sciv 0.0.105__py3-none-any.whl → 0.0.107__py3-none-any.whl

sciv/preprocessing/_scvi_.py

@@ -13,7 +13,7 @@ from torch.cuda import OutOfMemoryError
 
 from .. import util as ul
 from ..tool import umap, tsne
-from ..util import path
+from ..util import path, check_gpu_availability
 
 __name__: str = "preprocessing_scvi"
 
@@ -77,53 +77,70 @@ def poisson_vi(
         scvi.external.POISSONVI.setup_anndata(adata, layer="fragments", batch_key=batch_key)
         _model_ = scvi.external.POISSONVI(adata)
 
-        try:
-            data_splitter_kwargs = {"drop_dataset_tail": True, "drop_last": False}
-            with warnings.catch_warnings():
-                warnings.simplefilter("ignore")
-                _model_.train(
-                    max_epochs=int(max_epochs),
-                    check_val_every_n_epoch=1,
-                    accelerator="gpu",
-                    devices=-1,
-                    datasplitter_kwargs=data_splitter_kwargs,
-                    strategy="ddp_notebook_find_unused_parameters_true",
-                    lr=lr,
-                    batch_size=int(batch_size),
-                    eps=eps,
-                    early_stopping=early_stopping,
-                    early_stopping_patience=int(early_stopping_patience)
-                )
-        except Exception as ex:
+        if check_gpu_availability():
 
             try:
-                ul.log(__name__).warning(f"Multiple GPU failed to run, attempting to run on one card.\n {ex}")
-                with warnings.catch_warnings():
-                    warnings.simplefilter("ignore")
-                    _model_.train(
-                        max_epochs=int(max_epochs),
-                        check_val_every_n_epoch=1,
-                        lr=lr,
-                        batch_size=int(batch_size),
-                        eps=eps,
-                        early_stopping=early_stopping,
-                        early_stopping_patience=int(early_stopping_patience)
-                    )
-            except Exception as exc:
-                ul.log(__name__).warning(f"GPU failed to run, try to switch to CPU running.\n {exc}")
+                data_splitter_kwargs = {"drop_dataset_tail": True, "drop_last": False}
                 with warnings.catch_warnings():
                     warnings.simplefilter("ignore")
-                    _model_.to_device('cpu')
                     _model_.train(
-                        max_epochs=int(max_epochs),
+                        max_epochs=max_epochs,
                         check_val_every_n_epoch=1,
+                        accelerator="gpu",
+                        devices=-1,
+                        datasplitter_kwargs=data_splitter_kwargs,
+                        strategy="ddp_notebook_find_unused_parameters_true",
                         lr=lr,
-                        batch_size=int(batch_size),
+                        batch_size=batch_size,
                         eps=eps,
                         early_stopping=early_stopping,
-                        early_stopping_patience=int(early_stopping_patience),
-                        accelerator="cpu"
+                        early_stopping_patience=early_stopping_patience
                     )
+            except Exception as ex:
+
+                try:
+                    ul.log(__name__).warning(f"Multiple GPU failed to run, attempting to run on one card.\n {ex}")
+                    with warnings.catch_warnings():
+                        warnings.simplefilter("ignore")
+                        _model_.train(
+                            max_epochs=max_epochs,
+                            check_val_every_n_epoch=1,
+                            lr=lr,
+                            batch_size=batch_size,
+                            eps=eps,
+                            early_stopping=early_stopping,
+                            early_stopping_patience=early_stopping_patience
+                        )
+                except Exception as exc:
+                    ul.log(__name__).warning(f"GPU failed to run, try to switch to CPU running.\n {exc}")
+                    with warnings.catch_warnings():
+                        warnings.simplefilter("ignore")
+                        _model_.to_device('cpu')
+                        _model_.train(
+                            max_epochs=max_epochs,
+                            check_val_every_n_epoch=1,
+                            lr=lr,
+                            batch_size=batch_size,
+                            eps=eps,
+                            early_stopping=early_stopping,
+                            early_stopping_patience=early_stopping_patience,
+                            accelerator="cpu"
+                        )
+        else:
+
+            with warnings.catch_warnings():
+                warnings.simplefilter("ignore")
+                _model_.to_device('cpu')
+                _model_.train(
+                    max_epochs=max_epochs,
+                    check_val_every_n_epoch=1,
+                    lr=lr,
+                    batch_size=batch_size,
+                    eps=eps,
+                    early_stopping=early_stopping,
+                    early_stopping_patience=early_stopping_patience,
+                    accelerator="cpu"
+                )
 
         return _model_
 

sciv/tool/_algorithm_.py

@@ -982,12 +982,16 @@ def overlap_sum(regions: AnnData, variants: dict, trait_info: DataFrame, n_jobs:
             # Collect non-zero values
             if matrix_sum.size == 1:
                 val = float(matrix_sum)
+
                 if val != 0:
                     local_row_indices.append(row_idx)
                     local_col_indices.append(col_idx)
                     local_data_vals.append(val)
+
             else:
+
                 for t_idx, v in enumerate(matrix_sum):
+
                     if v != 0:
                         local_row_indices.append(row_idx)
                         local_col_indices.append(col_idx + t_idx)

sciv/tool/_random_walk_.py

@@ -86,7 +86,7 @@ class RandomWalkModel(nn.Module):
         self.p = p
         self.pbar = pbar
 
-        is_gpu_available = check_gpu_availability(verbose=False)
+        is_gpu_available = check_gpu_availability()
 
         self.device = 'cuda' if (device == 'gpu' or (device == 'auto' and is_gpu_available)) else 'cpu'
 
@@ -197,7 +197,7 @@ def random_walk(
     device: str = 'auto'
 ) -> matrix_data:
 
-    availability = check_gpu_availability(False)
+    availability = check_gpu_availability()
 
     if device == 'cpu' or (device == 'auto' and not availability):
         sample_count = seed_cell_weight.shape[1]
@@ -339,7 +339,7 @@ class RandomWalk:
         self.benchmark_count = benchmark_count
         self._enrichment_seed_cell_min_count_ = 3
 
-        self.is_gpu_available = check_gpu_availability(False)
+        self.is_gpu_available = check_gpu_availability()
 
         if not is_simple and self.is_ablation:
             if "cell_mutual_knn" not in cc_adata.layers:

sciv/util/_core_.py

@@ -466,7 +466,7 @@ def add_cluster_info(data: DataFrame, data_ref: DataFrame, cluster: str) -> Data
     return new_data
 
 
-def check_gpu_availability(verbose: bool = True) -> bool:
+def check_gpu_availability(verbose: bool = False) -> bool:
     available = torch.cuda.is_available()
 
     if verbose:

{sciv-0.0.105.dist-info → sciv-0.0.107.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sciv
-Version: 0.0.105
+Version: 0.0.107
 Summary: Unveiling the pivotal cell types involved in variant function regulation at a single-cell resolution
 Project-URL: github, https://github.com/YuZhengM/sciv
 Author-email: Zheng-Min Yu <yuzmbio@163.com>

{sciv-0.0.105.dist-info → sciv-0.0.107.dist-info}/RECORD

@@ -24,16 +24,16 @@ sciv/preprocessing/_anndata_.py,sha256=3d1cHFs1YA9UZkIPE089nZHFi-DjK-c1fRyi2shfS
 sciv/preprocessing/_gencode_.py,sha256=HKhRgK76khGepdv4FaKiOvTys1uJTbvIyrKUta5W0K8,2108
 sciv/preprocessing/_gsea_.py,sha256=AH_PpUNfMN7WkF0pqAuUhEC6qZwKrtQm6VaaYu6JLfI,3803
 sciv/preprocessing/_scanpy_.py,sha256=95kt6ylfqKmjLr8GZ1XH1khP9Q3uBYy4vqc4_2Zf1J4,11629
-sciv/preprocessing/_scvi_.py,sha256=vtMzaAuy3LZXPmg92MNJfFw5jIZv71_tZdpeEyl3JHc,10782
+sciv/preprocessing/_scvi_.py,sha256=RfphdljTzA0sNuae44rKuP8YCOv5Cl23e6txCIXK_mI,11544
 sciv/preprocessing/_snapatac_.py,sha256=Dq8CHF7Psl3CQszaEokQYO56Oe2uzyWOy_cGlaOywfc,27798
 sciv/tool/__init__.py,sha256=WgyiWuCvBadTZfNEB70VurbN1YfpNodnVCIcAABTYEA,2512
-sciv/tool/_algorithm_.py,sha256=5H_LHGyRfx0ng9wm8pux0_dEXuJJJVdnGLM9MjD12Fo,53682
+sciv/tool/_algorithm_.py,sha256=pQuCI7Jxdnd6KxenFNIx6iVHR2_UK04oyDleP5aOuPU,53690
 sciv/tool/_matrix_.py,sha256=SnC3sXic_ufuEXStcD_HncvYH6apBdNK6nhG6jFLmjA,24324
-sciv/tool/_random_walk_.py,sha256=2DPBRtd4A2MadfYDhkTw33WOATriiCULozNsUPE6lsY,48912
+sciv/tool/_random_walk_.py,sha256=M73RmnP-CZAiIxMl5YfkP6wkH2xQ10dCkjfwsJTKrBQ,48889
 sciv/util/__init__.py,sha256=nER-hRRaBGl3CGgl2rLZAUOQ3uY6-m9wQ9VNrjO5EFE,2014
 sciv/util/_constant_.py,sha256=w0wKQd8guLd1ZTW24_5aECrWsIWDiNQmEpLsWlHar1A,3000
-sciv/util/_core_.py,sha256=3sWLYFdGln-HQzM3BY4iYHjVjuiZaR4P4jOvZlXXrAY,16766
-sciv-0.0.105.dist-info/METADATA,sha256=I_-ayhIMfZyJxrBvwk9zLtYR5OsXXYLvWCJjJ3luBDU,3495
-sciv-0.0.105.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
-sciv-0.0.105.dist-info/licenses/LICENSE,sha256=4UvHVf3qCOZjHLs4LkYz8u96XRpXnZrpTKrkUQPs5_A,1075
-sciv-0.0.105.dist-info/RECORD,,
+sciv/util/_core_.py,sha256=00v5MrSqZxLnJvbfijD7DuR7innW_R0J6CAwivAsNNs,16767
+sciv-0.0.107.dist-info/METADATA,sha256=COiUTIK3Dd14FAZBJWZDwnDQPQ0JSsXZ5I_3k9x2fC8,3495
+sciv-0.0.107.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
+sciv-0.0.107.dist-info/licenses/LICENSE,sha256=4UvHVf3qCOZjHLs4LkYz8u96XRpXnZrpTKrkUQPs5_A,1075
+sciv-0.0.107.dist-info/RECORD,,