"""数据清洗脚本：修正原始数据中的问题"""

from pathlib import Path

import numpy as np
import pandas as pd
import typer
from loguru import logger

from lnp_ml.config import RAW_DATA_DIR, INTERIM_DATA_DIR


app = typer.Typer()


@app.command()
def main(
    input_path: Path = RAW_DATA_DIR / "internal.xlsx",
    output_path: Path = INTERIM_DATA_DIR / "internal.csv",
):
    """
    预处理内部数据，按给药途径分组进行 z-score 标准化，对 size 列取 log。
    
    修正内容：
    1. 按给药途径分组进行 z-score 标准化
    2. 对 size 列取 log
    """
    logger.info(f"Loading data from {input_path}")
    df = pd.read_excel(input_path, header=2)
    logger.info(f"Loaded {len(df)} samples")

    # 分别对肌肉注射组和静脉注射组重新进行 z-score 标准化
    logger.info("Z-score normalizing delivery by Route_of_administration...")
    df["unnormalized_delivery"] = pd.to_numeric(df["unnormalized_delivery"], errors="coerce")
    df["quantified_delivery"] = (
        df.groupby("Route_of_administration")["unnormalized_delivery"]
          .transform(lambda x: (x - x.mean()) / x.std())
    )

    # 对 size 列取 log
    logger.info("Log-transforming size column...")
    df["size"] = pd.to_numeric(df["size"], errors="coerce")
    df["size"] = np.log(df["size"].replace(0, np.nan))  # 避免 log(0)

    # 保存
    output_path.parent.mkdir(parents=True, exist_ok=True)
    df.to_csv(output_path, index=False)
    logger.success(f"Saved cleaned data to {output_path}")


if __name__ == "__main__":
    app()