Parameterize some metric tests

tests/ignite/metrics/test_fbeta.py

@@ -12,26 +12,32 @@
 torch.manual_seed(12)
 
 
-def test_wrong_inputs():
-
-    with pytest.raises(ValueError, match=r"Beta should be a positive integer"):
-        Fbeta(0.0)
-
-    with pytest.raises(ValueError, match=r"Input precision metric should have average=False"):
-        p = Precision(average=True)
-        Fbeta(1.0, precision=p)
-
-    with pytest.raises(ValueError, match=r"Input recall metric should have average=False"):
-        r = Recall(average=True)
-        Fbeta(1.0, recall=r)
-
-    with pytest.raises(ValueError, match=r"If precision argument is provided, output_transform should be None"):
-        p = Precision(average=False)
-        Fbeta(1.0, precision=p, output_transform=lambda x: x)
-
-    with pytest.raises(ValueError, match=r"If recall argument is provided, output_transform should be None"):
-        r = Recall(average=False)
-        Fbeta(1.0, recall=r, output_transform=lambda x: x)
+@pytest.mark.parametrize(
+    "match, fbeta, precision, recall, output_transform",
+    [
+        ("Beta should be a positive integer", 0.0, None, None, None),
+        ("Input precision metric should have average=False", 1.0, Precision(average=True), None, None),
+        ("Input recall metric should have average=False", 1.0, None, Recall(average=True), None),
+        (
+            "If precision argument is provided, output_transform should be None",
+            1.0,
+            Precision(average=False),
+            None,
+            lambda x: x,
+        ),
+        (
+            "If recall argument is provided, output_transform should be None",
+            1.0,
+            None,
+            Recall(average=False),
+            lambda x: x,
+        ),
+    ],
+)
+def test_wrong_inputs(match, fbeta, precision, recall, output_transform):
+
+    with pytest.raises(ValueError, match=fr"{match}"):
+        Fbeta(fbeta, precision=precision, recall=recall, output_transform=output_transform)
 
 
 def test_integration():

tests/ignite/metrics/test_loss.py

@@ -17,22 +17,9 @@ def test_zero_div():
         loss.compute()
 
 
-def test_compute():
-    loss = Loss(nll_loss)
-
-    y_pred = torch.tensor([[0.1, 0.4, 0.5], [0.1, 0.7, 0.2]]).log()
-    y = torch.tensor([2, 2]).long()
-    loss.update((y_pred, y))
-    assert_almost_equal(loss.compute(), 1.1512925625)
-
-    y_pred = torch.tensor([[0.1, 0.3, 0.6], [0.6, 0.2, 0.2], [0.2, 0.7, 0.1]]).log()
-    y = torch.tensor([2, 0, 2]).long()
-    loss.update((y_pred, y))
-    assert_almost_equal(loss.compute(), 1.1253643036)  # average
-
-
-def test_compute_on_criterion():
-    loss = Loss(nn.NLLLoss())
+@pytest.mark.parametrize("loss_function", [nll_loss, nn.NLLLoss()])
+def test_compute(loss_function):
+    loss = Loss(loss_function)
 
     y_pred = torch.tensor([[0.1, 0.4, 0.5], [0.1, 0.7, 0.2]]).log()
     y = torch.tensor([2, 2]).long()
@@ -135,16 +122,6 @@ def _test_distrib_accumulator_device(device):
         ), f"{type(loss._sum.device)}:{loss._sum.device} vs {type(metric_device)}:{metric_device}"
 
 
-def test_sum_detached():
-    loss = Loss(nll_loss)
-
-    y_pred = torch.tensor([[0.1, 0.4, 0.5], [0.1, 0.7, 0.2]], requires_grad=True).log()
-    y = torch.tensor([2, 2]).long()
-    loss.update((y_pred, y))
-
-    assert not loss._sum.requires_grad
-
-
 @pytest.mark.distributed
 @pytest.mark.skipif(not idist.has_native_dist_support, reason="Skip if no native dist support")
 @pytest.mark.skipif(torch.cuda.device_count() < 1, reason="Skip if no GPU")