points2vols test fix

bottler · facebook-github-bot · commit feb5d36394b2 · 2022-02-21T07:24:21.000-08:00
Summary: Fix tests which depended on output tensors being identical to input ones, which now fail in main PyTorch branch because of some change in autograd. The functions still work in-place.

Reviewed By: patricklabatut

Differential Revision: D34375817

fbshipit-source-id: 295ae195f75eab6c7abab412c997470d8de8add1
diff --git a/tests/test_points_to_volumes.py b/tests/test_points_to_volumes.py
@@ -567,8 +567,8 @@ def single_point(self, device, splat: bool, align_corners: bool):
             splat,
         )
 
-        self.assertIs(volume_densities, volume_densities_)
-        self.assertIs(volume_features, volume_features_)
+        self.assertTrue(volume_densities_.is_set_to(volume_densities))
+        self.assertTrue(volume_features_.is_set_to(volume_features))
 
         if align_corners:
             volume_densities_expected[0, 0, 2, 3, 4] = point_weight

Original file line number	Diff line number	Diff line change
`@@ -567,8 +567,8 @@ def single_point(self, device, splat: bool, align_corners: bool):`
`567`	`567`	`splat,`
`568`	`568`	`)`
`569`	`569`
`570`		`- self.assertIs(volume_densities, volume_densities_)`
`571`		`- self.assertIs(volume_features, volume_features_)`
	`570`	`+ self.assertTrue(volume_densities_.is_set_to(volume_densities))`
	`571`	`+ self.assertTrue(volume_features_.is_set_to(volume_features))`
`572`	`572`
`573`	`573`	`if align_corners:`
`574`	`574`	`volume_densities_expected[0, 0, 2, 3, 4] = point_weight`