import os
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt


# ============================================================
# 設定
# ============================================================

INPUT_FILE = "results/trade_base.csv"
OUTPUT_DIR = "results"

os.makedirs(OUTPUT_DIR, exist_ok=True)

D_LOWER = 0.60
D_UPPER = 1.00

C_LOWER = 0.50
C_UPPER = 1.00


# ============================================================
# 出力ファイル
# ============================================================

PERIOD_CSV = os.path.join(
    OUTPUT_DIR,
    "d_time_decay_period_analysis.csv"
)

YEAR_CSV = os.path.join(
    OUTPUT_DIR,
    "d_time_decay_year_analysis.csv"
)

COMPARE_CSV = os.path.join(
    OUTPUT_DIR,
    "relative_band_year_comparison.csv"
)

PF_GRAPH = os.path.join(
    OUTPUT_DIR,
    "yearly_pf_comparison.png"
)

TRADES_GRAPH = os.path.join(
    OUTPUT_DIR,
    "yearly_trade_count.png"
)

WIN_RATE_GRAPH = os.path.join(
    OUTPUT_DIR,
    "yearly_win_rate.png"
)

AVG_PROFIT_GRAPH = os.path.join(
    OUTPUT_DIR,
    "yearly_avg_profit.png"
)

AVG_WIN_LOSS_GRAPH = os.path.join(
    OUTPUT_DIR,
    "yearly_avg_win_loss.png"
)


# ============================================================
# データ読み込み
# ============================================================

df = pd.read_csv(INPUT_FILE)

print("=" * 60)
print("RELATIVE BAND DISTANCE - D TIME DECAY ANALYSIS")
print("=" * 60)

print(f"\n読み込み件数: {len(df)}")


# ============================================================
# 必要列確認
# ============================================================

required_columns = [
    "entry_time",
    "entry_band_distance_pips",
    "entry_atr14_pips",
    "profit",
]

missing_columns = [
    column
    for column in required_columns
    if column not in df.columns
]

if missing_columns:
    raise ValueError(
        "必要な列がありません: "
        + ", ".join(missing_columns)
    )


# ============================================================
# 型変換
# ============================================================

df["entry_time"] = pd.to_datetime(
    df["entry_time"],
    errors="coerce"
)

numeric_columns = [
    "entry_band_distance_pips",
    "entry_atr14_pips",
    "profit",
]

for column in numeric_columns:

    df[column] = pd.to_numeric(
        df[column],
        errors="coerce"
    )


# ============================================================
# 分析対象データ
# ============================================================

analysis_df = df[
    df["entry_time"].notna()
    &
    df["entry_band_distance_pips"].notna()
    &
    df["entry_atr14_pips"].notna()
    &
    df["profit"].notna()
    &
    (df["entry_atr14_pips"] > 0)
].copy()


analysis_df["relative_band_distance"] = (
    analysis_df["entry_band_distance_pips"]
    /
    analysis_df["entry_atr14_pips"]
)


analysis_df["relative_band_distance"] = (
    analysis_df["relative_band_distance"]
    .replace(
        [np.inf, -np.inf],
        np.nan
    )
)


analysis_df = analysis_df[
    analysis_df["relative_band_distance"].notna()
].copy()


analysis_df = (
    analysis_df
    .sort_values("entry_time")
    .reset_index(drop=True)
)


analysis_df["year"] = (
    analysis_df["entry_time"].dt.year
)


print(f"分析対象件数: {len(analysis_df)}")
print(
    "期間:",
    analysis_df["entry_time"].min(),
    "～",
    analysis_df["entry_time"].max()
)


# ============================================================
# 条件抽出
# ============================================================

base_df = analysis_df.copy()


c_df = analysis_df[
    analysis_df["relative_band_distance"]
    .between(
        C_LOWER,
        C_UPPER,
        inclusive="both"
    )
].copy()


d_df = analysis_df[
    analysis_df["relative_band_distance"]
    .between(
        D_LOWER,
        D_UPPER,
        inclusive="both"
    )
].copy()


print(f"\nBASE取引数: {len(base_df)}")
print(f"C取引数:    {len(c_df)}")
print(f"D取引数:    {len(d_df)}")


# ============================================================
# 統計計算
# ============================================================

def calculate_stats(source_df):

    profits = source_df["profit"]

    trades = len(profits)

    if trades == 0:

        return {
            "trades": 0,
            "wins": 0,
            "losses": 0,
            "win_rate": np.nan,
            "gross_profit": 0,
            "gross_loss": 0,
            "profit_factor": np.nan,
            "avg_win": np.nan,
            "avg_loss": np.nan,
            "avg_loss_abs": np.nan,
            "avg_profit": np.nan,
            "total_profit": 0,
            "max_win": np.nan,
            "max_loss": np.nan,
        }

    wins = profits[profits > 0]
    losses = profits[profits < 0]

    win_count = len(wins)
    loss_count = len(losses)

    win_rate = (
        win_count
        /
        trades
        *
        100
    )

    gross_profit = wins.sum()

    gross_loss = -losses.sum()

    if gross_loss > 0:

        profit_factor = (
            gross_profit
            /
            gross_loss
        )

    elif gross_profit > 0:

        profit_factor = np.inf

    else:

        profit_factor = np.nan

    avg_win = (
        wins.mean()
        if win_count > 0
        else np.nan
    )

    avg_loss = (
        losses.mean()
        if loss_count > 0
        else np.nan
    )

    avg_loss_abs = (
        abs(avg_loss)
        if pd.notna(avg_loss)
        else np.nan
    )

    avg_profit = profits.mean()

    total_profit = profits.sum()

    max_win = (
        wins.max()
        if win_count > 0
        else np.nan
    )

    max_loss = (
        losses.min()
        if loss_count > 0
        else np.nan
    )

    return {
        "trades": trades,
        "wins": win_count,
        "losses": loss_count,
        "win_rate": win_rate,
        "gross_profit": gross_profit,
        "gross_loss": gross_loss,
        "profit_factor": profit_factor,
        "avg_win": avg_win,
        "avg_loss": avg_loss,
        "avg_loss_abs": avg_loss_abs,
        "avg_profit": avg_profit,
        "total_profit": total_profit,
        "max_win": max_win,
        "max_loss": max_loss,
    }


# ============================================================
# Dを時系列で4分割
# ============================================================

d_sorted = (
    d_df
    .sort_values("entry_time")
    .reset_index(drop=True)
)


# DataFrameそのものではなく、
# 行番号を4分割する
period_indices = np.array_split(
    np.arange(len(d_sorted)),
    4
)


period_rows = []


for i, indices in enumerate(
    period_indices,
    start=1
):

    # 行番号を使ってDataFrameとして抽出
    period_df = (
        d_sorted
        .iloc[indices]
        .copy()
        .reset_index(drop=True)
    )

    stats = calculate_stats(
        period_df
    )

    stats["period"] = (
        f"Period{i}"
    )

    stats["start_date"] = (
        period_df["entry_time"].min()
    )

    stats["end_date"] = (
        period_df["entry_time"].max()
    )

    period_rows.append(stats)


period_result_df = pd.DataFrame(
    period_rows
)


period_columns = [
    "period",
    "start_date",
    "end_date",
    "trades",
    "wins",
    "losses",
    "win_rate",
    "gross_profit",
    "gross_loss",
    "profit_factor",
    "avg_win",
    "avg_loss",
    "avg_loss_abs",
    "avg_profit",
    "total_profit",
    "max_win",
    "max_loss",
]


period_result_df = period_result_df[
    period_columns
]


period_result_df.to_csv(
    PERIOD_CSV,
    index=False
)

# ============================================================
# D 年別分析
# ============================================================

year_rows = []


for year, year_df in d_df.groupby("year"):

    stats = calculate_stats(
        year_df
    )

    stats["year"] = year

    year_rows.append(stats)


year_result_df = pd.DataFrame(
    year_rows
)


year_columns = [
    "year",
    "trades",
    "wins",
    "losses",
    "win_rate",
    "gross_profit",
    "gross_loss",
    "profit_factor",
    "avg_win",
    "avg_loss",
    "avg_loss_abs",
    "avg_profit",
    "total_profit",
    "max_win",
    "max_loss",
]


year_result_df = (
    year_result_df[
        year_columns
    ]
    .sort_values("year")
)


year_result_df.to_csv(
    YEAR_CSV,
    index=False
)


# ============================================================
# BASE / C / D 年別比較
# ============================================================

comparison_rows = []


comparison_sets = {
    "BASE": base_df,
    "C_0.50_1.00": c_df,
    "D_0.60_1.00": d_df,
}


for candidate, candidate_df in comparison_sets.items():

    for year, year_df in candidate_df.groupby("year"):

        stats = calculate_stats(
            year_df
        )

        stats["candidate"] = candidate
        stats["year"] = year

        comparison_rows.append(stats)


comparison_df = pd.DataFrame(
    comparison_rows
)


comparison_columns = [
    "candidate",
    "year",
    "trades",
    "wins",
    "losses",
    "win_rate",
    "gross_profit",
    "gross_loss",
    "profit_factor",
    "avg_win",
    "avg_loss",
    "avg_loss_abs",
    "avg_profit",
    "total_profit",
]


comparison_df = (
    comparison_df[
        comparison_columns
    ]
    .sort_values(
        [
            "year",
            "candidate"
        ]
    )
)


comparison_df.to_csv(
    COMPARE_CSV,
    index=False
)


# ============================================================
# グラフ1：BASE / C / D 年別PF
# ============================================================

plt.figure(
    figsize=(12, 7)
)


for candidate in comparison_sets.keys():

    plot_df = comparison_df[
        comparison_df["candidate"]
        ==
        candidate
    ]

    plt.plot(
        plot_df["year"],
        plot_df["profit_factor"],
        marker="o",
        label=candidate
    )


plt.axhline(
    1.0,
    linestyle="--",
    linewidth=1
)


plt.title(
    "Yearly Profit Factor Comparison"
)

plt.xlabel("Year")
plt.ylabel("Profit Factor")

plt.legend()

plt.grid(alpha=0.3)

plt.tight_layout()

plt.savefig(
    PF_GRAPH,
    dpi=150
)

plt.close()


# ============================================================
# グラフ2：D 年別取引数
# ============================================================

plt.figure(
    figsize=(12, 7)
)


plt.bar(
    year_result_df["year"],
    year_result_df["trades"]
)


plt.title(
    "D: Yearly Trade Count"
)

plt.xlabel("Year")
plt.ylabel("Trades")

plt.grid(
    axis="y",
    alpha=0.3
)

plt.tight_layout()

plt.savefig(
    TRADES_GRAPH,
    dpi=150
)

plt.close()


# ============================================================
# グラフ3：D 年別勝率
# ============================================================

plt.figure(
    figsize=(12, 7)
)


plt.plot(
    year_result_df["year"],
    year_result_df["win_rate"],
    marker="o"
)


plt.title(
    "D: Yearly Win Rate"
)

plt.xlabel("Year")
plt.ylabel("Win Rate (%)")

plt.grid(alpha=0.3)

plt.tight_layout()

plt.savefig(
    WIN_RATE_GRAPH,
    dpi=150
)

plt.close()


# ============================================================
# グラフ4：D 年別平均損益
# ============================================================

plt.figure(
    figsize=(12, 7)
)


plt.plot(
    year_result_df["year"],
    year_result_df["avg_profit"],
    marker="o"
)


plt.axhline(
    0,
    linestyle="--",
    linewidth=1
)


plt.title(
    "D: Yearly Average Profit per Trade"
)

plt.xlabel("Year")
plt.ylabel("Average Profit")

plt.grid(alpha=0.3)

plt.tight_layout()

plt.savefig(
    AVG_PROFIT_GRAPH,
    dpi=150
)

plt.close()


# ============================================================
# グラフ5：D 平均利益 / 平均損失
# ============================================================

plt.figure(
    figsize=(12, 7)
)


plt.plot(
    year_result_df["year"],
    year_result_df["avg_win"],
    marker="o",
    label="Average Win"
)


plt.plot(
    year_result_df["year"],
    year_result_df["avg_loss_abs"],
    marker="o",
    label="Average Loss (Absolute)"
)


plt.title(
    "D: Yearly Average Win / Loss"
)

plt.xlabel("Year")
plt.ylabel("Profit / Loss")

plt.legend()

plt.grid(alpha=0.3)

plt.tight_layout()

plt.savefig(
    AVG_WIN_LOSS_GRAPH,
    dpi=150
)

plt.close()


# ============================================================
# コンソール表示
# ============================================================

print("\n")
print("=" * 60)
print("D PERIOD ANALYSIS")
print("=" * 60)

print(
    period_result_df.to_string(
        index=False
    )
)


print("\n")
print("=" * 60)
print("D YEAR ANALYSIS")
print("=" * 60)

print(
    year_result_df.to_string(
        index=False
    )
)


print("\n")
print("=" * 60)
print("OUTPUT FILES")
print("=" * 60)

print(PERIOD_CSV)
print(YEAR_CSV)
print(COMPARE_CSV)

print(PF_GRAPH)
print(TRADES_GRAPH)
print(WIN_RATE_GRAPH)
print(AVG_PROFIT_GRAPH)
print(AVG_WIN_LOSS_GRAPH)