File size: 2,108 Bytes
d2efaca |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 |
from __future__ import annotations
import argparse
from pathlib import Path
import pandas as pd
def verify_split(split_dir: Path) -> int:
csv_path = split_dir / "metadata.csv"
if not csv_path.exists():
print(f"No metadata.csv in {split_dir}; skipping.")
return 0
df = pd.read_csv(csv_path)
# Normalize column access
lower = {c.lower(): c for c in df.columns}
ic = next((lower[c] for c in ("image", "file_name", "filename", "path") if c in lower), None)
latc = next((lower[c] for c in ("latitude", "lat") if c in lower), None)
lonc = next((lower[c] for c in ("longitude", "lon", "long") if c in lower), None)
if ic is None or latc is None or lonc is None:
print(f"Missing expected columns in {csv_path}: {df.columns.tolist()}")
return 0
bad = 0
for i, row in df.iterrows():
img_rel = str(row[ic])
img_path = (split_dir / img_rel).resolve()
if not img_path.exists():
print(f"[missing] {split_dir.name} row {i}: {img_rel}")
bad += 1
continue
lat = float(row[latc])
lon = float(row[lonc])
if not (-90.0 <= lat <= 90.0) or not (-180.0 <= lon <= 180.0):
print(f"[range] {split_dir.name} row {i}: lat={lat}, lon={lon}")
bad += 1
ok = len(df) - bad
print(f"Split {split_dir.name}: {ok}/{len(df)} rows valid")
return ok
def main():
ap = argparse.ArgumentParser(description="Verify dataset metadata and paths for splits")
ap.add_argument("--data-dir", type=Path, default=Path("data"))
args = ap.parse_args()
total_ok = 0
total = 0
for split in ("train", "validation", "test"):
sd = args.data_dir / split
if sd.exists():
ok = verify_split(sd)
total_ok += ok
csv_path = sd / "metadata.csv"
if csv_path.exists():
import pandas as pd # local import
total += len(pd.read_csv(csv_path))
if total:
print(f"Overall: {total_ok}/{total} rows valid")
if __name__ == "__main__":
main()
|