Fix map_blocks HLG layering (#3598)

* Fix map_blocks HLG layering

This fixes an issue with the HighLevelGraph noted in
#3584, and exposed by a recent
change in Dask to do more HLG fusion.

* update

* black

* update

Author: TomAugspurger

doc/whats-new.rst

@@ -36,6 +36,8 @@ Bug fixes
 ~~~~~~~~~
 - Fix plotting with transposed 2D non-dimensional coordinates. (:issue:`3138`, :pull:`3441`)
   By `Deepak Cherian <https://github.com/dcherian>`_.
+- Fix issue with Dask-backed datasets raising a ``KeyError`` on some computations involving ``map_blocks`` (:pull:`3598`)
+  By `Tom Augspurger <https://github.com/TomAugspurger>`_.
 
 Documentation
 ~~~~~~~~~~~~~

xarray/core/parallel.py

@@ -7,12 +7,14 @@
 except ImportError:
     pass
 
+import collections
 import itertools
 import operator
 from typing import (
     Any,
     Callable,
     Dict,
+    DefaultDict,
     Hashable,
     Mapping,
     Sequence,
@@ -221,7 +223,12 @@ def _wrapper(func, obj, to_array, args, kwargs):
     indexes = {dim: dataset.indexes[dim] for dim in preserved_indexes}
     indexes.update({k: template.indexes[k] for k in new_indexes})
 
+    # We're building a new HighLevelGraph hlg. We'll have one new layer
+    # for each variable in the dataset, which is the result of the
+    # func applied to the values.
+
     graph: Dict[Any, Any] = {}
+    new_layers: DefaultDict[str, Dict[Any, Any]] = collections.defaultdict(dict)
     gname = "{}-{}".format(
         dask.utils.funcname(func), dask.base.tokenize(dataset, args, kwargs)
     )
@@ -310,9 +317,20 @@ def _wrapper(func, obj, to_array, args, kwargs):
                     # unchunked dimensions in the input have one chunk in the result
                     key += (0,)
 
-            graph[key] = (operator.getitem, from_wrapper, name)
+            # We're adding multiple new layers to the graph:
+            # The first new layer is the result of the computation on
+            # the array.
+            # Then we add one layer per variable, which extracts the
+            # result for that variable, and depends on just the first new
+            # layer.
+            new_layers[gname_l][key] = (operator.getitem, from_wrapper, name)
+
+    hlg = HighLevelGraph.from_collections(gname, graph, dependencies=[dataset])
 
-    graph = HighLevelGraph.from_collections(gname, graph, dependencies=[dataset])
+    for gname_l, layer in new_layers.items():
+        # This adds in the getitems for each variable in the dataset.
+        hlg.dependencies[gname_l] = {gname}
+        hlg.layers[gname_l] = layer
 
     result = Dataset(coords=indexes, attrs=template.attrs)
     for name, gname_l in var_key_map.items():
@@ -325,7 +343,7 @@ def _wrapper(func, obj, to_array, args, kwargs):
                 var_chunks.append((len(indexes[dim]),))
 
         data = dask.array.Array(
-            graph, name=gname_l, chunks=var_chunks, dtype=template[name].dtype
+            hlg, name=gname_l, chunks=var_chunks, dtype=template[name].dtype
         )
         result[name] = (dims, data, template[name].attrs)
 

xarray/tests/test_dask.py

@@ -1189,6 +1189,19 @@ def func(obj):
     assert_identical(expected.compute(), actual.compute())
 
 
+def test_map_blocks_hlg_layers():
+    # regression test for #3599
+    ds = xr.Dataset(
+        {
+            "x": (("a",), dask.array.ones(10, chunks=(5,))),
+            "z": (("b",), dask.array.ones(10, chunks=(5,))),
+        }
+    )
+    mapped = ds.map_blocks(lambda x: x)
+
+    xr.testing.assert_equal(mapped, ds)
+
+
 def test_make_meta(map_ds):
     from ..core.parallel import make_meta