File size: 2,402 Bytes
f03ba44 | 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 | """Check a ForensicBench submission file before uploading it.
Usage:
python tools/validate_submission.py my_flags.csv --data-dir data
Expected columns: sector, document_id, scheme_type, plus replicate for the multi run format
(`rationale` is optional). Sectors: energy, healthcare, luxurygoods, manufacturing, transport.
Scheme types: fictitious_ap_disbursements, revenue_manipulation, vendor_collusion,
shadow_payroll, inventory_manipulation.
"""
import argparse, sys
import pandas as pd
SECTORS = ["energy", "healthcare", "luxurygoods", "manufacturing", "transport"]
TYPES = {"fictitious_ap_disbursements", "revenue_manipulation", "vendor_collusion", "shadow_payroll", "inventory_manipulation", "unknown"}
def main():
ap = argparse.ArgumentParser()
ap.add_argument("file")
ap.add_argument("--data-dir", default="data")
a = ap.parse_args()
df = pd.read_csv(a.file, dtype=str) if a.file.endswith(".csv") else pd.read_json(a.file, lines=True, dtype=str)
errors = []
for c in ["sector", "document_id", "scheme_type"]:
if c not in df.columns:
errors.append(f"missing column: {c}")
if errors:
print("\n".join(errors)); sys.exit(1)
bad = set(df["sector"]) - set(SECTORS)
if bad: errors.append(f"unknown sectors: {sorted(bad)}")
bad = set(df["scheme_type"]) - TYPES
if bad: errors.append(f"unknown scheme types: {sorted(bad)}")
for s in sorted(set(df["sector"]) & set(SECTORS)):
ids = set(pd.read_parquet(f"{a.data_dir}/{s}/je_header.parquet", columns=["document_id"])["document_id"].str.lower())
sub = df[df["sector"] == s]["document_id"].str.lower()
unknown = set(sub) - ids
if unknown:
print(f"warning: {s}: {len(unknown)} document_id values are not in the ledger (they count as false positives)")
if len(set(df["sector"]) & set(SECTORS)) < 5:
errors.append("a submission must cover all 5 sectors")
if "replicate" in df.columns:
reps = df.groupby("sector")["replicate"].nunique().to_dict()
print("replicates per sector:", reps)
if any(v < 5 for v in reps.values()):
print("note: the full protocol uses 5 replicates per sector (25 runs)")
else:
print("single run format (no replicate column)")
if errors:
print("\n".join(errors)); sys.exit(1)
print("OK")
if __name__ == "__main__":
main()
|