Update src/analyze_portfolio_risk.py

src/analyze_portfolio_risk.py

@@ -1,83 +1,83 @@
-# compute_my_srisk.py
-# ==================================================
-# 💼 개인 포트폴리오 Srisk 계산기 (최종 안정화 버전)
-# ==================================================
-import pandas as pd
-import numpy as np
-import re
-
-# === 문자열에서 숫자만 추출 ===
-def extract_number(x):
-    if isinstance(x, str):
-        match = re.search(r"[-+]?\d*\.\d+|\d+", x)
-        return float(match.group()) if match else np.nan
-    return x
-
-
-# === Robust Z-score (IQR 기준 + 클리핑) ===
-def robust_zscore(val, series):
-    series = series.dropna()
-    q1, q3 = series.quantile([0.25, 0.75])
-    iqr = q3 - q1
-    med = series.median()
-    if iqr == 0:
-        return 0
-    z = (val - med) / iqr
-    return np.clip(z, -3, 3)
-
-
-# === Srisk 계산 함수 ===
-def compute_srisk_from_portfolio(portfolio, full_df):
-    df = full_df[full_df["Ticker"].isin(portfolio.keys())].copy()
-
-    # === 가중 평균 ===
-    df["Weight"] = df["Ticker"].map(portfolio)
-    sigma_p = np.average(df["Sigma"], weights=df["Weight"])
-    mdd_p   = np.average(df["MDD"],  weights=df["Weight"])
-    beta_p  = np.average(df["Beta"], weights=df["Weight"])
-
-    # === 섹터 기반 HHI ===
-    sector_map = df.set_index("Ticker")["Sector"].to_dict()
-    sector_weights = {}
-    for ticker, w in portfolio.items():
-        sector = sector_map.get(ticker, "Unknown")
-        sector_weights[sector] = sector_weights.get(sector, 0) + w
-    hhi_p = sum([w**2 for w in sector_weights.values()])
-
-    # === 시장 전체 기준 robust Z-score 계산 ===
-    S_sigma =  robust_zscore(sigma_p, full_df["Sigma"])              # 변동성 ↑ → 위험 ↑
-    S_mdd   = -robust_zscore(abs(mdd_p), abs(full_df["MDD"]))        # 낙폭 ↑ → 위험 ↑
-    S_beta  =  robust_zscore(beta_p, full_df["Beta"])                # 민감도 ↑ → 위험 ↑
-    S_hhi   =  robust_zscore(hhi_p, pd.Series([(1/len(portfolio))**2]*len(full_df)))
-
-    # === Srisk 계산 (조정된 가중치)
-    Srisk = 0.35*S_sigma + 0.25*S_mdd + 0.20*S_beta + 0.10*S_hhi
-
-    # === 구간 분류
-    if Srisk < 0.33:
-        category = "SAFE"
-    elif Srisk < 0.66:
-        category = "NEUTRAL"
-    else:
-        category = "AGGRESSIVE"
-
-    return {
-        "Srisk": Srisk,
-        "Category": category,
-        "Sigma_p": sigma_p,
-        "MDD_p": mdd_p,
-        "Beta_p": beta_p,
-        "HHI": hhi_p,
-        "Sectors": sector_weights
-    }
-
-def classify_investment_style(full_df, portfolio_list):
-    for col in ["Sigma", "MDD", "Beta"]:
-        full_df[col] = full_df[col].apply(extract_number)
-
-    df_port = pd.DataFrame(portfolio_list)
-    total_value = df_port['total_value'].sum()
-    portfolio_weights = (df_port.groupby('ticker')['total_value'].sum() / total_value).to_dict()
-    
-    r = compute_srisk_from_portfolio(portfolio_weights, full_df)
+# compute_my_srisk.py
+# ==================================================
+# 💼 개인 포트폴리오 Srisk 계산기 (최종 안정화 버전)
+# ==================================================
+import pandas as pd
+import numpy as np
+import re
+
+# === 문자열에서 숫자만 추출 ===
+def extract_number(x):
+    if isinstance(x, str):
+        match = re.search(r"[-+]?\d*\.\d+|\d+", x)
+        return float(match.group()) if match else np.nan
+    return x
+
+
+# === Robust Z-score (IQR 기준 + 클리핑) ===
+def robust_zscore(val, series):
+    series = series.dropna()
+    q1, q3 = series.quantile([0.25, 0.75])
+    iqr = q3 - q1
+    med = series.median()
+    if iqr == 0:
+        return 0
+    z = (val - med) / iqr
+    return np.clip(z, -3, 3)
+
+
+# === Srisk 계산 함수 ===
+def compute_srisk_from_portfolio(portfolio, full_df):
+    df = full_df[full_df["Ticker"].isin(portfolio.keys())].copy()
+
+    # === 가중 평균 ===
+    df["Weight"] = df["Ticker"].map(portfolio)
+    sigma_p = np.average(df["Sigma"], weights=df["Weight"])
+    mdd_p   = np.average(df["MDD"],  weights=df["Weight"])
+    beta_p  = np.average(df["Beta"], weights=df["Weight"])
+
+    # === 섹터 기반 HHI ===
+    sector_map = df.set_index("Ticker")["Sector"].to_dict()
+    sector_weights = {}
+    for ticker, w in portfolio.items():
+        sector = sector_map.get(ticker, "Unknown")
+        sector_weights[sector] = sector_weights.get(sector, 0) + w
+    hhi_p = sum([w**2 for w in sector_weights.values()])
+
+    # === 시장 전체 기준 robust Z-score 계산 ===
+    S_sigma =  robust_zscore(sigma_p, full_df["Sigma"])              # 변동성 ↑ → 위험 ↑
+    S_mdd   = -robust_zscore(abs(mdd_p), abs(full_df["MDD"]))        # 낙폭 ↑ → 위험 ↑
+    S_beta  =  robust_zscore(beta_p, full_df["Beta"])                # 민감도 ↑ → 위험 ↑
+    S_hhi   =  robust_zscore(hhi_p, pd.Series([(1/len(portfolio))**2]*len(full_df)))
+
+    # === Srisk 계산 (조정된 가중치)
+    Srisk = 0.35*S_sigma + 0.25*S_mdd + 0.20*S_beta + 0.10*S_hhi
+
+    # === 구간 분류
+    if Srisk < 0.33:
+        category = "SAFE"
+    elif Srisk < 0.66:
+        category = "NEUTRAL"
+    else:
+        category = "RISKY"
+
+    return {
+        "Srisk": Srisk,
+        "Category": category,
+        "Sigma_p": sigma_p,
+        "MDD_p": mdd_p,
+        "Beta_p": beta_p,
+        "HHI": hhi_p,
+        "Sectors": sector_weights
+    }
+
+def classify_investment_style(full_df, portfolio_list):
+    for col in ["Sigma", "MDD", "Beta"]:
+        full_df[col] = full_df[col].apply(extract_number)
+
+    df_port = pd.DataFrame(portfolio_list)
+    total_value = df_port['total_value'].sum()
+    portfolio_weights = (df_port.groupby('ticker')['total_value'].sum() / total_value).to_dict()
+    
+    r = compute_srisk_from_portfolio(portfolio_weights, full_df)
     return r['Srisk'], r['Category']