analyze_strategy_debug.py

import argparse

import pandas as pd


def analyze(path: str) -> pd.DataFrame:
    df = pd.read_csv(path, parse_dates=["date"]).set_index("date")

    print("Span:", df.index.min().date(), "to", df.index.max().date())
    print("Avg deployed:", df["deployed_p"].mean())
    print("Pct days >80% deployed:", (df["deployed_p"] > 0.8).mean())
    print("Pct days 0% deployed:", (df["deployed_p"] < 1e-9).mean())

    # Buys above T>1.3
    viol_buy = (df["deployed_p"].diff() > 1e-3) & (df["temp"] > 1.3)
    print("buys_above_1.3:", int(viol_buy.sum()))
    if viol_buy.any():
        print(df.loc[viol_buy, ["temp", "deployed_p", "target_p", "base_p"]].head())

    # Sells when sell-block was active
    viol_sell = (df["deployed_p"].diff() < -1e-3) & (df["block_sell"] == 1)
    print("sells_when_blocked:", int(viol_sell.sum()))
    if viol_sell.any():
        print(
            df.loc[
                viol_sell,
                ["temp", "deployed_p", "target_p", "base_p", "ret_3", "ret_6", "ret_12", "ret_22"],
            ].head()
        )

    print("first_nonzero_deploy_date:", df.index[df["deployed_p"] > 1e-9].min())
    print("forced_derisk_events:", int(df["forced_derisk"].sum()))
    if (df["forced_derisk"] == 1).any():
        print("first_forced_derisk_dates:", df.index[df["forced_derisk"] == 1][:5].tolist())

    return df


def main():
    parser = argparse.ArgumentParser(description="Inspect strategy reserve debug CSV")
    parser.add_argument(
        "--csv",
        default="strategy_tqqq_reserve_debug.csv",
        help="Path to the debug CSV produced by strategy_tqqq_reserve.py",
    )
    args = parser.parse_args()
    analyze(args.csv)


if __name__ == "__main__":
    main()