Merge branch 'main' into open-dtree-kwargs

ci/requirements/all-but-dask.yml

@@ -22,12 +22,12 @@ dependencies:
   - netcdf4
   - numba
   - numbagg
-  - numpy
+  - numpy<2
   - packaging
   - pandas
   - pint>=0.22
   - pip
-  - pydap
+  # - pydap
   - pytest
   - pytest-cov
   - pytest-env

ci/requirements/doc.yml

@@ -21,7 +21,7 @@ dependencies:
   - nbsphinx
   - netcdf4>=1.5
   - numba
-  - numpy>=1.21
+  - numpy>=1.21,<2
   - packaging>=21.3
   - pandas>=1.4,!=2.1.0
   - pooch

ci/requirements/environment-windows.yml

@@ -23,13 +23,13 @@ dependencies:
   - netcdf4
   - numba
   - numbagg
-  - numpy
+  - numpy<2
   - packaging
   - pandas
   # - pint>=0.22
   - pip
   - pre-commit
-  - pydap
+  # - pydap
   - pytest
   - pytest-cov
   - pytest-env

ci/requirements/environment.yml

@@ -26,7 +26,7 @@ dependencies:
   - numba
   - numbagg
   - numexpr
-  - numpy
+  - numpy<2
   - opt_einsum
   - packaging
   - pandas
@@ -35,7 +35,7 @@ dependencies:
   - pooch
   - pre-commit
   - pyarrow # pandas raises a deprecation warning without this, breaking doctests
-  - pydap
+  # - pydap
   - pytest
   - pytest-cov
   - pytest-env

doc/whats-new.rst

@@ -35,6 +35,8 @@ Performance
 
 - Allow chunking for arrays with duplicated dimension names (:issue:`8759`, :pull:`9099`).
   By `Martin Raspaud <https://github.com/mraspaud>`_.
+- Extract the source url from fsspec objects (:issue:`9142`, :pull:`8923`).
+  By `Justus Magin <https://github.com/keewis>`_.
 
 Breaking changes
 ~~~~~~~~~~~~~~~~
@@ -46,7 +48,16 @@ Deprecations
 
 Bug fixes
 ~~~~~~~~~
-
+- Make :py:func:`testing.assert_allclose` work with numpy 2.0 (:issue:`9165`, :pull:`9166`).
+  By `Pontus Lurcock <https://github.com/pont-us>`_.
+- Allow diffing objects with array attributes on variables (:issue:`9153`, :pull:`9169`).
+  By `Justus Magin <https://github.com/keewis>`_.
+- Promote floating-point numeric datetimes before decoding (:issue:`9179`, :pull:`9182`).
+  By `Justus Magin <https://github.com/keewis>`_.
+- Fiy static typing of tolerance arguments by allowing `str` type (:issue:`8892`, :pull:`9194`).
+  By `Michael Niklas <https://github.com/headtr1ck>`_.
+- Dark themes are now properly detected for ``html[data-theme=dark]``-tags (:pull:`9200`).
+  By `Dieter Werthmüller <https://github.com/prisae>`_.
 
 Documentation
 ~~~~~~~~~~~~~

properties/test_properties.py

@@ -0,0 +1,17 @@
+import pytest
+
+pytest.importorskip("hypothesis")
+
+from hypothesis import given
+
+import xarray as xr
+import xarray.testing.strategies as xrst
+
+
+@given(attrs=xrst.simple_attrs)
+def test_assert_identical(attrs):
+    v = xr.Variable(dims=(), data=0, attrs=attrs)
+    xr.testing.assert_identical(v, v.copy(deep=True))
+
+    ds = xr.Dataset(attrs=attrs)
+    xr.testing.assert_identical(ds, ds.copy(deep=True))

xarray/backends/api.py

@@ -382,8 +382,11 @@ def _dataset_from_backend_dataset(
     ds.set_close(backend_ds._close)
 
     # Ensure source filename always stored in dataset object
-    if "source" not in ds.encoding and isinstance(filename_or_obj, (str, os.PathLike)):
-        ds.encoding["source"] = _normalize_path(filename_or_obj)
+    if "source" not in ds.encoding:
+        path = getattr(filename_or_obj, "path", filename_or_obj)
+
+        if isinstance(path, (str, os.PathLike)):
+            ds.encoding["source"] = _normalize_path(path)
 
     return ds
 

xarray/backends/file_manager.py

@@ -63,15 +63,15 @@ class CachingFileManager(FileManager):
     FileManager.close(), which ensures that closed files are removed from the
     cache as well.
 
-    Example usage:
+    Example usage::
 
         manager = FileManager(open, 'example.txt', mode='w')
         f = manager.acquire()
         f.write(...)
         manager.close()  # ensures file is closed
 
     Note that as long as previous files are still cached, acquiring a file
-    multiple times from the same FileManager is essentially free:
+    multiple times from the same FileManager is essentially free::
 
         f1 = manager.acquire()
         f2 = manager.acquire()

xarray/coding/times.py

@@ -278,6 +278,8 @@ def _decode_datetime_with_pandas(
     # timedelta64 value, and therefore would raise an error in the lines above.
     if flat_num_dates.dtype.kind in "iu":
         flat_num_dates = flat_num_dates.astype(np.int64)
+    elif flat_num_dates.dtype.kind in "f":
+        flat_num_dates = flat_num_dates.astype(np.float64)
 
     # Cast input ordinals to integers of nanoseconds because pd.to_timedelta
     # works much faster when dealing with integers (GH 1399).

xarray/core/alignment.py

@@ -137,7 +137,7 @@ def __init__(
         exclude_dims: str | Iterable[Hashable] = frozenset(),
         exclude_vars: Iterable[Hashable] = frozenset(),
         method: str | None = None,
-        tolerance: int | float | Iterable[int | float] | None = None,
+        tolerance: float | Iterable[float] | str | None = None,
         copy: bool = True,
         fill_value: Any = dtypes.NA,
         sparse: bool = False,
@@ -965,7 +965,7 @@ def reindex(
     obj: T_Alignable,
     indexers: Mapping[Any, Any],
     method: str | None = None,
-    tolerance: int | float | Iterable[int | float] | None = None,
+    tolerance: float | Iterable[float] | str | None = None,
     copy: bool = True,
     fill_value: Any = dtypes.NA,
     sparse: bool = False,
@@ -1004,7 +1004,7 @@ def reindex_like(
     obj: T_Alignable,
     other: Dataset | DataArray,
     method: str | None = None,
-    tolerance: int | float | Iterable[int | float] | None = None,
+    tolerance: float | Iterable[float] | str | None = None,
     copy: bool = True,
     fill_value: Any = dtypes.NA,
 ) -> T_Alignable:

xarray/core/dataarray.py

@@ -1909,7 +1909,7 @@ def reindex_like(
         other: T_DataArrayOrSet,
         *,
         method: ReindexMethodOptions = None,
-        tolerance: int | float | Iterable[int | float] | None = None,
+        tolerance: float | Iterable[float] | str | None = None,
         copy: bool = True,
         fill_value=dtypes.NA,
     ) -> Self:
@@ -1936,7 +1936,7 @@ def reindex_like(
             - backfill / bfill: propagate next valid index value backward
             - nearest: use nearest valid index value
 
-        tolerance : optional
+        tolerance : float | Iterable[float] | str | None, default: None
             Maximum distance between original and new labels for inexact
             matches. The values of the index at the matching locations must
             satisfy the equation ``abs(index[indexer] - target) <= tolerance``.
@@ -2096,7 +2096,7 @@ def reindex(
         indexers: Mapping[Any, Any] | None = None,
         *,
         method: ReindexMethodOptions = None,
-        tolerance: float | Iterable[float] | None = None,
+        tolerance: float | Iterable[float] | str | None = None,
         copy: bool = True,
         fill_value=dtypes.NA,
         **indexers_kwargs: Any,
@@ -2126,7 +2126,7 @@ def reindex(
             - backfill / bfill: propagate next valid index value backward
             - nearest: use nearest valid index value
 
-        tolerance : float | Iterable[float] | None, default: None
+        tolerance : float | Iterable[float] | str | None, default: None
             Maximum distance between original and new labels for inexact
             matches. The values of the index at the matching locations must
             satisfy the equation ``abs(index[indexer] - target) <= tolerance``.

xarray/core/dataset.py

@@ -3499,7 +3499,7 @@ def reindex_like(
         self,
         other: T_Xarray,
         method: ReindexMethodOptions = None,
-        tolerance: int | float | Iterable[int | float] | None = None,
+        tolerance: float | Iterable[float] | str | None = None,
         copy: bool = True,
         fill_value: Any = xrdtypes.NA,
     ) -> Self:
@@ -3526,7 +3526,7 @@ def reindex_like(
             - "backfill" / "bfill": propagate next valid index value backward
             - "nearest": use nearest valid index value
 
-        tolerance : optional
+        tolerance : float | Iterable[float] | str | None, default: None
             Maximum distance between original and new labels for inexact
             matches. The values of the index at the matching locations must
             satisfy the equation ``abs(index[indexer] - target) <= tolerance``.
@@ -3569,7 +3569,7 @@ def reindex(
         self,
         indexers: Mapping[Any, Any] | None = None,
         method: ReindexMethodOptions = None,
-        tolerance: int | float | Iterable[int | float] | None = None,
+        tolerance: float | Iterable[float] | str | None = None,
         copy: bool = True,
         fill_value: Any = xrdtypes.NA,
         **indexers_kwargs: Any,
@@ -3594,7 +3594,7 @@ def reindex(
             - "backfill" / "bfill": propagate next valid index value backward
             - "nearest": use nearest valid index value
 
-        tolerance : optional
+        tolerance : float | Iterable[float] | str | None, default: None
             Maximum distance between original and new labels for inexact
             matches. The values of the index at the matching locations must
             satisfy the equation ``abs(index[indexer] - target) <= tolerance``.

xarray/core/datatree.py

@@ -1314,11 +1314,12 @@ def match(self, pattern: str) -> DataTree:
         ...     }
         ... )
         >>> dt.match("*/B")
-        DataTree('None', parent=None)
-        ├── DataTree('a')
-        │   └── DataTree('B')
-        └── DataTree('b')
-            └── DataTree('B')
+        <xarray.DataTree>
+        Group: /
+        ├── Group: /a
+        │   └── Group: /a/B
+        └── Group: /b
+            └── Group: /b/B
         """
         matching_nodes = {
             node.path: node.ds

xarray/core/datatree_render.py

@@ -57,11 +57,12 @@ def __init__(self):
         >>> s0a = DataTree(name="sub0A", parent=s0)
         >>> s1 = DataTree(name="sub1", parent=root)
         >>> print(RenderDataTree(root))
-        DataTree('root', parent=None)
-        ├── DataTree('sub0')
-        │   ├── DataTree('sub0B')
-        │   └── DataTree('sub0A')
-        └── DataTree('sub1')
+        <xarray.DataTree 'root'>
+        Group: /
+        ├── Group: /sub0
+        │   ├── Group: /sub0/sub0B
+        │   └── Group: /sub0/sub0A
+        └── Group: /sub1
         """
         super().__init__("\u2502   ", "\u251c\u2500\u2500 ", "\u2514\u2500\u2500 ")
 

xarray/core/formatting.py

@@ -765,6 +765,12 @@ def _diff_mapping_repr(
     a_indexes=None,
     b_indexes=None,
 ):
+    def compare_attr(a, b):
+        if is_duck_array(a) or is_duck_array(b):
+            return array_equiv(a, b)
+        else:
+            return a == b
+
     def extra_items_repr(extra_keys, mapping, ab_side, kwargs):
         extra_repr = [
             summarizer(k, mapping[k], col_width, **kwargs[k]) for k in extra_keys
@@ -801,11 +807,7 @@ def extra_items_repr(extra_keys, mapping, ab_side, kwargs):
             is_variable = True
         except AttributeError:
             # compare attribute value
-            if is_duck_array(a_mapping[k]) or is_duck_array(b_mapping[k]):
-                compatible = array_equiv(a_mapping[k], b_mapping[k])
-            else:
-                compatible = a_mapping[k] == b_mapping[k]
-
+            compatible = compare_attr(a_mapping[k], b_mapping[k])
             is_variable = False
 
         if not compatible:
@@ -821,7 +823,11 @@ def extra_items_repr(extra_keys, mapping, ab_side, kwargs):
 
                 attrs_to_print = set(a_attrs) ^ set(b_attrs)
                 attrs_to_print.update(
-                    {k for k in set(a_attrs) & set(b_attrs) if a_attrs[k] != b_attrs[k]}
+                    {
+                        k
+                        for k in set(a_attrs) & set(b_attrs)
+                        if not compare_attr(a_attrs[k], b_attrs[k])
+                    }
                 )
                 for m in (a_mapping, b_mapping):
                     attr_s = "\n".join(
@@ -1023,20 +1029,21 @@ def diff_datatree_repr(a: DataTree, b: DataTree, compat):
 
 def _single_node_repr(node: DataTree) -> str:
     """Information about this node, not including its relationships to other nodes."""
-    node_info = f"DataTree('{node.name}')"
-
     if node.has_data or node.has_attrs:
         ds_info = "\n" + repr(node.ds)
     else:
         ds_info = ""
-    return node_info + ds_info
+    return f"Group: {node.path}{ds_info}"
 
 
 def datatree_repr(dt: DataTree):
     """A printable representation of the structure of this entire tree."""
     renderer = RenderDataTree(dt)
 
-    lines = []
+    name_info = "" if dt.name is None else f" {dt.name!r}"
+    header = f"<xarray.DataTree{name_info}>"
+
+    lines = [header]
     for pre, fill, node in renderer:
         node_repr = _single_node_repr(node)
 
@@ -1051,12 +1058,6 @@ def datatree_repr(dt: DataTree):
                 else:
                     lines.append(f"{fill}{' ' * len(renderer.style.vertical)}{line}")
 
-    # Tack on info about whether or not root node has a parent at the start
-    first_line = lines[0]
-    parent = f'"{dt.parent.name}"' if dt.parent is not None else "None"
-    first_line_with_parent = first_line[:-1] + f", parent={parent})"
-    lines[0] = first_line_with_parent
-
     return "\n".join(lines)
 
 

xarray/core/iterators.py

@@ -39,15 +39,16 @@ class LevelOrderIter(Iterator):
     >>> i = DataTree(name="i", parent=g)
     >>> h = DataTree(name="h", parent=i)
     >>> print(f)
-    DataTree('f', parent=None)
-    ├── DataTree('b')
-    │   ├── DataTree('a')
-    │   └── DataTree('d')
-    │       ├── DataTree('c')
-    │       └── DataTree('e')
-    └── DataTree('g')
-        └── DataTree('i')
-            └── DataTree('h')
+    <xarray.DataTree 'f'>
+    Group: /
+    ├── Group: /b
+    │   ├── Group: /b/a
+    │   └── Group: /b/d
+    │       ├── Group: /b/d/c
+    │       └── Group: /b/d/e
+    └── Group: /g
+        └── Group: /g/i
+            └── Group: /g/i/h
     >>> [node.name for node in LevelOrderIter(f)]
     ['f', 'b', 'g', 'a', 'd', 'i', 'c', 'e', 'h']
     >>> [node.name for node in LevelOrderIter(f, maxlevel=3)]