Merge pull request #1719 from IntelPython/support-scalars-in-where

Add Python scalar support to `dpt.where`

Author: ndgrigorian

dpctl/tensor/_search_functions.py

@@ -17,11 +17,81 @@
 import dpctl
 import dpctl.tensor as dpt
 import dpctl.tensor._tensor_impl as ti
-from dpctl.tensor._manipulation_functions import _broadcast_shapes
+from dpctl.tensor._elementwise_common import (
+    _get_dtype,
+    _get_queue_usm_type,
+    _get_shape,
+    _validate_dtype,
+)
+from dpctl.tensor._manipulation_functions import _broadcast_shape_impl
 from dpctl.utils import ExecutionPlacementError, SequentialOrderManager
 
 from ._copy_utils import _empty_like_orderK, _empty_like_triple_orderK
-from ._type_utils import _all_data_types, _can_cast
+from ._type_utils import (
+    WeakBooleanType,
+    WeakComplexType,
+    WeakFloatingType,
+    WeakIntegralType,
+    _all_data_types,
+    _can_cast,
+    _is_weak_dtype,
+    _strong_dtype_num_kind,
+    _to_device_supported_dtype,
+    _weak_type_num_kind,
+)
+
+
+def _default_dtype_from_weak_type(dt, dev):
+    if isinstance(dt, WeakBooleanType):
+        return dpt.bool
+    if isinstance(dt, WeakIntegralType):
+        return dpt.dtype(ti.default_device_int_type(dev))
+    if isinstance(dt, WeakFloatingType):
+        return dpt.dtype(ti.default_device_fp_type(dev))
+    if isinstance(dt, WeakComplexType):
+        return dpt.dtype(ti.default_device_complex_type(dev))
+
+
+def _resolve_two_weak_types(o1_dtype, o2_dtype, dev):
+    "Resolves two weak data types per NEP-0050"
+    if _is_weak_dtype(o1_dtype):
+        if _is_weak_dtype(o2_dtype):
+            return _default_dtype_from_weak_type(
+                o1_dtype, dev
+            ), _default_dtype_from_weak_type(o2_dtype, dev)
+        o1_kind_num = _weak_type_num_kind(o1_dtype)
+        o2_kind_num = _strong_dtype_num_kind(o2_dtype)
+        if o1_kind_num > o2_kind_num:
+            if isinstance(o1_dtype, WeakIntegralType):
+                return dpt.dtype(ti.default_device_int_type(dev)), o2_dtype
+            if isinstance(o1_dtype, WeakComplexType):
+                if o2_dtype is dpt.float16 or o2_dtype is dpt.float32:
+                    return dpt.complex64, o2_dtype
+                return (
+                    _to_device_supported_dtype(dpt.complex128, dev),
+                    o2_dtype,
+                )
+            return _to_device_supported_dtype(dpt.float64, dev), o2_dtype
+        else:
+            return o2_dtype, o2_dtype
+    elif _is_weak_dtype(o2_dtype):
+        o1_kind_num = _strong_dtype_num_kind(o1_dtype)
+        o2_kind_num = _weak_type_num_kind(o2_dtype)
+        if o2_kind_num > o1_kind_num:
+            if isinstance(o2_dtype, WeakIntegralType):
+                return o1_dtype, dpt.dtype(ti.default_device_int_type(dev))
+            if isinstance(o2_dtype, WeakComplexType):
+                if o1_dtype is dpt.float16 or o1_dtype is dpt.float32:
+                    return o1_dtype, dpt.complex64
+                return o1_dtype, _to_device_supported_dtype(dpt.complex128, dev)
+            return (
+                o1_dtype,
+                _to_device_supported_dtype(dpt.float64, dev),
+            )
+        else:
+            return o1_dtype, o1_dtype
+    else:
+        return o1_dtype, o2_dtype
 
 
 def _where_result_type(dt1, dt2, dev):
@@ -51,16 +121,17 @@ def where(condition, x1, x2, /, *, order="K", out=None):
             and otherwise yields from ``x2``.
             Must be compatible with ``x1`` and ``x2`` according
             to broadcasting rules.
-        x1 (usm_ndarray): Array from which values are chosen when
-            ``condition`` is ``True``.
+        x1 (Union[usm_ndarray, bool, int, float, complex]):
+            Array from which values are chosen when ``condition`` is ``True``.
             Must be compatible with ``condition`` and ``x2`` according
             to broadcasting rules.
-        x2 (usm_ndarray): Array from which values are chosen when
-            ``condition`` is not ``True``.
+        x2 (Union[usm_ndarray, bool, int, float, complex]):
+            Array from which values are chosen when ``condition`` is not
+            ``True``.
             Must be compatible with ``condition`` and ``x2`` according
             to broadcasting rules.
         order (``"K"``, ``"C"``, ``"F"``, ``"A"``, optional):
-            Memory layout of the new output arra,
+            Memory layout of the new output array,
             if parameter ``out`` is ``None``.
             Default: ``"K"``.
         out (Optional[usm_ndarray]):
@@ -81,36 +152,90 @@ def where(condition, x1, x2, /, *, order="K", out=None):
         raise TypeError(
             "Expecting dpctl.tensor.usm_ndarray type, " f"got {type(condition)}"
         )
-    if not isinstance(x1, dpt.usm_ndarray):
-        raise TypeError(
-            "Expecting dpctl.tensor.usm_ndarray type, " f"got {type(x1)}"
+    if order not in ["K", "C", "F", "A"]:
+        order = "K"
+    q1, condition_usm_type = condition.sycl_queue, condition.usm_type
+    q2, x1_usm_type = _get_queue_usm_type(x1)
+    q3, x2_usm_type = _get_queue_usm_type(x2)
+    if q2 is None and q3 is None:
+        exec_q = q1
+        out_usm_type = condition_usm_type
+    elif q3 is None:
+        exec_q = dpctl.utils.get_execution_queue((q1, q2))
+        if exec_q is None:
+            raise ExecutionPlacementError(
+                "Execution placement can not be unambiguously inferred "
+                "from input arguments."
+            )
+        out_usm_type = dpctl.utils.get_coerced_usm_type(
+            (
+                condition_usm_type,
+                x1_usm_type,
+            )
         )
-    if not isinstance(x2, dpt.usm_ndarray):
+    elif q2 is None:
+        exec_q = dpctl.utils.get_execution_queue((q1, q3))
+        if exec_q is None:
+            raise ExecutionPlacementError(
+                "Execution placement can not be unambiguously inferred "
+                "from input arguments."
+            )
+        out_usm_type = dpctl.utils.get_coerced_usm_type(
+            (
+                condition_usm_type,
+                x2_usm_type,
+            )
+        )
+    else:
+        exec_q = dpctl.utils.get_execution_queue((q1, q2, q3))
+        if exec_q is None:
+            raise ExecutionPlacementError(
+                "Execution placement can not be unambiguously inferred "
+                "from input arguments."
+            )
+        out_usm_type = dpctl.utils.get_coerced_usm_type(
+            (
+                condition_usm_type,
+                x1_usm_type,
+                x2_usm_type,
+            )
+        )
+    dpctl.utils.validate_usm_type(out_usm_type, allow_none=False)
+    condition_shape = condition.shape
+    x1_shape = _get_shape(x1)
+    x2_shape = _get_shape(x2)
+    if not all(
+        isinstance(s, (tuple, list))
+        for s in (
+            x1_shape,
+            x2_shape,
+        )
+    ):
         raise TypeError(
-            "Expecting dpctl.tensor.usm_ndarray type, " f"got {type(x2)}"
+            "Shape of arguments can not be inferred. "
+            "Arguments are expected to be "
+            "lists, tuples, or both"
         )
-    if order not in ["K", "C", "F", "A"]:
-        order = "K"
-    exec_q = dpctl.utils.get_execution_queue(
-        (
-            condition.sycl_queue,
-            x1.sycl_queue,
-            x2.sycl_queue,
+    try:
+        res_shape = _broadcast_shape_impl(
+            [
+                condition_shape,
+                x1_shape,
+                x2_shape,
+            ]
         )
-    )
-    if exec_q is None:
-        raise dpctl.utils.ExecutionPlacementError
-    out_usm_type = dpctl.utils.get_coerced_usm_type(
-        (
-            condition.usm_type,
-            x1.usm_type,
-            x2.usm_type,
+    except ValueError:
+        raise ValueError(
+            "operands could not be broadcast together with shapes "
+            f"{condition_shape}, {x1_shape}, and {x2_shape}"
         )
-    )
-
-    x1_dtype = x1.dtype
-    x2_dtype = x2.dtype
-    out_dtype = _where_result_type(x1_dtype, x2_dtype, exec_q.sycl_device)
+    sycl_dev = exec_q.sycl_device
+    x1_dtype = _get_dtype(x1, sycl_dev)
+    x2_dtype = _get_dtype(x2, sycl_dev)
+    if not all(_validate_dtype(o) for o in (x1_dtype, x2_dtype)):
+        raise ValueError("Operands have unsupported data types")
+    x1_dtype, x2_dtype = _resolve_two_weak_types(x1_dtype, x2_dtype, sycl_dev)
+    out_dtype = _where_result_type(x1_dtype, x2_dtype, sycl_dev)
     if out_dtype is None:
         raise TypeError(
             "function 'where' does not support input "
@@ -119,8 +244,6 @@ def where(condition, x1, x2, /, *, order="K", out=None):
             "to any supported types according to the casting rule ''safe''."
         )
 
-    res_shape = _broadcast_shapes(condition, x1, x2)
-
     orig_out = out
     if out is not None:
         if not isinstance(out, dpt.usm_ndarray):
@@ -149,16 +272,25 @@ def where(condition, x1, x2, /, *, order="K", out=None):
                 "Input and output allocation queues are not compatible"
             )
 
-        if ti._array_overlap(condition, out):
-            if not ti._same_logical_tensors(condition, out):
-                out = dpt.empty_like(out)
+        if ti._array_overlap(condition, out) and not ti._same_logical_tensors(
+            condition, out
+        ):
+            out = dpt.empty_like(out)
 
-        if ti._array_overlap(x1, out):
-            if not ti._same_logical_tensors(x1, out):
+        if isinstance(x1, dpt.usm_ndarray):
+            if (
+                ti._array_overlap(x1, out)
+                and not ti._same_logical_tensors(x1, out)
+                and x1_dtype == out_dtype
+            ):
                 out = dpt.empty_like(out)
 
-        if ti._array_overlap(x2, out):
-            if not ti._same_logical_tensors(x2, out):
+        if isinstance(x2, dpt.usm_ndarray):
+            if (
+                ti._array_overlap(x2, out)
+                and not ti._same_logical_tensors(x2, out)
+                and x2_dtype == out_dtype
+            ):
                 out = dpt.empty_like(out)
 
     if order == "A":
@@ -174,6 +306,10 @@ def where(condition, x1, x2, /, *, order="K", out=None):
             )
             else "C"
         )
+    if not isinstance(x1, dpt.usm_ndarray):
+        x1 = dpt.asarray(x1, dtype=x1_dtype, sycl_queue=exec_q)
+    if not isinstance(x2, dpt.usm_ndarray):
+        x2 = dpt.asarray(x2, dtype=x2_dtype, sycl_queue=exec_q)
 
     if condition.size == 0:
         if out is not None:
@@ -236,9 +372,12 @@ def where(condition, x1, x2, /, *, order="K", out=None):
                 sycl_queue=exec_q,
             )
 
-    condition = dpt.broadcast_to(condition, res_shape)
-    x1 = dpt.broadcast_to(x1, res_shape)
-    x2 = dpt.broadcast_to(x2, res_shape)
+    if condition_shape != res_shape:
+        condition = dpt.broadcast_to(condition, res_shape)
+    if x1_shape != res_shape:
+        x1 = dpt.broadcast_to(x1, res_shape)
+    if x2_shape != res_shape:
+        x2 = dpt.broadcast_to(x2, res_shape)
 
     dep_evs = _manager.submitted_events
     hev, where_ev = ti._where(

dpctl/tests/test_usm_ndarray_search_functions.py

@@ -14,6 +14,9 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+import ctypes
+import itertools
+
 import numpy as np
 import pytest
 from helper import get_queue_or_skip, skip_if_dtype_not_supported
@@ -350,9 +353,9 @@ def test_where_arg_validation():
 
     with pytest.raises(TypeError):
         dpt.where(check, x1, x2)
-    with pytest.raises(TypeError):
+    with pytest.raises(ValueError):
         dpt.where(x1, check, x2)
-    with pytest.raises(TypeError):
+    with pytest.raises(ValueError):
         dpt.where(x1, x2, check)
 
 
@@ -522,3 +525,54 @@ def test_where_out_arg_validation():
         dpt.where(condition, x1, x2, out=out_wrong_shape)
     with pytest.raises(ValueError):
         dpt.where(condition, x1, x2, out=out_not_writable)
+
+
+@pytest.mark.parametrize("arr_dt", _all_dtypes)
+def test_where_python_scalar(arr_dt):
+    q = get_queue_or_skip()
+    skip_if_dtype_not_supported(arr_dt, q)
+
+    n1, n2 = 10, 10
+    condition = dpt.tile(
+        dpt.reshape(
+            dpt.asarray([True, False], dtype="?", sycl_queue=q), (1, 2)
+        ),
+        (n1, n2 // 2),
+    )
+    x = dpt.zeros((n1, n2), dtype=arr_dt, sycl_queue=q)
+    py_scalars = (
+        bool(0),
+        int(0),
+        float(0),
+        complex(0),
+        np.float32(0),
+        ctypes.c_int(0),
+    )
+    for sc in py_scalars:
+        r = dpt.where(condition, x, sc)
+        assert isinstance(r, dpt.usm_ndarray)
+        r = dpt.where(condition, sc, x)
+        assert isinstance(r, dpt.usm_ndarray)
+
+
+def test_where_two_python_scalars():
+    get_queue_or_skip()
+
+    n1, n2 = 10, 10
+    condition = dpt.tile(
+        dpt.reshape(dpt.asarray([True, False], dtype="?"), (1, 2)),
+        (n1, n2 // 2),
+    )
+
+    py_scalars = [
+        bool(0),
+        int(0),
+        float(0),
+        complex(0),
+        np.float32(0),
+        ctypes.c_int(0),
+    ]
+
+    for sc1, sc2 in itertools.product(py_scalars, repeat=2):
+        r = dpt.where(condition, sc1, sc2)
+        assert isinstance(r, dpt.usm_ndarray)