Make xr.corr and xr.map_blocks work without dask (#5731)

Co-authored-by: Mathias Hauser <[email protected]>
Co-authored-by: dcherian <[email protected]>
Co-authored-by: Deepak Cherian <[email protected]>

doc/whats-new.rst

@@ -34,6 +34,8 @@ Deprecations
 
 Bug fixes
 ~~~~~~~~~
+- :py:func:`xr.map_blocks` and :py:func:`xr.corr` now work when dask is not installed (:issue:`3391`, :issue:`5715`, :pull:`5731`).
+  By `Gijom <https://github.com/Gijom>`_.
 - Fix plot.line crash for data of shape ``(1, N)`` in _title_for_slice on format_item (:pull:`5948`).
   By `Sebastian Weigand <https://github.com/s-weigand>`_.
 - Fix a regression in the removal of duplicate backend entrypoints (:issue:`5944`, :pull:`5959`)

xarray/core/computation.py

@@ -1359,7 +1359,8 @@ def _get_valid_values(da, other):
             da = da.where(~missing_vals)
             return da
         else:
-            return da
+            # ensure consistent return dtype
+            return da.astype(float)
 
     da_a = da_a.map_blocks(_get_valid_values, args=[da_b])
     da_b = da_b.map_blocks(_get_valid_values, args=[da_a])

xarray/core/parallel.py

@@ -23,6 +23,7 @@
 from .alignment import align
 from .dataarray import DataArray
 from .dataset import Dataset
+from .pycompat import is_dask_collection
 
 try:
     import dask
@@ -328,13 +329,13 @@ def _wrapper(
         raise TypeError("kwargs must be a mapping (for example, a dict)")
 
     for value in kwargs.values():
-        if dask.is_dask_collection(value):
+        if is_dask_collection(value):
             raise TypeError(
                 "Cannot pass dask collections in kwargs yet. Please compute or "
                 "load values before passing to map_blocks."
             )
 
-    if not dask.is_dask_collection(obj):
+    if not is_dask_collection(obj):
         return func(obj, *args, **kwargs)
 
     all_args = [obj] + list(args)

xarray/core/pycompat.py

@@ -43,15 +43,19 @@ def __init__(self, mod):
         self.available = duck_array_module is not None
 
 
-def is_duck_dask_array(x):
+def is_dask_collection(x):
     if DuckArrayModule("dask").available:
         from dask.base import is_dask_collection
 
-        return is_duck_array(x) and is_dask_collection(x)
+        return is_dask_collection(x)
     else:
         return False
 
 
+def is_duck_dask_array(x):
+    return is_duck_array(x) and is_dask_collection(x)
+
+
 dsk = DuckArrayModule("dask")
 dask_version = dsk.version
 dask_array_type = dsk.type

xarray/tests/test_computation.py

@@ -24,8 +24,6 @@
 
 from . import has_dask, raise_if_dask_computes, requires_dask
 
-dask = pytest.importorskip("dask")
-
 
 def assert_identical(a, b):
     """A version of this function which accepts numpy arrays"""
@@ -1420,6 +1418,7 @@ def arrays_w_tuples():
     ],
 )
 @pytest.mark.parametrize("dim", [None, "x", "time"])
+@requires_dask
 def test_lazy_corrcov(da_a, da_b, dim, ddof) -> None:
     # GH 5284
     from dask import is_dask_collection
@@ -1554,6 +1553,28 @@ def test_covcorr_consistency(da_a, da_b, dim) -> None:
     assert_allclose(actual, expected)
 
 
+@requires_dask
+@pytest.mark.parametrize("da_a, da_b", arrays_w_tuples()[1])
+@pytest.mark.parametrize("dim", [None, "time", "x"])
+def test_corr_lazycorr_consistency(da_a, da_b, dim) -> None:
+    da_al = da_a.chunk()
+    da_bl = da_b.chunk()
+    c_abl = xr.corr(da_al, da_bl, dim=dim)
+    c_ab = xr.corr(da_a, da_b, dim=dim)
+    c_ab_mixed = xr.corr(da_a, da_bl, dim=dim)
+    assert_allclose(c_ab, c_abl)
+    assert_allclose(c_ab, c_ab_mixed)
+
+
+@requires_dask
+def test_corr_dtype_error():
+    da_a = xr.DataArray([[1, 2], [2, 1]], dims=["x", "time"])
+    da_b = xr.DataArray([[1, 2], [1, np.nan]], dims=["x", "time"])
+
+    xr.testing.assert_equal(xr.corr(da_a, da_b), xr.corr(da_a.chunk(), da_b.chunk()))
+    xr.testing.assert_equal(xr.corr(da_a, da_b), xr.corr(da_a, da_b.chunk()))
+
+
 @pytest.mark.parametrize(
     "da_a",
     arrays_w_tuples()[0],