kalpesh77/Multiverse-field-coordinates

1 2"""3Generate one layer + one dimension.4Usage example (inside Colab or local):5    python generate_single.py --layer 1 --dim 86"""7import argparse8import numpy as np9import pandas as pd10from pathlib import Path11from utils.orthogonal_bases import generate_orthogonal_matrix12from utils.hypersphere import point_on_hypersphere13 14def generate_coords(n_dim, layer=1, ids=range(1, 109), angles=np.arange(0, 361)):15    radius = 0.2 + (layer - 1) * 0.1516    rows = []17    for id_num in ids:18        Q = generate_orthogonal_matrix(n_dim, id_num)19        row = {"ID": id_num}20        for angle_deg in angles:21            theta = np.deg2rad(angle_deg)22            base = point_on_hypersphere(n_dim, theta)23            final = (Q @ base) * radius24            row[str(angle_deg)] = ",".join(f"{v:.6f}" for v in final)25        rows.append(row)26    df = pd.DataFrame(rows)27    return df[["ID"] + [str(a) for a in angles]]28 29if __name__ == "__main__":30    parser = argparse.ArgumentParser()31    parser.add_argument("--layer", type=int, default=1)32    parser.add_argument("--dim", type=int, default=3)33    parser.add_argument("--out_dir", type=str, default="data")34    args = parser.parse_args()35 36    df = generate_coords(args.dim, layer=args.layer)37    out = Path(args.out_dir) / f"layer_{args.layer:02d}" / f"coords_{args.dim}d.csv"38    out.parent.mkdir(parents=True, exist_ok=True)39    df.to_csv(out, index=False)40    print(f"Saved: {out} | shape={df.shape}")41