax_func , dim = None if dim is None else [dim ], ** kwargs
8363+ )
83658364 else :
83668365 raise ValueError (
83678366 "When dim is a sequence or ..., DataArray.argmin() returns a dict. "
@@ -8469,7 +8468,7 @@ def curvefit(
84698468 self : T_Dataset ,
84708469 coords : str | DataArray | Iterable [str | DataArray ],
84718470 func : Callable [..., Any ],
8472- reduce_dims : Hashable | Iterable [ Hashable ] | None = None ,
8471+ reduce_dims : Dims = None ,
84738472 skipna : bool = True ,
84748473 p0 : dict [str , Any ] | None = None ,
84758474 bounds : dict [str , Any ] | None = None ,
@@ -8494,7 +8493,7 @@ def curvefit(
84948493 array of length `len(x)`. `params` are the fittable parameters which are optimized
84958494 by scipy curve_fit. `x` can also be specified as a sequence containing multiple
84968495 coordinates, e.g. `f((x0, x1), *params)`.
8497- reduce_dims : hashable or sequence of hashable
8496+ reduce_dims : str, Iterable of Hashable or None, optional
84988497 Additional dimension(s) over which to aggregate while fitting. For example,
84998498 calling `ds.curvefit(coords='time', reduce_dims=['lat', 'lon'], ...)` will
85008499 aggregate all lat and lon points and fit the specified function along the
@@ -8545,6 +8544,7 @@ def curvefit(
85458544 if kwargs is None :
85468545 kwargs = {}
85478546
8547+ reduce_dims_ : list [Hashable ]
85488548 if not reduce_dims :
85498549 reduce_dims_ = []
85508550 elif isinstance (reduce_dims , str ) or not isinstance (reduce_dims , Iterable ):
0 commit comments