CT/lung2/scripts/create_synthetic_smoke.py

from __future__ import annotations

import argparse
import sys
from pathlib import Path

ROOT = Path(__file__).resolve().parents[1]
if str(ROOT) not in sys.path:
    sys.path.insert(0, str(ROOT))

import numpy as np
import pandas as pd

from src.features.radiomics import clean_radiomics_matrix
from src.utils import ensure_dir, seed_everything


def main() -> None:
    parser = argparse.ArgumentParser()
    parser.add_argument("--out-dir", default="data/processed/synthetic_smoke")
    parser.add_argument("--n", type=int, default=16)
    parser.add_argument("--seed", type=int, default=17)
    args = parser.parse_args()
    seed_everything(args.seed)
    out = ensure_dir(args.out_dir)
    features = ensure_dir("data/features")

    rng = np.random.default_rng(args.seed)
    patient_ids = [f"TCGA-SYN-{i:03d}" for i in range(args.n)]
    rad = pd.DataFrame(rng.normal(size=(args.n, 20)), columns=[f"rad_{i}" for i in range(20)])
    rad.insert(0, "patient_id", patient_ids)
    rad = clean_radiomics_matrix(rad, missing_threshold=1.0, correlation_threshold=0.99)

    dl_bbox = pd.DataFrame(rng.normal(size=(args.n, 8)), columns=[f"bbox_{i}" for i in range(8)])
    dl_bbox.insert(0, "patient_id", patient_ids)
    dl_patch = pd.DataFrame(rng.normal(size=(args.n, 8)), columns=[f"patch_{i}" for i in range(8)])
    dl_patch.insert(0, "patient_id", patient_ids)
    labels = pd.DataFrame({"patient_id": patient_ids, "label": rng.integers(0, 2, size=args.n)})

    rad.to_csv(features / "radiomics_clean.csv", index=False)
    dl_bbox.to_csv(features / "dl_bbox_medicalnet.csv", index=False)
    dl_patch.to_csv(features / "dl_patch_triregion.csv", index=False)
    labels.to_csv(features / "labels.csv", index=False)
    print(f"created synthetic smoke features in {features}")


if __name__ == "__main__":
    main()