Simplify with list comprehension

Signed-off-by: R. Garcia-Dias <[email protected]>

Author: garciadias

tests/apps/detection/networks/test_retinanet.py

@@ -90,13 +90,8 @@
 CASE_LIST = [TEST_CASE_1, TEST_CASE_2, TEST_CASE_3, TEST_CASE_2_A, TEST_CASE_3_A]
 MODEL_LIST = [resnet10, resnet18, resnet34, resnet50, resnet101, resnet152, resnet200]
 
-TEST_CASES = []
-for params in dict_product(model=MODEL_LIST, case=CASE_LIST):
-    TEST_CASES.append([params["model"], *params["case"]])
-
-TEST_CASES_TS = []
-for params in dict_product(model=MODEL_LIST, case=[TEST_CASE_1]):
-    TEST_CASES_TS.append([params["model"], *params["case"]])
+TEST_CASES = [[params["model"], *params["case"]] for params in dict_product(model=MODEL_LIST, case=CASE_LIST)]
+TEST_CASES_TS = [[params["model"], *params["case"]] for params in dict_product(model=MODEL_LIST, case=[TEST_CASE_1])]
 
 
 @SkipIfBeforePyTorchVersion((1, 12))

tests/data/meta_tensor/test_meta_tensor.py

@@ -37,9 +37,8 @@
 DTYPES = [[torch.float32], [torch.float64], [torch.float16], [torch.int64], [torch.int32], [None]]
 
 # Replace nested loops with dict_product
-TESTS = []
-for params in dict_product(device=TEST_DEVICES, dtype=DTYPES):
-    TESTS.append((*params["device"], *params["dtype"]))  # type: ignore
+
+TESTS = [(*params["device"], *params["dtype"]) for params in dict_product(device=TEST_DEVICES, dtype=DTYPES)]
 
 
 def rand_string(min_len=5, max_len=10):

tests/networks/blocks/test_patchembedding.py

@@ -26,36 +26,39 @@
 einops, has_einops = optional_import("einops")
 
 
-TEST_CASE_PATCHEMBEDDINGBLOCK = dict_product(
-    dropout_rate=[0.5],
-    in_channels=[1, 4],
-    hidden_size=[96, 288],
-    img_size=[32, 64],
-    patch_size=[8, 16],
-    num_heads=[8, 12],
-    proj_type=["conv", "perceptron"],
-    pos_embed_type=["none", "learnable", "sincos"],
-    spatial_dims=[2, 3],
-)
 TEST_CASE_PATCHEMBEDDINGBLOCK = [
     [
         params,
         (2, params["in_channels"], *([params["img_size"]] * params["spatial_dims"])),
         (2, (params["img_size"] // params["patch_size"]) ** params["spatial_dims"], params["hidden_size"]),
     ]
-    for params in TEST_CASE_PATCHEMBEDDINGBLOCK
+    for params in dict_product(
+        dropout_rate=[0.5],
+        in_channels=[1, 4],
+        hidden_size=[96, 288],
+        img_size=[32, 64],
+        patch_size=[8, 16],
+        num_heads=[8, 12],
+        proj_type=["conv", "perceptron"],
+        pos_embed_type=["none", "learnable", "sincos"],
+        spatial_dims=[2, 3],
+    )
 ]
 
-TEST_CASE_PATCHEMBED = dict_product(
-    patch_size=[2], in_chans=[1, 4], img_size=[96], embed_dim=[6, 12], norm_layer=[nn.LayerNorm], spatial_dims=[2, 3]
-)
 TEST_CASE_PATCHEMBED = [
     [
         params,
         (2, params["in_chans"], *([params["img_size"]] * params["spatial_dims"])),
         (2, (params["img_size"] // params["patch_size"]) ** params["spatial_dims"], params["embed_dim"]),
     ]
-    for params in TEST_CASE_PATCHEMBED
+    for params in dict_product(
+        patch_size=[2],
+        in_chans=[1, 4],
+        img_size=[96],
+        embed_dim=[6, 12],
+        norm_layer=[nn.LayerNorm],
+        spatial_dims=[2, 3],
+    )
 ]
 
 

tests/networks/blocks/test_selfattention.py

@@ -26,31 +26,22 @@
 
 einops, has_einops = optional_import("einops")
 
-TEST_CASE_SABLOCK = []
-for params in dict_product(
-    dropout_rate=np.linspace(0, 1, 4),
-    hidden_size=[360, 480, 600, 768],
-    num_heads=[4, 6, 8, 12],
-    rel_pos_embedding=[None, RelPosEmbedding.DECOMPOSED],
-    input_size=[(16, 32), (8, 8, 8)],
-    include_fc=[True, False],
-    use_combined_linear=[True, False],
-):
-    test_case = [
-        {
-            "hidden_size": params["hidden_size"],
-            "num_heads": params["num_heads"],
-            "dropout_rate": params["dropout_rate"],
-            "rel_pos_embedding": params["rel_pos_embedding"],
-            "input_size": params["input_size"],
-            "include_fc": params["include_fc"],
-            "use_combined_linear": params["use_combined_linear"],
-            "use_flash_attention": True if params["rel_pos_embedding"] is None else False,
-        },
+TEST_CASE_SABLOCK = [
+    [
+        params,
         (2, 512, params["hidden_size"]),
         (2, 512, params["hidden_size"]),
     ]
-    TEST_CASE_SABLOCK.append(test_case)
+    for params in dict_product(
+        dropout_rate=np.linspace(0, 1, 4),
+        hidden_size=[360, 480, 600, 768],
+        num_heads=[4, 6, 8, 12],
+        rel_pos_embedding=[None, RelPosEmbedding.DECOMPOSED],
+        input_size=[(16, 32), (8, 8, 8)],
+        include_fc=[True, False],
+        use_combined_linear=[True, False],
+    )
+]
 
 
 class TestResBlock(unittest.TestCase):