"""E1b drift: does the clean-air baseline change from day to day, and does subtracting it help?

    python experiments/e1b_drift.py data/demo_calibration --gas ethanol --concentration 1.0

Compares, for the same standard solution measured on different days:
    raw steady value         (what you would get without a baseline)
    relative change (r-b)/b  (what features.py uses)
A smaller day-to-day CV for the relative change means the baseline correction works.
"""
from __future__ import annotations

import argparse
import sys
from pathlib import Path

import pandas as pd

sys.path.insert(0, str(Path(__file__).resolve().parents[1]))

from enose.common import output_dir, plt, save_figure  # noqa: E402
from enose.features import _tail_mean  # noqa: E402


def main() -> None:
    parser = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
    parser.add_argument("folder", type=Path)
    parser.add_argument("--gas", default="ethanol")
    parser.add_argument("--concentration", type=float, default=1.0)
    parser.add_argument("--heater-step", type=int, default=2)
    arguments = parser.parse_args()
    raw = pd.read_csv(arguments.folder / "raw.csv")
    labels = pd.read_csv(arguments.folder / "labels.csv")
    results = output_dir(arguments.folder / "results")

    chosen = labels[(labels["label"] == arguments.gas) & (labels["concentration"] == arguments.concentration)]
    if "rh_level" in chosen:
        chosen = chosen[chosen["rh_level"] == "mid"]
    records = []
    for sample in chosen.itertuples():
        rows = raw[(raw["sample_id"] == sample.sample_id) & (raw["heater_step"] == arguments.heater_step)]
        baseline = _tail_mean(rows[rows["phase"] == "purge"], "gas_ohm")
        steady = _tail_mean(rows[rows["phase"] == "sample"], "gas_ohm")
        records.append({"date": sample.date, "baseline": baseline, "steady": steady,
                        "relative_change": (steady - baseline) / baseline})
    table = pd.DataFrame(records).sort_values("date")
    daily = table.groupby("date").mean()
    daily.to_csv(results / "e1b_daily.csv")

    cv_raw = 100 * daily["steady"].std() / daily["steady"].mean()
    cv_relative = 100 * daily["relative_change"].std() / abs(daily["relative_change"].mean())
    print(daily.round(4).to_string())
    print(f"\n跨日變異係數：原始穩定值 {cv_raw:.1f}%　相對變化量 {cv_relative:.1f}%")
    print("相對變化量的 CV 比較小 → 扣基準有效" if cv_relative < cv_raw else "扣基準沒有改善，請檢查採樣流程")

    figure, axes = plt.subplots(1, 2, figsize=(10, 3.6))
    axes[0].plot(daily.index, daily["baseline"], "o-", label="乾淨空氣基準 b")
    axes[0].plot(daily.index, daily["steady"], "s-", label="樣本穩定值 r")
    axes[0].set_title("BME688 氣體電阻（Ω）")
    axes[0].legend()
    axes[1].plot(daily.index, daily["relative_change"], "o-")
    axes[1].set_title("相對變化量 (r − b) / b")
    for axis in axes:
        axis.tick_params(axis="x", rotation=45)
    save_figure(figure, results / "e1b_drift.png")


if __name__ == "__main__":
    main()