update elementwise tests for exp, real, imag, and conj

dpctl/tensor/_elementwise_funcs.py

@@ -36,8 +36,11 @@
 Returns:
     usm_narray:
         An array containing the element-wise absolute values.
-        For complex input, the absolute value is its magnitude. The data type
-        of the returned array is determined by the Type Promotion Rules.
+        For complex input, the absolute value is its magnitude.
+        If `x` has a real-valued data type, the returned array has the
+        same data type as `x`. If `x` has a complex floating-point data type,
+        the returned array has a real-valued floating-point data type whose
+        precision matches the precision of `x`.
 """
 
 abs = UnaryElementwiseFunc("abs", ti._abs_result_type, ti._abs, _abs_docstring_)
@@ -133,8 +136,8 @@
         Default: "K".
 Returns:
     usm_narray:
-        An array containing the element-wise conjugate values. The data type
-        of the returned array is determined by the Type Promotion Rules.
+        An array containing the element-wise conjugate values.
+        The returned array has the same data type as `x`.
 """
 
 conj = UnaryElementwiseFunc(
@@ -216,8 +219,7 @@
 Returns:
     usm_narray:
         An array containing the result of element-wise equality comparison.
-        The data type of the returned array is determined by the
-        Type Promotion Rules.
+        The returned array has a data type of `bool`.
 """
 
 equal = BinaryElementwiseFunc(
@@ -264,6 +266,8 @@
 Return:
     usm_ndarray:
         An array containing the element-wise exp(x)-1 values.
+        The data type of the returned array is determined by the Type
+        Promotion Rules.
 """
 
 expm1 = UnaryElementwiseFunc(
@@ -288,7 +292,7 @@
         Second input array, also expected to have numeric data type.
 Returns:
     usm_narray:
-        an array containing the result of element-wise  floor division.
+        an array containing the result of element-wise floor division.
         The data type of the returned array is determined by the Type
         Promotion Rules.
 """
@@ -319,8 +323,7 @@
 Returns:
     usm_narray:
         An array containing the result of element-wise greater-than comparison.
-        The data type of the returned array is determined by the
-        Type Promotion Rules.
+        The returned array has a data type of `bool`.
 """
 
 greater = BinaryElementwiseFunc(
@@ -347,8 +350,7 @@
     usm_narray:
         An array containing the result of element-wise greater-than or equal-to
         comparison.
-        The data type of the returned array is determined by the
-        Type Promotion Rules.
+        The returned array has a data type of `bool`.
 """
 
 greater_equal = BinaryElementwiseFunc(
@@ -376,8 +378,10 @@
 Returns:
     usm_narray:
         An array containing the element-wise imaginary component of input.
-        The data type of the returned array is determined
-        by the Type Promotion Rules.
+        If the input is a real-valued data type, the returned array has
+        the same datat type. If the input is a complex floating-point
+        data type, the returned array has a floating-point data type
+        with the same floating-point precision as complex input.
 """
 
 imag = UnaryElementwiseFunc(
@@ -403,7 +407,7 @@
     usm_narray:
         An array which is True where `x` is not positive infinity,
         negative infinity, or NaN, False otherwise.
-        The data type of the returned array is boolean.
+        The data type of the returned array is `bool`.
 """
 
 isfinite = UnaryElementwiseFunc(
@@ -428,7 +432,7 @@
 Returns:
     usm_narray:
         An array which is True where `x` is positive or negative infinity,
-        False otherwise. The data type of the returned array is boolean.
+        False otherwise. The data type of the returned array is `bool`.
 """
 
 isinf = UnaryElementwiseFunc(
@@ -453,7 +457,7 @@
 Returns:
     usm_narray:
         An array which is True where x is NaN, False otherwise.
-        The data type of the returned array is boolean.
+        The data type of the returned array is `bool`.
 """
 
 isnan = UnaryElementwiseFunc(
@@ -481,8 +485,7 @@
 Returns:
     usm_narray:
         An array containing the result of element-wise less-than comparison.
-        The data type of the returned array is determined by the
-        Type Promotion Rules.
+        The returned array has a data type of `bool`.
 """
 
 less = BinaryElementwiseFunc(
@@ -508,9 +511,7 @@
 Returns:
     usm_narray:
         An array containing the result of element-wise less-than or equal-to
-        comparison.
-        The data type of the returned array is determined by the
-        Type Promotion Rules.
+        comparison. The returned array has a data type of `bool`.
 """
 
 less_equal = BinaryElementwiseFunc(
@@ -536,6 +537,8 @@
 Return:
     usm_ndarray:
         An array containing the element-wise natural logarithm values.
+        The data type of the returned array is determined by the Type
+        Promotion Rules.
 """
 
 log = UnaryElementwiseFunc("log", ti._log_result_type, ti._log, _log_docstring)
@@ -555,7 +558,8 @@
         Default: "K".
 Return:
     usm_ndarray:
-        An array containing the element-wise log(1+x) values.
+        An array containing the element-wise log(1+x) values. The data type
+        of the returned array is determined by the Type Promotion Rules.
 """
 
 log1p = UnaryElementwiseFunc(
@@ -804,8 +808,7 @@
 Returns:
     usm_narray:
         an array containing the result of element-wise inequality comparison.
-        The data type of the returned array is determined by the
-        Type Promotion Rules.
+        The returned array has a data type of `bool`.
 """
 
 not_equal = BinaryElementwiseFunc(
@@ -873,8 +876,8 @@
         Default: "K".
 Returns:
     usm_narray:
-        An array containing the element-wise projection. The data
-        type of the returned array is determined by the Type Promotion Rules.
+        An array containing the element-wise projection.
+        The returned array has the same data type as `x`.
 """
 
 proj = UnaryElementwiseFunc(
@@ -898,8 +901,11 @@
         Default: "K".
 Returns:
     usm_narray:
-        An array containing the element-wise real component of input. The data
-        type of the returned array is determined by the Type Promotion Rules.
+        An array containing the element-wise real component of input.
+        If the input is a real-valued data type, the returned array has
+        the same datat type. If the input is a complex floating-point
+        data type, the returned array has a floating-point data type
+        with the same floating-point precision as complex input.
 """
 
 real = UnaryElementwiseFunc(

dpctl/tensor/libtensor/include/kernels/elementwise_functions/exp.hpp

@@ -64,7 +64,58 @@ template <typename argT, typename resT> struct ExpFunctor
 
     resT operator()(const argT &in)
     {
-        return std::exp(in);
+        if constexpr (is_complex<argT>::value) {
+            using realT = typename argT::value_type;
+
+            constexpr realT q_nan = std::numeric_limits<realT>::quiet_NaN();
+
+            const realT x = std::real(in);
+            const realT y = std::imag(in);
+            if (std::isfinite(x)) {
+                if (std::isfinite(y)) {
+                    return std::exp(in);
+                }
+                else {
+                    return resT{q_nan, q_nan};
+                }
+            }
+            else if (std::isnan(x)) {
+                /* x is nan */
+                if (y == realT(0)) {
+                    return resT{in};
+                }
+                else {
+                    return resT{x, q_nan};
+                }
+            }
+            else {
+                if (!std::signbit(x)) { /* x is +inf */
+                    if (y == realT(0)) {
+                        return resT{x, y};
+                    }
+                    else if (std::isfinite(y)) {
+                        return resT{x * std::cos(y), x * std::sin(y)};
+                    }
+                    else {
+                        /* x = +inf, y = +-inf || nan */
+                        return resT{x, q_nan};
+                    }
+                }
+                else { /* x is -inf */
+                    if (std::isfinite(y)) {
+                        realT exp_x = std::exp(x);
+                        return resT{exp_x * std::cos(y), exp_x * std::sin(y)};
+                    }
+                    else {
+                        /* x = -inf, y = +-inf || nan */
+                        return resT{0, 0};
+                    }
+                }
+            }
+        }
+        else {
+            return std::exp(in);
+        }
     }
 };
 

dpctl/tests/elementwise/test_complex.py

@@ -114,11 +114,11 @@ def test_complex_order(np_call, dpt_call, dtype):
     X[..., 0::2] = np.pi / 6 + 1j * np.pi / 3
     X[..., 1::2] = np.pi / 3 + 1j * np.pi / 6
 
-    for ord in ["C", "F", "A", "K"]:
-        for perms in itertools.permutations(range(4)):
-            U = dpt.permute_dims(X[:, ::-1, ::-1, :], perms)
+    for perms in itertools.permutations(range(4)):
+        U = dpt.permute_dims(X[:, ::-1, ::-1, :], perms)
+        expected_Y = np_call(dpt.asnumpy(U))
+        for ord in ["C", "F", "A", "K"]:
             Y = dpt_call(U, order=ord)
-            expected_Y = np_call(dpt.asnumpy(U))
             assert np.allclose(dpt.asnumpy(Y), expected_Y)
 
 
@@ -164,35 +164,51 @@ def test_projection(dtype):
     "np_call, dpt_call",
     [(np.real, dpt.real), (np.imag, dpt.imag), (np.conj, dpt.conj)],
 )
-@pytest.mark.parametrize("dtype", ["f4", "f8"])
-@pytest.mark.parametrize("stride", [-1, 1, 2, 4, 5])
-def test_complex_strided(np_call, dpt_call, dtype, stride):
+@pytest.mark.parametrize("dtype", ["c8", "c16"])
+def test_complex_strided(np_call, dpt_call, dtype):
     q = get_queue_or_skip()
     skip_if_dtype_not_supported(dtype, q)
 
-    N = 100
-    rng = np.random.default_rng(42)
-    x1 = rng.standard_normal(N, dtype)
-    x2 = 1j * rng.standard_normal(N, dtype)
-    x = x1 + x2
-    y = np_call(x[::stride])
-    z = dpt_call(dpt.asarray(x[::stride]))
+    np.random.seed(42)
+    strides = np.array([-4, -3, -2, -1, 1, 2, 3, 4])
+    sizes = [2, 4, 6, 8, 9, 24, 72]
+    tol = 8 * dpt.finfo(dtype).resolution
 
-    tol = 8 * dpt.finfo(y.dtype).resolution
-    assert_allclose(y, dpt.asnumpy(z), atol=tol, rtol=tol)
+    low = -1000.0
+    high = 1000.0
+    for ii in sizes:
+        x1 = np.random.uniform(low=low, high=high, size=ii)
+        x2 = np.random.uniform(low=low, high=high, size=ii)
+        Xnp = np.array([complex(v1, v2) for v1, v2 in zip(x1, x2)], dtype=dtype)
+        X = dpt.asarray(Xnp)
+        Ynp = np_call(Xnp)
+        for jj in strides:
+            assert_allclose(
+                dpt.asnumpy(dpt_call(X[::jj])),
+                Ynp[::jj],
+                atol=tol,
+                rtol=tol,
+            )
 
 
-@pytest.mark.parametrize("dtype", ["f2", "f4", "f8"])
+@pytest.mark.parametrize("dtype", ["c8", "c16"])
 def test_complex_special_cases(dtype):
     q = get_queue_or_skip()
     skip_if_dtype_not_supported(dtype, q)
 
-    x = [np.nan, -np.nan, np.inf, -np.inf]
-    with np.errstate(all="ignore"):
-        Xnp = 1j * np.array(x, dtype=dtype)
-    X = dpt.asarray(Xnp, dtype=Xnp.dtype)
+    x = [np.nan, -np.nan, np.inf, -np.inf, +0.0, -0.0]
+    xc = [complex(*val) for val in itertools.product(x, repeat=2)]
+
+    Xc_np = np.array(xc, dtype=dtype)
+    Xc = dpt.asarray(Xc_np, dtype=dtype, sycl_queue=q)
 
     tol = 8 * dpt.finfo(dtype).resolution
-    assert_allclose(dpt.asnumpy(dpt.real(X)), np.real(Xnp), atol=tol, rtol=tol)
-    assert_allclose(dpt.asnumpy(dpt.imag(X)), np.imag(Xnp), atol=tol, rtol=tol)
-    assert_allclose(dpt.asnumpy(dpt.conj(X)), np.conj(Xnp), atol=tol, rtol=tol)
+    assert_allclose(
+        dpt.asnumpy(dpt.real(Xc)), np.real(Xc_np), atol=tol, rtol=tol
+    )
+    assert_allclose(
+        dpt.asnumpy(dpt.imag(Xc)), np.imag(Xc_np), atol=tol, rtol=tol
+    )
+    assert_allclose(
+        dpt.asnumpy(dpt.conj(Xc)), np.conj(Xc_np), atol=tol, rtol=tol
+    )