Euler: use float64 for test

naver · Apr 23, 2024 · 0768dfd · 0768dfd
1 parent 8f6a2ed
commit 0768dfd
Showing 1 changed file with 6 additions and 3 deletions.
diff --git a/test/test_euler.py b/test/test_euler.py
@@ -20,6 +20,7 @@ def test_euler(self):
 
     def test_euler_unitquat_consistency(self):
         device = torch.device(0) if torch.cuda.is_available() else torch.device('cpu')
+        dtype = torch.float64
         for degrees in (True, False):
             for batch_shape in [tuple(),
                                 torch.Size((30,)),
@@ -28,7 +29,7 @@ def test_euler_unitquat_consistency(self):
                     for convention in ["".join(permutation) for permutation in itertools.permutations('xyz')] + ["xyx", "xzx", "yxy", "yzy", "zxz", "zyz"]:
                         if intrinsics:
                             convention = convention.upper()
-                        q = roma.random_unitquat(batch_shape, device=device)
+                        q = roma.random_unitquat(batch_shape, device=device, dtype=dtype)
                         angles = roma.unitquat_to_euler(convention, q, degrees=degrees)
                         self.assertTrue(len(angles) == 3)
                         self.assertTrue(all([angle.shape == batch_shape for angle in angles]))
@@ -63,6 +64,7 @@ def test_euler_rotvec_consistency(self):
 
     def test_euler_rotmat_consistency(self):
         device = torch.device(0) if torch.cuda.is_available() else torch.device('cpu')
+        dtype = torch.float64
         for degrees in (True, False):
             for batch_shape in [tuple(),
                                 torch.Size((30,)),
@@ -71,7 +73,7 @@ def test_euler_rotmat_consistency(self):
                     for convention in ["".join(permutation) for permutation in itertools.permutations('xyz')] + ["xyx", "xzx", "yxy", "yzy", "zxz", "zyz"]:
                         if intrinsics:
                             convention = convention.upper()
-                        q = roma.random_rotmat(batch_shape, device=device)
+                        q = roma.random_rotmat(batch_shape, device=device, dtype=dtype)
                         angles = roma.rotmat_to_euler(convention, q, degrees=degrees)
                         self.assertTrue(len(angles) == 3)
                         self.assertTrue(all([angle.shape == batch_shape for angle in angles]))
@@ -96,7 +98,8 @@ def test_euler_tensor(self):
         Test that Euler conversion methods support both list and tensor inputs.
         """
         batch_shape = (10,34)
-        q = roma.random_unitquat(batch_shape, device=device)
+        dtype = torch.float64
+        q = roma.random_unitquat(batch_shape, device=device, dtype=dtype)
         convention = 'xyz'
         angles = roma.unitquat_to_euler(convention, q)
         angles_tensor = roma.unitquat_to_euler(convention, q, as_tensor=True)