python3.9

Author: xingzhongyu

.pre-commit-config.yaml

@@ -3,40 +3,40 @@ repos:
     rev: v4.6.0
     hooks:
       - id: trailing-whitespace
-        language_version: python3.11
+        language_version: python3.9
       - id: end-of-file-fixer
         exclude: ^docs/
-        language_version: python3.11
+        language_version: python3.9
 
   - repo: https://github.com/adrienverge/yamllint.git
     rev: v1.35.1
     hooks:
       - id: yamllint
         args: [-c=.yamllint.yml]
-        language_version: python3.11
+        language_version: python3.9
 
   - repo: https://github.com/asottile/pyupgrade
     rev: v3.16.0
     hooks:
       - id: pyupgrade
-        args: [--py3-plus]
-        language_version: python3.11
+        args: [--py39-plus]
+        language_version: python3.9
 
   - repo: https://github.com/google/yapf
     rev: v0.43.0
     hooks:
       - id: yapf
         name: Format code
         additional_dependencies: [toml]
-        language_version: python3.11
+        language_version: python3.9
 
   - repo: https://github.com/pycqa/isort
     rev: 5.13.2
     hooks:
       - id: isort
         name: Sort imports
         # args: [--line-width, "120", --profile, black]
-        language_version: python3.11
+        language_version: python3.9
 
   - repo: https://github.com/PyCQA/docformatter
     rev: eb1df34
@@ -45,7 +45,7 @@ repos:
         name: Format docstring
         additional_dependencies: [tomli]
         args: [--config, pyproject.toml]
-        language_version: python3.11
+        language_version: python3.9
 
   - repo: https://github.com/executablebooks/mdformat
     rev: 0.7.17
@@ -55,4 +55,4 @@ repos:
         additional_dependencies:
           - mdformat-gfm
           - mdformat-tables
-        language_version: python3.11
+        language_version: python3.9

dance/atlas/sc_similarity/anndata_similarity.py

@@ -458,9 +458,9 @@ def get_targets(dataset_truth: str):
         sim_targets.append((sum(x for x, y in sim_targets), sum(y for x, y in sim_targets)))
         return sim_targets
 
-    def compute_similarity(self, random_state: int, methods: List[str] = [
+    def compute_similarity(self, random_state: int, methods: list[str] = [
         'cosine', 'pearson', 'jaccard', 'js_distance', 'otdd', 'common_genes_num', "ground_truth", "metadata_sim"
-    ], origin=False) -> Dict[str, float]:
+    ], origin=False) -> dict[str, float]:
         """Compute multiple similarity metrics between datasets.
 
         Parameters
@@ -526,11 +526,11 @@ def compute_similarity(self, random_state: int, methods: List[str] = [
         return results
 
     def get_similarity_matrix_A2B(
-        self, methods: List[str] = [
+        self, methods: list[str] = [
             "wasserstein", "Hausdorff", "chamfer", "energy", "sinkhorn2", "bures", "spectral", "common_genes_num",
             "ground_truth", "metadata_sim", "mmd"
         ]
-    ) -> Dict[str, float]:
+    ) -> dict[str, float]:
         """Same as compute_similarity, keeping method name consistency."""
         cumulative_results = {method: 0.0 for method in methods}
 

dance/datasets/singlemodality.py

@@ -518,7 +518,7 @@ def is_complete(self):
 
         for i in check:
             if not osp.exists(i):
-                logger.info("file {} doesn't exist".format(i))
+                logger.info(f"file {i} doesn't exist")
                 return False
         return True
 

dance/datasets/spatial.py

@@ -51,7 +51,7 @@ def __init__(self, root=".", full_download=False, data_id="151673", data_dir="da
         super().__init__(root, full_download)
 
         self.data_id = data_id
-        self.data_dir = data_dir + "/{}".format(data_id)
+        self.data_dir = data_dir + f"/{data_id}"
         self.sample_file = sample_file
 
     def download_all(self):

dance/modules/multi_modality/joint_embedding/dcca.py

@@ -102,7 +102,7 @@ def build_multi_layers(layers, use_batch_norm=True, dropout_rate=0.1):
     if dropout_rate > 0:
         fc_layers = nn.Sequential(
             collections.OrderedDict([(
-                "Layer {}".format(i),
+                f"Layer {i}",
                 nn.Sequential(
                     nn.Linear(n_in, n_out),
                     nn.BatchNorm1d(n_out, momentum=0.01, eps=0.001),
@@ -113,7 +113,7 @@ def build_multi_layers(layers, use_batch_norm=True, dropout_rate=0.1):
     else:
         fc_layers = nn.Sequential(
             collections.OrderedDict([(
-                "Layer {}".format(i),
+                f"Layer {i}",
                 nn.Sequential(
                     nn.Linear(n_in, n_out),
                     nn.BatchNorm1d(n_out, momentum=0.01, eps=0.001),

dance/modules/multi_modality/joint_embedding/scmvae.py

@@ -129,7 +129,7 @@ def build_multi_layers(layers, use_batch_norm=True, dropout_rate=0.1):
     if dropout_rate > 0:
         fc_layers = nn.Sequential(
             collections.OrderedDict([(
-                "Layer {}".format(i),
+                f"Layer {i}",
                 nn.Sequential(
                     nn.Linear(n_in, n_out),
                     nn.BatchNorm1d(n_out, momentum=0.01, eps=0.001),
@@ -141,7 +141,7 @@ def build_multi_layers(layers, use_batch_norm=True, dropout_rate=0.1):
     else:
         fc_layers = nn.Sequential(
             collections.OrderedDict([(
-                "Layer {}".format(i),
+                f"Layer {i}",
                 nn.Sequential(
                     nn.Linear(n_in, n_out),
                     nn.BatchNorm1d(n_out, momentum=0.01, eps=0.001),

dance/modules/multi_modality/match_modality/cmae.py

@@ -85,7 +85,7 @@ def calc_gen_loss_half(self, input_fake):
                 all1 = Variable(torch.ones_like(out0.data).cuda(), requires_grad=False)
                 loss += torch.mean(F.binary_cross_entropy(F.sigmoid(out0), all1))
             else:
-                assert 0, "Unsupported GAN type: {}".format(self.gan_type)
+                assert 0, f"Unsupported GAN type: {self.gan_type}"
         return loss
 
 
@@ -221,7 +221,7 @@ def init_fun(m):
             elif init_type == 'default':
                 pass
             else:
-                assert 0, "Unsupported initialization: {}".format(init_type)
+                assert 0, f"Unsupported initialization: {init_type}"
             if hasattr(m, 'bias') and m.bias is not None:
                 init.constant_(m.bias.data, 0.0)
 

dance/modules/multi_modality/match_modality/scmm.py

@@ -549,15 +549,15 @@ def fit(self, x_train, y_train, val_ratio=0.15):
                 optimizer.step()
                 b_loss += loss.item()
                 if self.params.print_freq > 0 and i % self.params.print_freq == 0:
-                    print("iteration {:04d}: loss: {:6.3f}".format(i, loss.item() / self.params.batch_size))
+                    print(f"iteration {i:04d}: loss: {loss.item() / self.params.batch_size:6.3f}")
             tr.append(b_loss / len(train_loader.dataset))
-            print('====> Epoch: {:03d} Train loss: {:.4f}'.format(epoch, tr[-1]))
+            print(f'====> Epoch: {epoch:03d} Train loss: {tr[-1]:.4f}')
 
             if torch.isnan(torch.tensor([b_loss])):
                 break
 
             vals.append(self.score(train_mod1[val_idx], train_mod2[val_idx], metric='loss'))
-            print('====>             Valid loss: {:.4f}'.format(vals[-1]))
+            print(f'====>             Valid loss: {vals[-1]:.4f}')
 
             if vals[-1] == min(vals):
                 if not os.path.exists('models'):

dance/modules/multi_modality/predict_modality/babel.py

@@ -249,7 +249,7 @@ class ChromEncoder(nn.Module):
     processing everything to be the same dimensionality, concatenate to form a single
     latent dimension."""
 
-    def __init__(self, num_inputs: List[int], latent_dim: int = 32, activation=nn.PReLU):
+    def __init__(self, num_inputs: list[int], latent_dim: int = 32, activation=nn.PReLU):
         super().__init__()
         self.num_inputs = num_inputs
         self.act = activation
@@ -290,7 +290,7 @@ class ChromDecoder(nn.Module):
 
     def __init__(
         self,
-        num_outputs: List[int],  # Per-chromosome list of output sizes
+        num_outputs: list[int],  # Per-chromosome list of output sizes
         latent_dim: int = 32,
         activation=nn.PReLU,
         final_activations=[Exp(), ClippedSoftplus()],
@@ -513,7 +513,7 @@ def forward_single(self, x, size_factors=None, in_domain: int = 1, out_domain: i
         decoded = decoder(encoded)
         return self._combine_output_and_encoded(decoded, encoded, num_non_latent_out)
 
-    def forward(self, x, size_factors=None, mode: Union[None, Tuple[int, int]] = None):
+    def forward(self, x, size_factors=None, mode: Union[None, tuple[int, int]] = None):
         if self.flat_mode:
             x = self.split_catted_input(x)
         assert isinstance(x, (tuple, list))
@@ -552,11 +552,11 @@ def __init__(
         input_dim1: int,
         input_dim2: int,
         hidden_dim: int = 16,
-        final_activations1: Union[Callable, List[Callable]] = [
+        final_activations1: Union[Callable, list[Callable]] = [
             Exp(),
             ClippedSoftplus(),
         ],
-        final_activations2: Union[Callable, List[Callable]] = nn.Sigmoid(),
+        final_activations2: Union[Callable, list[Callable]] = nn.Sigmoid(),
         flat_mode: bool = True,  # Controls if we have to re-split inputs
         seed: int = 182822,
     ):
@@ -633,7 +633,7 @@ def forward_single(self, x, size_factors=None, in_domain: int = 1, out_domain: i
         assert isinstance(retval[0], (torch.TensorType, torch.Tensor))
         return retval
 
-    def forward(self, x, size_factors=None, mode: Union[None, Tuple[int, int]] = None):
+    def forward(self, x, size_factors=None, mode: Union[None, tuple[int, int]] = None):
         if self.flat_mode:
             x = self.split_catted_input(x)
         assert isinstance(x, (tuple, list))
@@ -662,7 +662,7 @@ class AssymSplicedAutoEncoder(SplicedAutoEncoder):
     def __init__(
         self,
         input_dim1: int,
-        input_dim2: List[int],
+        input_dim2: list[int],
         hidden_dim: int = 16,
         final_activations1: list = [Exp(), ClippedSoftplus()],
         final_activations2=nn.Sigmoid(),

dance/modules/multi_modality/predict_modality/cmae.py

@@ -84,7 +84,7 @@ def calc_gen_loss_half(self, input_fake):
                 all1 = Variable(torch.ones_like(out0.data).cuda(), requires_grad=False)
                 loss += torch.mean(F.binary_cross_entropy(F.sigmoid(out0), all1))
             else:
-                assert 0, "Unsupported GAN type: {}".format(self.gan_type)
+                assert 0, f"Unsupported GAN type: {self.gan_type}"
         return loss
 
 
@@ -219,7 +219,7 @@ def init_fun(m):
             elif init_type == 'default':
                 pass
             else:
-                assert 0, "Unsupported initialization: {}".format(init_type)
+                assert 0, f"Unsupported initialization: {init_type}"
             if hasattr(m, 'bias') and m.bias is not None:
                 init.constant_(m.bias.data, 0.0)