ModuFlow/mask-ddpm
3
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Win and linux can run the code

Browse Source
This commit is contained in:
MZ YANG
2026-01-22 17:39:31 +08:00
parent c3f750cd9d
commit f37a8ce179
22 changed files with 32572 additions and 87 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
example/prepare_data.py
View File

@@ -6,28 +6,30 @@ from pathlib import Path
from typing import Optional
from data_utils import compute_cont_stats, build_vocab, load_split
from platform_utils import safe_path, ensure_dir
BASE_DIR = Path(__file__).resolve().parent
REPO_DIR = BASE_DIR.parent.parent
DATA_PATH = str(REPO_DIR / "dataset" / "hai" / "hai-21.03" / "train1.csv.gz")
SPLIT_PATH = str(BASE_DIR / "feature_split.json")
OUT_STATS = str(BASE_DIR / "results" / "cont_stats.json")
OUT_VOCAB = str(BASE_DIR / "results" / "disc_vocab.json")
DATA_PATH = REPO_DIR / "dataset" / "hai" / "hai-21.03" / "train1.csv.gz"
SPLIT_PATH = BASE_DIR / "feature_split.json"
OUT_STATS = BASE_DIR / "results" / "cont_stats.json"
OUT_VOCAB = BASE_DIR / "results" / "disc_vocab.json"
def main(max_rows: Optional[int] = None):
split = load_split(SPLIT_PATH)
split = load_split(safe_path(SPLIT_PATH))
time_col = split.get("time_column", "time")
cont_cols = [c for c in split["continuous"] if c != time_col]
disc_cols = [c for c in split["discrete"] if not c.startswith("attack") and c != time_col]
mean, std = compute_cont_stats(DATA_PATH, cont_cols, max_rows=max_rows)
vocab = build_vocab(DATA_PATH, disc_cols, max_rows=max_rows)
mean, std = compute_cont_stats(safe_path(DATA_PATH), cont_cols, max_rows=max_rows)
vocab = build_vocab(safe_path(DATA_PATH), disc_cols, max_rows=max_rows)
with open(OUT_STATS, "w", encoding="ascii") as f:
ensure_dir(OUT_STATS.parent)
with open(safe_path(OUT_STATS), "w", encoding="utf-8") as f:
json.dump({"mean": mean, "std": std, "max_rows": max_rows}, f, indent=2)
with open(OUT_VOCAB, "w", encoding="ascii") as f:
with open(safe_path(OUT_VOCAB), "w", encoding="utf-8") as f:
json.dump({"vocab": vocab, "max_rows": max_rows}, f, indent=2)