Finalize scripts for OME challenge submission

Author: constantinpape

scripts/ome_challenge/create_csv.py

@@ -0,0 +1,47 @@
+import os
+import numpy as np
+import pandas as pd
+
+ROOT = "/mnt/lustre-emmy-hdd/usr/u12086/data/flamingo/ngff-v3"
+URL_ROOT = "https://radosgw.public.os.wwu.de/n4bi-goe"
+
+
+def get_directory_size(directory):
+    total_size = 0
+    # Walk through all subdirectories and files
+    for dirpath, dirnames, filenames in os.walk(directory):
+        for filename in filenames:
+            filepath = os.path.join(dirpath, filename)
+            # Only add file size if it is a file (skip if it's a broken symlink)
+            if os.path.isfile(filepath):
+                total_size += os.path.getsize(filepath)
+
+    size_rd = np.round(total_size / 1e9, 2)
+    size_rd = f"{size_rd} GB"
+    return total_size, size_rd
+
+
+names = [
+    "Platynereis-H2B-TL.ome.zarr",
+    "Zebrafish-H2B-short-timelapse.ome.zarr",
+    "Zebrafish-XSPIM-multiview.ome.zarr",
+]
+
+urls = []
+written = []
+written_human_readable = []
+
+for name in names:
+    url = f"{URL_ROOT}/{name}"
+    urls.append(url)
+    folder = os.path.join(ROOT, name)
+    size, size_rd = get_directory_size(folder)
+    written.append(size)
+    written_human_readable.append(size_rd)
+
+df = {
+    "url": urls, "written": written, "written_human_readable": written_human_readable,
+}
+
+df = pd.DataFrame(df)
+df.to_csv("flamingo.csv", index=False)

scripts/ome_challenge/create_metadata.py

@@ -0,0 +1,60 @@
+# Create additional top-level metadata.
+
+import argparse
+import os
+import json
+from glob import glob
+
+
+def get_series(path):
+    setups = sorted(glob(os.path.join(path, "**/timepoint*")))
+    setups = [os.path.relpath(p, path) for p in setups]
+    return setups
+
+
+def create_metadata(path):
+    # TOP LEVEL METADATA
+    bf_to_raw = {
+        "attributes": {
+            "ome": {
+                "version": "0.5",
+                "bioformats2raw.layout": 3
+            }
+        },
+        "zarr_format": 3,
+        "node_type": "group",
+    }
+    meta_path = os.path.join(path, "zarr.json")
+
+    # This can be safely over-written.
+    with open(meta_path, "w") as f:
+        json.dump(bf_to_raw, f)
+
+    # OME METADATA
+    series = get_series(path)
+    ome_metadata = {
+        "attributes": {
+            "ome": {
+                "version": "0.5",
+                "series": series
+            }
+        },
+        "zarr_format": 3,
+        "node_type": "group",
+    }
+    meta_folder = os.path.join(path, "OME")
+    os.makedirs(meta_folder, exist_ok=True)
+    meta_path = os.path.join(meta_folder, "zarr.json")
+    with open(meta_path, "w") as f:
+        json.dump(ome_metadata, f)
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("path")
+    args = parser.parse_args()
+    create_metadata(args.path)
+
+
+if __name__ == "__main__":
+    main()

scripts/ome_challenge/upload_data.py

@@ -25,7 +25,10 @@ def upload_data(name):
 
 
 def main():
-    name = "Platynereis-H2B-TL.ome.zarr"
+    # name = "Platynereis-H2B-TL.ome.zarr"
+    # name = "Zebrafish-H2B-short-timelapse.ome.zarr"
+    name = "Zebrafish-XSPIM-multiview.ome.zarr"
+
     upload_data(name)