Update app.py

app.py

@@ -430,8 +430,70 @@ def generate_portfolio_overview(as_of_month, segment):
     """
     df = _filter_master_by_month(as_of_month)
 
-    # If segment filtering required in future, `segment` param can be used to filter further
+    # If a segmentation column is provided, return per-segment breakdown
+    valid_segments = [
+        "fico_band",
+        "sourcing_channel",
+        "city_tier",
+        "occupation_type"
+    ]
 
+    if segment in valid_segments and segment in df.columns:
+        grp = segment
+
+        total_accounts = df.groupby(grp)["account_id"].nunique()
+
+        if "balance" in df.columns:
+            open_accounts = (
+                df.loc[df["balance"] > 0].groupby(grp)["account_id"].nunique()
+            )
+            total_balance = df.groupby(grp)["balance"].sum()
+        else:
+            open_accounts = pd.Series(0, index=total_accounts.index)
+            total_balance = pd.Series(0, index=total_accounts.index)
+
+        if "dpd" in df.columns:
+            bad_accounts = (
+                df.loc[df["dpd"].fillna(0) >= 30].groupby(grp)["account_id"].nunique()
+            )
+            bad_balance = (
+                df.assign(_bad_balance=df["balance"].where(df["dpd"].fillna(0) >= 30, 0))
+                .groupby(grp)["_bad_balance"].sum()
+            ) if "balance" in df.columns else pd.Series(0, index=total_accounts.index)
+        else:
+            bad_accounts = pd.Series(0, index=total_accounts.index)
+            bad_balance = pd.Series(0, index=total_accounts.index)
+
+        # NCL if present
+        ncl_cols = [c for c in df.columns if "ncl" in c.lower()]
+        if len(ncl_cols) > 0:
+            total_ncl = df.groupby(grp)[ncl_cols[0]].sum()
+            ncl_rate = (total_ncl / total_balance * 100).round(2).fillna(0)
+        else:
+            ncl_rate = (bad_balance / total_balance * 100).round(2).fillna(0)
+
+        # average fico per group if available
+        if "fico_score" in df.columns:
+            avg_fico = df.groupby(grp)["fico_score"].mean().round(1)
+        else:
+            avg_fico = pd.Series(float("nan"), index=total_accounts.index)
+
+        result = pd.DataFrame({
+            "Segment": total_accounts.index,
+            "Total_Accounts": total_accounts.values,
+            "Open_Accounts": open_accounts.reindex(total_accounts.index).fillna(0).astype(int).values,
+            "Bad_Accounts": bad_accounts.reindex(total_accounts.index).fillna(0).astype(int).values,
+            "Total_Balance": total_balance.reindex(total_accounts.index).fillna(0).values,
+            "NCL_Rate_pct": ncl_rate.reindex(total_accounts.index).fillna(0).values,
+            "Avg_FICO": avg_fico.reindex(total_accounts.index).fillna(float("nan")).values
+        })
+
+        # Sort by NCL rate descending
+        result = result.sort_values("NCL_Rate_pct", ascending=False).reset_index(drop=True)
+
+        return result
+
+    # Default: single-line overview
     total_accounts = df["account_id"].nunique() if "account_id" in df.columns else 0
 
     open_accounts = (
@@ -470,7 +532,6 @@ def generate_portfolio_overview(as_of_month, segment):
     if "fico_score" in df.columns:
         avg_fico = round(df["fico_score"].dropna().mean(), 1)
     elif "fico_band" in df.columns:
-        # attempt to map band midpoints like "600-650" -> 625
         def band_mid(b):
             try:
                 parts = b.split("-")