Feature/optimize cp features

src/ops_model/data/data_loader.py

@@ -451,9 +451,17 @@ def __getitem__(self, index):
         nuc_mask = np.expand_dims(nuc_mask, axis=0)
         sc_mask = cell_mask == ci.segmentation_id
 
+        # cell_seg and nuclear_seg zarr arrays can differ by a few pixels at FOV
+        # boundaries due to independent segmentation pipelines. Clip to the smaller
+        # extent so downstream mask multiplication doesn't broadcast-error.
+        min_h = min(nuc_mask.shape[1], sc_mask.shape[1])
+        min_w = min(nuc_mask.shape[2], sc_mask.shape[2])
+        nuc_mask = nuc_mask[:, :min_h, :min_w]
+        sc_mask = sc_mask[:, :min_h, :min_w]
+
         # ensure that all output masks as binary
         if self.cell_masks:
-            data = data * sc_mask
+            data = data[:, :min_h, :min_w]
             nuc_mask = (nuc_mask * sc_mask) > 0
 
         cyto_mask = sc_mask & (nuc_mask == 0)

src/ops_model/data/paths.py

@@ -3,6 +3,7 @@
 
 """
 
+import os
 from pathlib import Path
 
 
@@ -14,11 +15,22 @@ def __init__(self, experiment: str, well: str = None):
         else:
             self.well_prefix = None
 
-        self.base = Path("/hpc/projects/icd.fast.ops")
+        # Allow override of base directory via environment variable
+        self.base = Path(
+            os.environ.get(
+                "OPS_OUTPUT_BASE_DIR",
+                "/hpc/projects/icd.fast.ops",
+            )
+        )
+
+        # Allow override of fast_ops base directory (defaults to base)
+        fast_base = Path(
+            os.environ.get("OPS_FAST_OUTPUT_BASE_DIR", str(self.base))
+        )
 
         self.stores = {
             "phenotyping": self.base / self.experiment / "3-assembly/phenotyping.zarr",
-            "phenotyping_v3": Path("/hpc/projects/icd.fast.ops")
+            "phenotyping_v3": fast_base
             / self.experiment
             / "3-assembly/phenotyping_v3.zarr",
         }