"""Phase 3 — calibration of the classical baseline: sweep a few behavioural parameters over seeds and
report the stylized facts (tracking error, return autocorrelation, kurtosis, volatility clustering,
spread/depth, bubbles). Writes results/calibration/sweep.parquet and a summary CSV. No API calls."""
from __future__ import annotations

import itertools
import sys
from concurrent.futures import ProcessPoolExecutor
from pathlib import Path

import pandas as pd

sys.path.insert(0, str(Path(__file__).resolve().parents[1] / "src"))
from market_agents.experiments.config import RunConfig  # noqa: E402
from market_agents.experiments.runner import run_config  # noqa: E402
from market_agents.metrics.outcomes import run_outcomes  # noqa: E402

ROOT = Path(__file__).resolve().parents[1]
GRID = {
    "fund_aggr": [10.0, 20.0, 40.0],          # fundamentalist lots per 1 % mispricing
    "trend_share": [0.10, 0.15, 0.25],        # trend+leveraged share (rebalanced against noise)
    "noise_pm": [0.3, 0.6],                   # noise traders' market-order probability
    "mm_qty": [10, 20],                       # market-maker quote size
}
SEEDS = list(range(1, 11))


def make_cfg(fund_aggr, trend_share, noise_pm, mm_qty, seed):
    tr = trend_share * 0.6
    lev = trend_share * 0.4
    noise = 1.0 - (0.25 + 0.10 + 0.05 + tr + lev)
    mix = {"fundamentalist": 0.25, "trend": round(tr, 4), "meanrev": 0.10, "noise": round(noise, 4), "market_maker": 0.05,
           "statarb": 0.0, "leveraged": round(lev, 4)}
    mix["noise"] = round(1.0 - sum(v for k, v in mix.items() if k != "noise"), 6)
    return RunConfig.from_yaml(ROOT / "configs/baseline.yaml", **{
        "name": f"cal_fa{fund_aggr}_ts{trend_share}_pm{noise_pm}_mm{mm_qty}", "seed": seed, "n_ticks": 600,
        "population.mix": mix,
        "population.policy_kwargs": {"fundamentalist": {"aggressiveness": fund_aggr}, "noise": {"p_market": noise_pm},
                                     "market_maker": {"quote_qty": mm_qty}},
        "treatment": {"fund_aggr": fund_aggr, "trend_share": trend_share, "noise_pm": noise_pm, "mm_qty": mm_qty}})


def one(args):
    cfg = make_cfg(*args)
    return run_outcomes(run_config(cfg))


if __name__ == "__main__":
    jobs = [(*g, s) for g in itertools.product(*GRID.values()) for s in SEEDS]
    print(f"{len(jobs)} runs")
    with ProcessPoolExecutor(max_workers=8) as ex:
        rows = list(ex.map(one, jobs, chunksize=4))
    df = pd.DataFrame(rows)
    out = ROOT / "results/calibration"
    out.mkdir(parents=True, exist_ok=True)
    df.to_parquet(out / "sweep.parquet", index=False)
    keys = ["tr_fund_aggr", "tr_trend_share", "tr_noise_pm", "tr_mm_qty"]
    cols = ["mad_pct", "acf_r_1", "vr5", "kurtosis", "acf_absr_1_10", "vol_bp", "spread_bp", "depth5", "lambda_bp_100",
            "n_bubbles", "max_dd", "time_to_absorb", "response_ratio", "volume", "pnl_market_maker"]
    summ = df.groupby(keys)[cols].mean().round(3)
    summ.to_csv(out / "sweep_summary.csv")
    print(summ.to_string())