DatasetView in map_over_subtree (#269)

* re-forbid initializing a DatasetView directly

* test that map_over_subtree definitely doesn't modify data in-place

* chnge behaviour to return an immmutable DatasetView within map_over_subtree

* improve error messages

* whatsew

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* remove unused return variable

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

datatree/datatree.py

@@ -108,13 +108,10 @@ class DatasetView(Dataset):
     An immutable Dataset-like view onto the data in a single DataTree node.
 
     In-place operations modifying this object should raise an AttributeError.
+    This requires overriding all inherited constructors.
 
     Operations returning a new result will return a new xarray.Dataset object.
     This includes all API on Dataset, which will be inherited.
-
-    This requires overriding all inherited private constructors.
-
-    We leave the public init constructor because it is used by type() in some xarray code (see datatree GH issue #188)
     """
 
     # TODO what happens if user alters (in-place) a DataArray they extracted from this object?
@@ -130,6 +127,14 @@ class DatasetView(Dataset):
         "_variables",
     )
 
+    def __init__(
+        self,
+        data_vars: Optional[Mapping[Any, Any]] = None,
+        coords: Optional[Mapping[Any, Any]] = None,
+        attrs: Optional[Mapping[Any, Any]] = None,
+    ):
+        raise AttributeError("DatasetView objects are not to be initialized directly")
+
     @classmethod
     def _from_node(
         cls,
@@ -150,14 +155,16 @@ def _from_node(
 
     def __setitem__(self, key, val) -> None:
         raise AttributeError(
-            "Mutation of the DatasetView is not allowed, please use __setitem__ on the wrapping DataTree node, "
-            "or use `DataTree.to_dataset()` if you want a mutable dataset"
+            "Mutation of the DatasetView is not allowed, please use `.__setitem__` on the wrapping DataTree node, "
+            "or use `dt.to_dataset()` if you want a mutable dataset. If calling this from within `map_over_subtree`,"
+            "use `.copy()` first to get a mutable version of the input dataset."
         )
 
     def update(self, other) -> None:
         raise AttributeError(
-            "Mutation of the DatasetView is not allowed, please use .update on the wrapping DataTree node, "
-            "or use `DataTree.to_dataset()` if you want a mutable dataset"
+            "Mutation of the DatasetView is not allowed, please use `.update` on the wrapping DataTree node, "
+            "or use `dt.to_dataset()` if you want a mutable dataset. If calling this from within `map_over_subtree`,"
+            "use `.copy()` first to get a mutable version of the input dataset."
         )
 
     # FIXME https://github.com/python/mypy/issues/7328

datatree/mapping.py

@@ -190,15 +190,14 @@ def _map_over_subtree(*args, **kwargs) -> DataTree | Tuple[DataTree, ...]:
             *args_as_tree_length_iterables,
             *list(kwargs_as_tree_length_iterables.values()),
         ):
-            node_args_as_datasets = [
-                a.to_dataset() if isinstance(a, DataTree) else a
-                for a in all_node_args[:n_args]
+            node_args_as_datasetviews = [
+                a.ds if isinstance(a, DataTree) else a for a in all_node_args[:n_args]
             ]
-            node_kwargs_as_datasets = dict(
+            node_kwargs_as_datasetviews = dict(
                 zip(
                     [k for k in kwargs_as_tree_length_iterables.keys()],
                     [
-                        v.to_dataset() if isinstance(v, DataTree) else v
+                        v.ds if isinstance(v, DataTree) else v
                         for v in all_node_args[n_args:]
                     ],
                 )
@@ -210,7 +209,7 @@ def _map_over_subtree(*args, **kwargs) -> DataTree | Tuple[DataTree, ...]:
             # Now we can call func on the data in this particular set of corresponding nodes
             results = (
                 func_with_error_context(
-                    *node_args_as_datasets, **node_kwargs_as_datasets
+                    *node_args_as_datasetviews, **node_kwargs_as_datasetviews
                 )
                 if node_of_first_tree.has_data
                 else None

datatree/tests/test_mapping.py

@@ -304,7 +304,7 @@ def weighted_mean(ds):
 
         dt.map_over_subtree(weighted_mean)
 
-    def test_alter_inplace(self):
+    def test_alter_inplace_forbidden(self):
         simpsons = DataTree.from_dict(
             d={
                 "/": xr.Dataset({"age": 83}),
@@ -322,7 +322,8 @@ def fast_forward(ds: xr.Dataset, years: float) -> xr.Dataset:
             ds["age"] = ds["age"] + years
             return ds
 
-        simpsons.map_over_subtree(fast_forward, years=10)
+        with pytest.raises(AttributeError):
+            simpsons.map_over_subtree(fast_forward, years=10)
 
 
 @pytest.mark.xfail

docs/source/whats-new.rst

@@ -32,6 +32,8 @@ Breaking changes
 
 - Nodes containing only attributes but no data are now ignored by :py:func:`map_over_subtree` (:issue:`262`, :pull:`263`)
   By `Tom Nicholas <https://github.com/TomNicholas>`_.
+- Disallow altering of given dataset inside function called by :py:func:`map_over_subtree` (:pull:`269`, reverts part of :pull:`194`).
+  By `Tom Nicholas <https://github.com/TomNicholas>`_.
 
 Deprecations
 ~~~~~~~~~~~~