Minor updates

Author: constantinpape

scripts/export_lower_resolution.py

@@ -13,14 +13,17 @@
 # from skimage.segmentation import relabel_sequential
 
 
-def filter_component(fs, segmentation, cochlea, seg_name, components):
+def filter_component(fs, segmentation, cochlea, seg_name, components, min_size):
     # First, we download the MoBIE table for this segmentation.
     internal_path = os.path.join(BUCKET_NAME, cochlea, "tables",  seg_name, "default.tsv")
     with fs.open(internal_path, "r") as f:
         table = pd.read_csv(f, sep="\t")
 
     # Then we get the ids for the components and us them to filter the segmentation.
     component_mask = np.isin(table.component_labels.values, components)
+    if min_size is not None:
+        component_mask = np.logical_and(component_mask, table.n_pixels > min_size)
+
     keep_label_ids = table.label_id.values[component_mask].astype("int64")
     if max(keep_label_ids) > np.iinfo("uint16").max:
         warnings.warn(f"Label ID exceeds maximum of data type 'uint16': {np.iinfo('uint16').max}.")
@@ -155,9 +158,13 @@ def export_lower_resolution(args):
             with zarr.open(s3_store, mode="r") as f:
                 data = f[input_key][:].astype("float32")
             print("Data shape", data.shape)
+
             if args.filter_by_components is not None:
                 print(f"Filtering channel {channel} by components {args.filter_by_components}.")
-                data = filter_component(fs, data, args.cochlea, channel, args.filter_by_components)
+                data = filter_component(
+                    fs, data, args.cochlea, channel, args.filter_by_components, min_size=args.min_size
+                )
+
             if args.filter_cochlea_channels is not None:
                 us_factor = ds_factor // (2 ** scale)
                 upscaled_filter = upscale_volume(data, filter_volume, upscale_factor=us_factor)
@@ -183,6 +190,7 @@ def main():
     parser.add_argument("--binarize", action="store_true")
     parser.add_argument("--filter_cochlea_channels", nargs="+", type=str, default=None)
     parser.add_argument("--filter_dilation_iterations", type=int, default=8)
+    parser.add_argument("--min_size", default=None, type=int)
     args = parser.parse_args()
 
     export_lower_resolution(args)

scripts/la-vision/predict_blocks.py

@@ -12,9 +12,11 @@
 
 
 def _get_files(sgn=True):
-    input_root = "/mnt/vast-nhr/home/pape41/u12086/Work/my_projects/flamingo-tools/scripts/more-annotations/LA_VISION_M04"  # noqa
-    input_root2 = "/mnt/vast-nhr/home/pape41/u12086/Work/my_projects/flamingo-tools/scripts/more-annotations/LA_VISION_M04_2"  # noqa
-    input_root3 = "/mnt/vast-nhr/home/pape41/u12086/Work/my_projects/flamingo-tools/scripts/more-annotations/LA_VISION_Mar05"  # noqa
+    # root = "/mnt/vast-nhr/home/pape41/u12086/Work/my_projects/flamingo-tools/scripts/more-annotations"
+    root = "/home/pape/Work/my_projects/flamingo-tools/scripts/more-annotation"
+    input_root = os.path.join(root, "LA_VISION_M04")
+    input_root2 = os.path.join(root, "LA_VISION_M04_2")
+    input_root3 = os.path.join(root, "LA_VISION_Mar05")
 
     input_root_ihc = "/mnt/vast-nhr/home/pape41/u12086/Work/my_projects/flamingo-tools/scripts/more-annotations/LA_VISION_Mar05-ihc"  # noqa
 

scripts/measurements/measure_sgns.py

@@ -34,6 +34,7 @@ def main():
     else:
         cochleae = args.cochleae
 
+    n_sgns = []
     for dataset in cochleae:
         if dataset not in project_info["datasets"]:
             print("Could not find cochleae", dataset)
@@ -60,11 +61,15 @@ def main():
                 component_ids, n_per_component = np.unique(
                     remaining_sgns, return_counts=True
                 )
-                print("Number of SGNs (largest component):", max(n_per_component))
+                n_sgn = max(n_per_component)
+                print("Number of SGNs (largest component):", n_sgn)
+                n_sgns.append(n_sgn)
             else:
                 print(source)
                 print("Number of SGNs (no postprocessing):", len(table))
 
+    print("Average number of SGNs:", np.mean(n_sgns), "+-", np.std(n_sgns))
+
 
 if __name__ == "__main__":
     main()