#!/usr/bin/env python3
"""基于修正后退费数据，重新预估6月端内营收"""
import math

# ---- 修正后的月度数据 ----
monthly = [
    ('2025-09', 4, 6096, 6096, 1529, 3.99),
    ('2025-10', 25, 30375, 30355, 2412, 12.58),
    ('2025-11', 24, 30676, 30676, 2971, 10.33),
    ('2025-12', 19, 19781, 19775, 3523, 5.61),
    ('2026-01', 42, 49358, 49314, 1937, 25.46),
    ('2026-02', 21, 30279, 30253, 1743, 17.36),
    ('2026-03', 65, 85431, 82234, 4166, 19.74),   # 修正
    ('2026-04', 90, 112906, 106810, 5667, 18.85),  # 修正
    ('2026-05', 50, 71740, 69142, 3300, 20.95),    # 修正
]

# 5月每日新增
may_daily_new = [
    ('2026-05-01', 63), ('2026-05-02', 60), ('2026-05-03', 60), ('2026-05-04', 52),
    ('2026-05-05', 70), ('2026-05-06', 98), ('2026-05-07', 133), ('2026-05-08', 86),
    ('2026-05-09', 89), ('2026-05-10', 106), ('2026-05-11', 98), ('2026-05-12', 82),
    ('2026-05-13', 99), ('2026-05-14', 98), ('2026-05-15', 166), ('2026-05-16', 116),
    ('2026-05-17', 123), ('2026-05-18', 88), ('2026-05-19', 115), ('2026-05-20', 132),
    ('2026-05-21', 130), ('2026-05-22', 127), ('2026-05-23', 200), ('2026-05-24', 121),
    ('2026-05-25', 67), ('2026-05-26', 104), ('2026-05-27', 99), ('2026-05-28', 112),
    ('2026-05-29', 107), ('2026-05-30', 121), ('2026-05-31', 178),
]

print("=" * 70)
print("端内营收预估分析（修正退费数据后）")
print("=" * 70)

# ---- 月度汇总 ----
print("\n## 月度汇总（2025-09 ~ 2026-05）")
print(f"{'月份':<10} {'订单':>5} {'GMV':>10} {'GSV':>10} {'新增':>7} {'GSV/新增':>9}")
print("-" * 58)
for m in monthly:
    print(f"{m[0]:<10} {m[1]:>5} {m[2]:>10,.0f} {m[3]:>10,.0f} {m[4]:>7,} {m[5]:>9.2f}")

# ---- 3/4/5月核心指标 ----
print("\n## 3/4/5月核心指标")
m345 = [m for m in monthly if m[0] >= '2026-03']
gsv_per_new_vals = [m[5] for m in m345]
avg_gsv_per_new = sum(gsv_per_new_vals) / len(gsv_per_new_vals)

for m in m345:
    print(f"  {m[0]}: GSV={m[3]:,.0f}元, 新增={m[4]:,}人, GSV/新增={m[5]:.2f}元, 订单={m[1]}单")

print(f"\n  GSV/新增用户均值: {avg_gsv_per_new:.2f} 元")
print(f"  范围: {min(gsv_per_new_vals):.2f} ~ {max(gsv_per_new_vals):.2f}")
cv = math.sqrt(sum((x-avg_gsv_per_new)**2 for x in gsv_per_new_vals)/len(gsv_per_new_vals))/avg_gsv_per_new*100
print(f"  CV(变异系数): {cv:.1f}%")

# ---- 5月每日新增趋势 ----
print("\n## 5月每日新增注册趋势")
may_total = sum(nu for _, nu in may_daily_new)
may_avg = may_total / len(may_daily_new)
may_late = [(dt, nu) for dt, nu in may_daily_new if dt >= '2026-05-16']
may_late_avg = sum(nu for _, nu in may_late) / len(may_late)
may_early = [(dt, nu) for dt, nu in may_daily_new if dt < '2026-05-16']
may_early_avg = sum(nu for _, nu in may_early) / len(may_early)

print(f"  5月总新增: {may_total} 人, 日均: {may_avg:.0f} 人")
print(f"  上半月(1-15日)日均: {may_early_avg:.0f} 人")
print(f"  下半月(16-31日)日均: {may_late_avg:.0f} 人")

# 按周
from datetime import date
weeks = {}
for dt, nu in may_daily_new:
    d = date.fromisoformat(dt)
    wk = d.isocalendar()[1]
    if wk not in weeks:
        weeks[wk] = {'total': 0, 'days': 0}
    weeks[wk]['total'] += nu
    weeks[wk]['days'] += 1

for wk in sorted(weeks):
    w = weeks[wk]
    print(f"  第{wk}周: 日均{w['total']/w['days']:.0f}人, 合计{w['total']}人")

# ---- 预估模型 ----
print("\n" + "=" * 70)
print("## 6月营收预估（6月共30天）")
print("=" * 70)

# 场景
scenarios = {
    '乐观': may_late_avg * 1.10,
    '基准': may_avg,
    '保守': may_late_avg,
    '悲观': may_late_avg * 0.85,
}

# 模型1: GSV/新增用户均值法
print("\n### 模型1: GSV/新增用户均值法")
print(f"  公式: 6月GSV = 6月预估新增 × {avg_gsv_per_new:.2f}")
for label, daily_new in scenarios.items():
    jun_new = daily_new * 30
    jun_gsv = jun_new * avg_gsv_per_new
    print(f"  {label}: 日均新增{daily_new:.0f}人 → 月新增{jun_new:.0f}人 → GSV≈{jun_gsv:,.0f}元")

# 模型2: 线性回归
print("\n### 模型2: 月度新增→GSV线性回归")
xs = [m[4] for m in m345]
ys = [m[3] for m in m345]
n = len(xs)
x_mean = sum(xs) / n
y_mean = sum(ys) / n
slope = sum((xs[i]-x_mean)*(ys[i]-y_mean) for i in range(n)) / sum((x-x_mean)**2 for x in xs)
intercept = y_mean - slope * x_mean
r2 = 1 - sum((ys[i]-(slope*xs[i]+intercept))**2 for i in range(n)) / sum((y-y_mean)**2 for y in ys)

print(f"  回归: GSV = {slope:.2f} × 新增 + {intercept:.0f}  (R²={r2:.4f})")
for label, daily_new in scenarios.items():
    jun_new = daily_new * 30
    jun_gsv = slope * jun_new + intercept
    print(f"  {label}: 新增{jun_new:.0f}人 → GSV≈{jun_gsv:,.0f}元")

# 综合预估
print("\n" + "=" * 70)
print("## 综合预估")
print("=" * 70)

print(f"\n{'场景':<8} {'日均新增':>8} {'月新增':>8} {'模型1 GSV':>12} {'模型2 GSV':>12} {'综合GSV':>12}")
print("-" * 58)
for label, daily_new in scenarios.items():
    jun_new = daily_new * 30
    gsv1 = jun_new * avg_gsv_per_new
    gsv2 = slope * jun_new + intercept
    gsv_avg = (gsv1 + gsv2) / 2
    print(f"{label:<8} {daily_new:>8.0f} {jun_new:>8.0f} {gsv1:>12,.0f} {gsv2:>12,.0f} {gsv_avg:>12,.0f}")

# 推荐区间
jun_new_base = may_avg * 30
gsv_base = (jun_new_base * avg_gsv_per_new + slope * jun_new_base + intercept) / 2
jun_new_conservative = may_late_avg * 30
gsv_conservative = (jun_new_conservative * avg_gsv_per_new + slope * jun_new_conservative + intercept) / 2
jun_new_optimistic = may_late_avg * 1.10 * 30
gsv_optimistic = (jun_new_optimistic * avg_gsv_per_new + slope * jun_new_optimistic + intercept) / 2

print(f"\n### 推荐预估区间")
print(f"  基准预估: 6月GSV ≈ {gsv_base:,.0f} 元（日均新增{may_avg:.0f}人）")
print(f"  保守预估: 6月GSV ≈ {gsv_conservative:,.0f} 元（日均新增{may_late_avg:.0f}人）")
print(f"  乐观预估: 6月GSV ≈ {gsv_optimistic:,.0f} 元（日均新增{may_late_avg*1.1:.0f}人）")
print(f"  预估区间: {gsv_conservative:,.0f} ~ {gsv_optimistic:,.0f} 元")

print("\n### 与修正前对比")
print(f"  修正前 GSV/新增均值: 20.71 → 修正后: {avg_gsv_per_new:.2f}")
print(f"  修正前基准预估: ~67,720 → 修正后: ~{gsv_base:,.0f}")

print("\n### 关键假设与风险")
print(f"  1. GSV/新增用户指标 CV={cv:.1f}%，稳定性好，假设6月延续")
print(f"  2. 5月新增3,300人，较4月(5,667)下降42%")
print(f"  3. 5月下半月日均{may_late_avg:.0f}人，略高于上半月({may_early_avg:.0f}人)")
print(f"  4. 转化率逐月下降(1.87%→1.29%→0.79%)，但客单价上升抵消了影响")
print(f"  5. 建议：6月第1周结束后用实际新增数据回测修正")