Merge branch 'master' into correlate_fft

Author: AlexanderKalistratov

.github/workflows/array-api-skips.txt

@@ -3,11 +3,6 @@
 # no 'uint8' dtype
 array_api_tests/test_array_object.py::test_getitem_masking
 
-# no 'isdtype' function
-array_api_tests/test_data_type_functions.py::test_isdtype
-array_api_tests/test_has_names.py::test_has_names[data_type-isdtype]
-array_api_tests/test_signatures.py::test_func_signature[isdtype]
-
 # missing unique-like functions
 array_api_tests/test_has_names.py::test_has_names[set-unique_all]
 array_api_tests/test_has_names.py::test_has_names[set-unique_counts]

doc/reference/dtype.rst

@@ -14,7 +14,6 @@ Data type routines
    dpnp.min_scalar_type
    dpnp.result_type
    dpnp.common_type
-   dpnp.obj2sctype
 
 Creating data types
 -------------------

dpnp/__init__.py

@@ -42,7 +42,7 @@
 # are not installed under any of default paths where Python is searching.
 from platform import system
 
-if system() == "Windows":
+if system() == "Windows":  # pragma: no cover
     if hasattr(os, "add_dll_directory"):
         os.add_dll_directory(mypath)
         os.add_dll_directory(dpctlpath)

dpnp/dpnp_iface.py

@@ -712,7 +712,7 @@ def is_cuda_backend(obj=None):
     if (
         sycl_device is not None
         and sycl_device.backend == dpctl.backend_type.cuda
-    ):
+    ):  # pragma: no cover
         return True
     return False
 

dpnp/dpnp_iface_histograms.py

@@ -91,7 +91,7 @@ def _align_dtypes(a_dtype, bins_dtype, ntype, supported_types, device):
             return sample_type, hist_type
 
     # should not happen
-    return None, None
+    return None, None  # pragma: no cover
 
 
 def _ravel_check_a_and_weights(a, weights):
@@ -392,7 +392,7 @@ def bincount(x, weights=None, minlength=None):
         x.dtype, x.dtype, ntype, supported_types, device
     )
 
-    if x_casted_dtype is None or ntype_casted is None:
+    if x_casted_dtype is None or ntype_casted is None:  # pragma: no cover
         raise ValueError(
             f"function '{bincount}' does not support input types "
             f"({x.dtype}, {ntype}), "
@@ -607,7 +607,7 @@ def histogram(a, bins=10, range=None, density=None, weights=None):
         a.dtype, bin_edges.dtype, ntype, supported_types, device
     )
 
-    if a_bin_dtype is None or hist_dtype is None:
+    if a_bin_dtype is None or hist_dtype is None:  # pragma: no cover
         raise ValueError(
             f"function '{histogram}' does not support input types "
             f"({a.dtype}, {bin_edges.dtype}, {ntype}), "

dpnp/dpnp_iface_indexing.py

@@ -138,7 +138,7 @@ def choose(x1, choices, out=None, mode="raise"):
         )
 
     if x1_desc:
-        if dpnp.is_cuda_backend(x1_desc.get_array()):
+        if dpnp.is_cuda_backend(x1_desc.get_array()):  # pragma: no cover
             raise NotImplementedError(
                 "Running on CUDA is currently not supported"
             )

dpnp/dpnp_iface_libmath.py

@@ -82,7 +82,7 @@ def erf(in_array1):
         in_array1, copy_when_strides=False, copy_when_nondefault_queue=False
     )
     if x1_desc:
-        if dpnp.is_cuda_backend(x1_desc.get_array()):
+        if dpnp.is_cuda_backend(x1_desc.get_array()):  # pragma: no cover
             raise NotImplementedError(
                 "Running on CUDA is currently not supported"
             )

dpnp/dpnp_iface_mathematical.py

@@ -2946,7 +2946,7 @@ def modf(x1, **kwargs):
 
     x1_desc = dpnp.get_dpnp_descriptor(x1, copy_when_nondefault_queue=False)
     if x1_desc:
-        if dpnp.is_cuda_backend(x1_desc.get_array()):
+        if dpnp.is_cuda_backend(x1_desc.get_array()):  # pragma: no cover
             raise NotImplementedError(
                 "Running on CUDA is currently not supported"
             )

dpnp/dpnp_iface_sorting.py

@@ -215,7 +215,7 @@ def partition(x1, kth, axis=-1, kind="introselect", order=None):
 
     x1_desc = dpnp.get_dpnp_descriptor(x1, copy_when_nondefault_queue=False)
     if x1_desc:
-        if dpnp.is_cuda_backend(x1_desc.get_array()):
+        if dpnp.is_cuda_backend(x1_desc.get_array()):  # pragma: no cover
             raise NotImplementedError(
                 "Running on CUDA is currently not supported"
             )

dpnp/dpnp_iface_statistics.py

@@ -112,7 +112,7 @@ def _count_reduce_items(arr, axis, where=True):
         for ax in axis:
             items *= arr.shape[normalize_axis_index(ax, arr.ndim)]
         items = dpnp.intp(items)
-    else:
+    else:  # pragma: no cover
         raise NotImplementedError(
             "where keyword argument is only supported with its default value."
         )

dpnp/dpnp_iface_types.py

@@ -64,6 +64,7 @@
     "integer",
     "intc",
     "intp",
+    "isdtype",
     "issubdtype",
     "is_type_supported",
     "nan",
@@ -194,11 +195,66 @@ def iinfo(dtype):
             smallest representable number.
 
     """
+
     if isinstance(dtype, dpnp_array):
         dtype = dtype.dtype
     return dpt.iinfo(dtype)
 
 
+def isdtype(dtype, kind):
+    """
+    Returns a boolean indicating whether a provided `dtype` is
+    of a specified data type `kind`.
+
+    Parameters
+    ----------
+    dtype : dtype
+        The input dtype.
+    kind : {dtype, str, tuple of dtypes or strs}
+        The input dtype or dtype kind. Allowed dtype kinds are:
+
+        * ``'bool'`` : boolean kind
+        * ``'signed integer'`` : signed integer data types
+        * ``'unsigned integer'`` : unsigned integer data types
+        * ``'integral'`` : integer data types
+        * ``'real floating'`` : real-valued floating-point data types
+        * ``'complex floating'`` : complex floating-point data types
+        * ``'numeric'`` : numeric data types
+
+    Returns
+    -------
+    out : bool
+        A boolean indicating whether a provided `dtype` is of a specified data
+        type `kind`.
+
+    See Also
+    --------
+    :obj:`dpnp.issubdtype` : Test if the first argument is a type code
+                             lower/equal in type hierarchy.
+
+    Examples
+    --------
+    >>> import dpnp as np
+    >>> np.isdtype(np.float32, np.float64)
+    False
+    >>> np.isdtype(np.float32, "real floating")
+    True
+    >>> np.isdtype(np.complex128, ("real floating", "complex floating"))
+    True
+
+    """
+
+    if isinstance(dtype, type):
+        dtype = dpt.dtype(dtype)
+
+    if isinstance(kind, type):
+        kind = dpt.dtype(kind)
+    elif isinstance(kind, tuple):
+        kind = tuple(dpt.dtype(k) if isinstance(k, type) else k for k in kind)
+
+    return dpt.isdtype(dtype, kind)
+
+
 def issubdtype(arg1, arg2):
     """
     Returns ``True`` if the first argument is a type code lower/equal

dpnp/dpnp_utils/dpnp_utils_common.py

@@ -78,4 +78,4 @@ def is_castable(dtype, stype):
             ):
                 return stypes
 
-    return None
+    return None  # pragma: no cover

dpnp/linalg/dpnp_utils_linalg.py

@@ -401,7 +401,7 @@ def _batched_qr(a, mode="reduced"):
     # w/a to avoid raice conditional on CUDA during multiple runs
     # TODO: Remove it ones the OneMath issue is resolved
     # https://github.com/uxlfoundation/oneMath/issues/626
-    if dpnp.is_cuda_backend(a_sycl_queue):
+    if dpnp.is_cuda_backend(a_sycl_queue):  # pragma: no cover
         ht_ev.wait()
     else:
         _manager.add_event_pair(ht_ev, geqrf_ev)
@@ -2479,7 +2479,7 @@ def dpnp_qr(a, mode="reduced"):
     # w/a to avoid raice conditional on CUDA during multiple runs
     # TODO: Remove it ones the OneMath issue is resolved
     # https://github.com/uxlfoundation/oneMath/issues/626
-    if dpnp.is_cuda_backend(a_sycl_queue):
+    if dpnp.is_cuda_backend(a_sycl_queue):  # pragma: no cover
         ht_ev.wait()
     else:
         _manager.add_event_pair(ht_ev, geqrf_ev)