Closes pydata#3409

ci/azure/install.yml

@@ -16,9 +16,9 @@ steps:
         --pre \
         --upgrade \
         matplotlib \
+        numpy \
         pandas \
         scipy
-        # numpy \  # FIXME https://github.com/pydata/xarray/issues/3409
     pip install \
         --no-deps \
         --upgrade \

doc/whats-new.rst

@@ -100,6 +100,12 @@ Bug fixes
   (:issue:`3402`). By `Deepak Cherian <https://github.com/dcherian/>`_
 - Allow appending datetime and bool data variables to zarr stores.
   (:issue:`3480`). By `Akihiro Matsukawa <https://github.com/amatsukawa/>`_.
+- Add support for numpy >=1.18 (:issue:`3409`).
+  By `Guido Imperiale <https://github.com/crusaderky>`_.
+- Add support for pandas >=0.26 (:issue:`3440`).
+  By `Deepak Cherian <https://github.com/dcherian>`_.
+- Add support for pseudonetcdf >=3.1 (:pull:`3485`).
+  By `Barron Henderson <https://github.com/barronh>`_.
 
 Documentation
 ~~~~~~~~~~~~~
@@ -118,7 +124,6 @@ Documentation
 
 Internal Changes
 ~~~~~~~~~~~~~~~~
-
 - Added integration tests against `pint <https://pint.readthedocs.io/>`_.
   (:pull:`3238`, :pull:`3447`, :pull:`3493`, :pull:`3508`)
   by `Justus Magin <https://github.com/keewis>`_.

xarray/core/duck_array_ops.py

@@ -10,6 +10,7 @@
 
 import numpy as np
 import pandas as pd
+from distutils.version import LooseVersion
 
 from . import dask_array_ops, dtypes, npcompat, nputils
 from .nputils import nanfirst, nanlast
@@ -351,6 +352,32 @@ def f(values, axis=None, skipna=None, **kwargs):
 _mean = _create_nan_agg_method("mean")
 
 
+def _datetime_crude_nanmin(array):
+    """Implement nanmin for datetime64 arrays, with caveats:
+
+    - can't accept an axis parameter
+    - will return incorrect results if the array exclusively contains NaT
+    """
+    if LooseVersion(np.__version__) < "1.18":
+        # numpy.min < 1.18 incorrectly skips NaT - which we exploit here
+        return min(array, skipna=False)
+    # This requires numpy >= 1.15
+
+    from .dataarray import DataArray
+    from .variable import Variable
+
+    if isinstance(array, (DataArray, Variable)):
+        array = array.data
+    array = array.ravel()
+    array = array[~pandas_isnull(array)]
+
+    assert array.dtype.kind in "Mm"
+    assert array.dtype.itemsize == 8
+    initial = np.array(2 ** 63 - 1, dtype=array.dtype)
+
+    return min(array, initial=initial, skipna=False)
+
+
 def datetime_to_numeric(array, offset=None, datetime_unit=None, dtype=float):
     """Convert an array containing datetime-like data to an array of floats.
 
@@ -370,7 +397,10 @@ def datetime_to_numeric(array, offset=None, datetime_unit=None, dtype=float):
     """
     # TODO: make this function dask-compatible?
     if offset is None:
-        offset = array.min()
+        if array.dtype.kind in "Mm":
+            offset = _datetime_crude_nanmin(array)
+        else:
+            offset = min(array)
     array = array - offset
 
     if not hasattr(array, "dtype"):  # scalar is converted to 0d-array
@@ -401,7 +431,8 @@ def mean(array, axis=None, skipna=None, **kwargs):
 
     array = asarray(array)
     if array.dtype.kind in "Mm":
-        offset = min(array)
+        offset = _datetime_crude_nanmin(array)
+
         # xarray always uses np.datetime64[ns] for np.datetime64 data
         dtype = "timedelta64[ns]"
         return (