from app.services.data_service import DataService


def test_prediction_basic_holdout_mae() -> None:
    ds = DataService()
    ds.load_data()
    assert ds.df is not None

    df = ds.df.dropna(
        subset=[
            "weave",
            "blend",
            "Greige EPI",
            "Greige PPI",
            "FINISH EPI",
            "FINISH PPI",
        ]
    ).copy()

    sample = df.sample(n=min(120, len(df)), random_state=42)

    abs_epi_errors = []
    abs_ppi_errors = []
    valid = 0

    for _, row in sample.iterrows():
        payload = {
            "weave": row["weave"],
            "blend": row["blend"],
            "warp_count": float(row["warp_count"])
            if row["warp_count"] == row["warp_count"]
            else None,
            "weft_count": float(row["weft_count"])
            if row["weft_count"] == row["weft_count"]
            else None,
            "finish_epi": float(row["FINISH EPI"]),
            "finish_ppi": float(row["FINISH PPI"]),
            "target_gsm": float(row["FINISH GSM"])
            if row["FINISH GSM"] == row["FINISH GSM"]
            else None,
        }
        out = ds.predict_construction(payload)
        rec = out.get("recommendation", {})

        if rec.get("greige_epi") is None or rec.get("greige_ppi") is None:
            continue

        valid += 1
        abs_epi_errors.append(abs(float(rec["greige_epi"]) - float(row["Greige EPI"])))
        abs_ppi_errors.append(abs(float(rec["greige_ppi"]) - float(row["Greige PPI"])))

    assert valid >= 80

    mae_epi = sum(abs_epi_errors) / len(abs_epi_errors)
    mae_ppi = sum(abs_ppi_errors) / len(abs_ppi_errors)

    assert mae_epi < 45
    assert mae_ppi < 20


def test_validation_report_endpoint_logic() -> None:
    ds = DataService()
    ds.load_data()
    report = ds.get_validation_report(sample_size=120, seed=9)
    assert "mae" in report
    assert report["scored_rows"] >= 80
    assert report["mae"]["greige_epi"] < 35
    assert report["mae"]["greige_ppi"] < 15