Flake8 updates

src/scip/segmentation/cellpose.py

@@ -22,6 +22,20 @@
 import torch
 
 
+def _get_gpu_device(worker):
+    if isinstance(get_client().cluster, LocalCluster):
+        gpu_id = '0'
+    else:
+        gpu_workers = [
+            address
+            for address, w in get_client().scheduler_info()["workers"].items()
+            if "cellpose" in w["resources"]
+        ]
+        gpu_id = gpu_workers.index(worker.address)
+
+    return torch.device(f'cuda:{gpu_id}')
+
+
 def segment_block(
     events: List[Mapping[str, Any]],
     *,
@@ -53,27 +67,16 @@ def segment_block(
     if len(events) == 0:
         return events
 
-    w = get_worker()
-    if hasattr(w, "cellpose"):
-        model = w.cellpose
+    worker = get_worker()
+    if hasattr(worker, "cellpose"):
+        model = worker.cellpose
     else:
         if gpu_accelerated:
-
-            if isinstance(get_client().cluster, LocalCluster):
-                gpu_id = '0'
-            else:
-                gpu_workers = [
-                    address
-                    for address, w in get_client().scheduler_info()["workers"].items()
-                    if "cellpose" in w["resources"]
-                ]
-                gpu_id = gpu_workers.index(w.address)
-
-            device = torch.device(f'cuda:{gpu_id}')
+            device = _get_gpu_device(worker)
             model = models.Cellpose(gpu=True, device=device, model_type='cyto2')
         else:
             model = models.Cellpose(gpu=False, model_type='cyto2')
-        w.cellpose = model
+        worker.cellpose = model
 
     parents, _, _, _ = model.eval(
         x=[e["pixels"][[parent_channel_index, dapi_channel_index]] for e in events],