import pandas as pd
from decimal import Decimal

def calc_dupont_factors(income_stmt: pd.DataFrame, balance_sheet: pd.DataFrame):
    """计算杜邦分析五因素指标，金额使用Decimal保证精度"""
    revenue = Decimal(str(income_stmt["营业收入"].iloc[0]))
    ebit = Decimal(str(income_stmt["息税前利润"].iloc[0]))
    ebt = Decimal(str(income_stmt["税前利润"].iloc[0]))
    net_income = Decimal(str(income_stmt["净利润"].iloc[0]))
    tax = Decimal(str(income_stmt["所得税"].iloc[0]))
    total_assets = Decimal(str(balance_sheet["总资产"].iloc[0]))
    equity = Decimal(str(balance_sheet["股东权益"].iloc[0]))

    # 五因素逐一计算
    tax_burden = ebt / (ebt + tax)       # 税负因素
    interest = ebt / ebit                 # 利息因素
    op_margin = ebit / revenue            # 经营利润率
    asset_turnover = revenue / total_assets  # 资产周转率
    leverage = total_assets / equity      # 权益乘数
    roe = tax_burden * interest * op_margin * asset_turnover * leverage
    return {"税负因素": tax_burden, "利息因素": interest,
            "经营利润率": op_margin, "资产周转率": asset_turnover,
            "权益乘数": leverage, "ROE": roe}

def dupont_trend_analysis(multi_year_data: dict) -> pd.DataFrame:
    """多年杜邦分析趋势对比，输入为 {年份: (利润表, 资产负债表)} 字典"""
    results = []
    for year, (istmt, bs) in multi_year_data.items():
        factors = calc_dupont_factors(istmt, bs)
        factors["年份"] = year
        results.append(factors)
    df = pd.DataFrame(results).set_index("年份")
    # 转为百分比展示，更直观
    pct_cols = ["税负因素", "利息因素", "经营利润率", "ROE"]
    df[pct_cols] = df[pct_cols].apply(lambda x: x.apply(lambda v: f"{v:.2%}"))
    return df

import matplotlib.pyplot as plt
import numpy as np

def plot_waterfall(base_factors, current_factors, year_label):
    """绘制ROE因素贡献瀑布图"""
    labels = ["税负", "利息", "经营利润率", "资产周转率", "权益乘数"]
    # 计算各因素变化对ROE的边际贡献
    contributions = []
    for key in labels:
        delta = current_factors[key] - base_factors[key]
        contributions.append(float(delta) * 100)

    # 绘制瀑布图
    fig, ax = plt.subplots(figsize=(10, 6))
    cumulative = np.cumsum(contributions)
    colors = ["#2ecc71" if c >= 0 else "#e74c3c" for c in contributions]
    ax.bar(labels, contributions, color=colors, edgecolor="white")
    ax.set_title(f"{year_label} ROE变动因素贡献（百分点）")
    ax.axhline(y=0, color="black", linewidth=0.5)
    plt.tight_layout()
    plt.savefig(f"dupont_waterfall_{year_label}.png", dpi=150)
    plt.close()

from decimal import Decimal

def cvp_analysis(price, variable_cost, fixed_cost, actual_volume):
    """CVP本量利分析：计算盈亏平衡点和安全边际"""
    p = Decimal(str(price))
    vc = Decimal(str(variable_cost))
    fc = Decimal(str(fixed_cost))
    vol = Decimal(str(actual_volume))

    # 边际贡献 = 单价 - 单位变动成本
    cm = p - vc
    # 盈亏平衡点（销量）= 固定成本 / 边际贡献
    breakeven = fc / cm
    # 安全边际率 = (实际销量 - 盈亏平衡销量) / 实际销量
    margin_of_safety = (vol - breakeven) / vol
    return {"边际贡献": cm, "盈亏平衡销量": breakeven,
            "安全边际率": f"{margin_of_safety:.1%}",
            "当前利润": cm * vol - fc}

def cvp_sensitivity(price, vc, fc, base_volume):
    """敏感性分析：测试价格和成本变动对利润的影响"""
    results = []
    # 价格变动 -10% 到 +10%
    for pct in [-0.10, -0.05, 0, 0.05, 0.10]:
        new_price = Decimal(str(price)) * (1 + Decimal(str(pct)))
        new_cm = new_price - Decimal(str(vc))
        profit = new_cm * Decimal(str(base_volume)) - Decimal(str(fc))
        results.append({"变动": f"价格{pct:+.0%}", "利润": profit})
    # 成本变动 -10% 到 +10%
    for pct in [-0.10, -0.05, 0, 0.05, 0.10]:
        new_vc = Decimal(str(vc)) * (1 + Decimal(str(pct)))
        new_cm = Decimal(str(price)) - new_vc
        profit = new_cm * Decimal(str(base_volume)) - Decimal(str(fc))
        results.append({"变动": f"成本{pct:+.0%}", "利润": profit})
    return pd.DataFrame(results)

import pandas as pd
from decimal import Decimal

def variance_analysis(budget_df, actual_df):
    """方差分析：拆解收入差异为价格、数量和混合因素"""
    merged = budget_df.merge(actual_df, on="产品", suffixes=("_预算", "_实际"))
    results = []
    for _, row in merged.iterrows():
        bp = Decimal(str(row["单价_预算"]))
        ap = Decimal(str(row["单价_实际"]))
        bq = Decimal(str(row["数量_预算"]))
        aq = Decimal(str(row["数量_实际"]))
        # 三类差异计算
        price_var = (ap - bp) * aq          # 价格差异
        qty_var = (aq - bq) * bp            # 数量差异
        total_var = ap * aq - bp * bq       # 总差异
        mix_var = total_var - price_var - qty_var  # 混合差异
        results.append({"产品": row["产品"], "价格差异": price_var,
                        "数量差异": qty_var, "混合差异": mix_var,
                        "总差异": total_var})
    return pd.DataFrame(results)

# 正确做法：分子分母口径一致
equity = Decimal(str(bs["归母股东权益"].iloc[0]))
# 不要用总权益，否则杠杆倍数会被低估
leverage = total_assets / equity

# 阶梯式固定成本：不同产能区间对应不同的固定成本
def stepped_fixed_cost(volume):
    """根据产量区间返回对应的固定成本"""
    if volume <= 10000:
        return Decimal("500000")   # 基础产能
    elif volume <= 20000:
        return Decimal("800000")   # 需新增设备
    else:
        return Decimal("1200000")  # 需新建产线