dask: Data.masked_invalid, Data.mask_invalid

cf/data/data.py

@@ -845,7 +845,7 @@ def __bool__(self):
                 "elements is ambiguous. Use d.any() or d.all()"
             )
 
-        return bool(self._get_dask())
+        return bool(self.to_dask_array())
 
     def __repr__(self):
         """Called by the `repr` built-in function.
@@ -5374,7 +5374,7 @@ def arctanh(self, inplace=False):
         >>> d.arctanh(inplace=True)
         >>> print(d.array)
         [nan inf 1.0986122886681098 0.6931471805599453 --]
-        >>> d.mask_invalid(inplace=True)
+        >>> d.masked_invalid(inplace=True)
         >>> print(d.array)
         [-- -- 1.0986122886681098 0.6931471805599453 --]
 
@@ -5429,7 +5429,7 @@ def arcsin(self, inplace=False):
         >>> d.arcsin(inplace=True)
         >>> print(d.array)
         [nan 1.5707963267948966 0.9272952180016123 0.6435011087932844 --]
-        >>> d.mask_invalid(inplace=True)
+        >>> d.masked_invalid(inplace=True)
         >>> print(d.array)
         [-- 1.5707963267948966 0.9272952180016123 0.6435011087932844 --]
 
@@ -5533,7 +5533,7 @@ def arccos(self, inplace=False):
         >>> d.arccos(inplace=True)
         >>> print(d.array)
         [nan 0.0 0.6435011087932843 0.9272952180016123 --]
-        >>> d.mask_invalid(inplace=True)
+        >>> d.masked_invalid(inplace=True)
         >>> print(d.array)
         [-- 0.0 0.6435011087932843 0.9272952180016123 --]
 
@@ -5588,7 +5588,7 @@ def arccosh(self, inplace=False):
         >>> d.arccosh(inplace=True)
         >>> print(d.array)
         [0.6223625037147786 0.0 nan nan --]
-        >>> d.mask_invalid(inplace=True)
+        >>> d.masked_invalid(inplace=True)
         >>> print(d.array)
         [0.6223625037147786 0.0 -- -- --]
 
@@ -5665,7 +5665,7 @@ def all(self, axis=None, keepdims=True, split_every=None):
 
         """
         d = self.copy(array=False)
-        dx = self._get_dask()
+        dx = self.to_dask_array()
         dx = da.all(dx, axis=axis, keepdims=keepdims, split_every=split_every)
         d._set_dask(dx, reset_mask_hardness=False)
         d.hardmask = _DEFAULT_HARDMASK
@@ -5782,7 +5782,7 @@ def any(self, axis=None, keepdims=True, split_every=None):
 
         """
         d = self.copy(array=False)
-        dx = self._get_dask()
+        dx = self.to_dask_array()
         dx = da.any(dx, axis=axis, keepdims=keepdims, split_every=split_every)
         d._set_dask(dx, reset_mask_hardness=False)
         d.hardmask = _DEFAULT_HARDMASK
@@ -9199,79 +9199,38 @@ def datum(self, *index):
 
         return cf_masked
 
-    @_deprecated_kwarg_check("i")
+    @daskified(_DASKIFIED_VERBOSE)
     @_inplace_enabled(default=False)
-    def mask_invalid(self, inplace=False, i=False):
+    def masked_invalid(self, inplace=False):
         """Mask the array where invalid values occur (NaN or inf).
 
-        Note that:
-
-        * Invalid values in the results of arithmetic operations may only
-          occur if the raising of `FloatingPointError` exceptions has been
-          suppressed by `cf.Data.seterr`.
-
-        * If the raising of `FloatingPointError` exceptions has been
-          allowed then invalid values in the results of arithmetic
-          operations it is possible for them to be automatically converted
-          to masked values, depending on the setting of
-          `cf.Data.mask_fpe`. In this case, such automatic conversion
-          might be faster than calling `mask_invalid`.
-
-        .. seealso:: `cf.Data.mask_fpe`, `cf.Data.seterr`
+        .. seealso:: `where`
 
         :Parameters:
 
             {{inplace: `bool`, optional}}
 
-            {{i: deprecated at version 3.0.0}}
-
         :Returns:
 
             `Data` or `None`
+                The masked data, or `None` if the operation was
+                in-place.
 
         **Examples**
 
-        >>> d = cf.Data([0., 1])
-        >>> e = cf.Data([1., 2])
-        >>> old = cf.Data.seterr('ignore')
-
-        >>> f = e/d
+        >>> d = cf.Data([0, 1, 2])
+        >>> e = cf.Data([0, 2, 0])
+        >>> f = d / e
         >>> f
-        <CF Data: [inf, 2.0] >
-        >>> f.mask_invalid()
-        <CF Data: [--, 2.0] >
-
-        >>> f=e**12345
-        >>> f
-        <CF Data: [1.0, inf] >
-        >>> f.mask_invalid()
-        <CF Data: [1.0, --] >
-
-        >>> old = cf.Data.seterr('raise')
-        >>> old = cf.Data.mask_fpe(True)
-        >>> e/d
-        <CF Data: [--, 2.0] >
-        >>> e**12345
-        <CF Data: [1.0, --] >
+        <CF Data(3): [nan, 0.5, inf]>
+        >>> f.masked_invalid()
+        <CF Data(3): [--, 0.5, --]>
 
         """
         d = _inplace_enabled_define_and_cleanup(self)
-
-        config = d.partition_configuration(readonly=False)
-
-        for partition in d.partitions.matrix.flat:
-            partition.open(config)
-            array = partition.array
-
-            array = np.ma.masked_invalid(array, copy=False)
-            array.shrink_mask()
-            if array.mask is np.ma.nomask:
-                array = array.data
-
-            partition.subarray = array
-
-            partition.close()
-
+        dx = self.to_dask_array()
+        dx = da.ma.masked_invalid(dx)
+        d._set_dask(dx, reset_mask_hardness=False)
         return d
 
     def del_calendar(self, default=ValueError()):

cf/data/mixin/deprecations.py

@@ -625,6 +625,45 @@ def mask_fpe(*arg):
             "https://github.com/dask/dask/issues/3245 for more details)."
         )
 
+    def mask_invalid(self, *args, **kwargs):
+        """Mask the array where invalid values occur (NaN or inf).
+
+        Deprecated at version TODODASK. Use the method
+        `masked_invalid` instead.
+
+        .. seealso:: `where`
+
+        :Parameters:
+
+            {{inplace: `bool`, optional}}
+
+            {{i: deprecated at version 3.0.0}}
+
+        :Returns:
+
+            `Data` or `None`
+                The masked data, or `None` if the operation was
+                in-place.
+
+        **Examples**
+
+        >>> d = cf.Data([0, 1, 2])
+        >>> e = cf.Data([0, 2, 0])
+        >>> f = d / e
+        >>> f
+        <CF Data(3): [nan, 0.5, inf]>
+        >>> f.mask_invalid()
+        <CF Data(3): [--, 0.5, --]>
+
+        """
+        _DEPRECATION_ERROR_METHOD(
+            self,
+            "mask_invalid",
+            message="Use the method 'masked_invalid' instead.",
+            version="TODODASK",
+            removed_at="5.0.0",
+        )  # pragma: no cover
+
     def partition_boundaries(self):
         """Return the partition boundaries for each partition matrix
         dimension.

cf/test/test_Data.py

@@ -3842,6 +3842,22 @@ def test_Data_tolist(self):
             self.assertEqual(e, np.array(x).tolist())
             self.assertTrue(d.equals(cf.Data(e)))
 
+    def test_Data_masked_invalid(self):
+        a = np.array([0, 1, 2])
+        b = np.array([0, 2, 0])
+        with warnings.catch_warnings():
+            warnings.simplefilter("ignore", category=RuntimeWarning)
+            i = a / b
+
+        d = cf.Data(i, "m")
+        e = d.masked_invalid().array
+        c = np.ma.masked_invalid(i)
+
+        self.assertTrue((e.mask == c.mask).all())
+        self.assertTrue((e == c).all())
+
+        self.assertIsNone(d.masked_invalid(inplace=True))
+
     def test_Data_uncompress(self):
         import cfdm