[add] test

cupyx/scipy/linalg/__init__.py

@@ -1 +1 @@
-from cupyx.scipy.linalg.solve import solve_triangular  # NOQA
+from cupyx.scipy.linalg.solve_triangular import solve_triangular  # NOQA

tests/cupyx_tests/scipy_tests/linalg_tests/test_solve_triangular.py

@@ -0,0 +1,56 @@
+import unittest
+
+import cupy
+from cupy import testing
+from cupy.testing import condition
+import cupyx.scipy.linalg
+
+import numpy
+
+try:
+    import scipy.linalg
+    _scipy_available = True
+except ImportError:
+    _scipy_available = False
+
+
+@testing.parameterize(*testing.product({
+    'trans': [0, 1, 2, 'N', 'T', 'C'],
+    'lower': [True, False],
+    'unit_diagonal': [True, False],
+}))
+@testing.gpu
+@testing.with_requires('scipy')
+class TestSolveTriangular(unittest.TestCase):
+
+    @testing.for_float_dtypes(no_float16=True)
+    @condition.retry(10)
+    def check_x(self, a_shape, b_shape, dtype):
+        a_cpu = numpy.random.randint(0, 10, size=a_shape).astype(dtype)
+        b_cpu = numpy.random.randint(0, 10, size=b_shape).astype(dtype)
+        a_cpu = numpy.tril(a_cpu)
+
+        if self.lower is False:
+            a_cpu = a_cpu.T
+        if self.unit_diagonal is True:
+            numpy.fill_diagonal(a_cpu, 1)
+
+        a_gpu = cupy.asarray(a_cpu)
+        b_gpu = cupy.asarray(b_cpu)
+        a_gpu_copy = a_gpu.copy()
+        b_gpu_copy = b_gpu.copy()
+        result_cpu = scipy.linalg.solve_triangular(
+            a_cpu, b_cpu, trans=self.trans, lower=self.lower,
+            unit_diagonal=self.unit_diagonal)
+        result_gpu = cupyx.scipy.linalg.solve_triangular(
+            a_gpu, b_gpu, trans=self.trans, lower=self.lower,
+            unit_diagonal=self.unit_diagonal)
+        self.assertEqual(result_cpu.dtype, result_gpu.dtype)
+        cupy.testing.assert_allclose(result_cpu, result_gpu, atol=1e-3)
+        cupy.testing.assert_array_equal(a_gpu_copy, a_gpu)
+        cupy.testing.assert_array_equal(b_gpu_copy, b_gpu)
+
+    def test_solve(self):
+        self.check_x((4, 4), (4,))
+        self.check_x((5, 5), (5, 2))
+        self.check_x((5, 5), (5, 5))