Update metrics/vintage_metrics.py

metrics/vintage_metrics.py

@@ -3,21 +3,22 @@
 import pandas as pd
 import numpy as np
 
+from helpers.filter_helpers import apply_filters
+
 
 def calculate_vintage_dpd(
     df,
     mob,
     dpd_threshold,
-    metric_type="count"
+    group_col=None,
+    filters=None
 ):
-    """
-    Generic DPD metric calculator.
-    Example:
-    30+@3
-    60+@6
-    """
 
-    mob_df = df[df["mob"] == mob].copy()
+    df = apply_filters(df, filters)
+
+    mob_df = df[
+        df["mob"] == mob
+    ].copy()
 
     mob_df["bad_flag"] = np.where(
         mob_df["dpd"] >= dpd_threshold,
@@ -25,51 +26,53 @@ def calculate_vintage_dpd(
         0
     )
 
-    vintage_summary = (
-        mob_df.groupby("booking_vintage")
+    group_columns = ["booking_vintage"]
+
+    if group_col is not None:
+        group_columns.append(group_col)
+
+    result = (
+        mob_df.groupby(group_columns)
         .agg(
             total_accounts=("account_id", "nunique"),
-            bad_accounts=("bad_flag", "sum"),
-            total_balance=("balance", "sum"),
-            bad_balance=(
-                "balance",
-                lambda x: x[
-                    mob_df.loc[x.index, "bad_flag"] == 1
-                ].sum()
-            )
+            bad_accounts=("bad_flag", "sum")
         )
         .reset_index()
     )
 
-    if metric_type == "count":
-
-        vintage_summary["dpd_rate"] = (
-            vintage_summary["bad_accounts"]
-            / vintage_summary["total_accounts"]
-        ) * 100
-
-    elif metric_type == "balance":
+    result["dpd_rate"] = (
+        result["bad_accounts"]
+        / result["total_accounts"]
+    ) * 100
 
-        vintage_summary["dpd_rate"] = (
-            vintage_summary["bad_balance"]
-            / vintage_summary["total_balance"]
-        ) * 100
+    result["dpd_rate"] = (
+        result["dpd_rate"]
+        .round(2)
+    )
 
-    return vintage_summary
+    return result
 
 
 def calculate_vintage_ncl(
     df,
-    mob
+    mob,
+    group_col=None,
+    filters=None
 ):
-    """
-    Calculates NCL rate at a given MOB.
-    """
 
-    mob_df = df[df["mob"] == mob].copy()
+    df = apply_filters(df, filters)
+
+    mob_df = df[
+        df["mob"] == mob
+    ].copy()
+
+    group_columns = ["booking_vintage"]
 
-    vintage_summary = (
-        mob_df.groupby("booking_vintage")
+    if group_col is not None:
+        group_columns.append(group_col)
+
+    result = (
+        mob_df.groupby(group_columns)
         .agg(
             total_balance=("balance", "sum"),
             total_ncl=("ncl_amount", "sum")
@@ -77,9 +80,14 @@ def calculate_vintage_ncl(
         .reset_index()
     )
 
-    vintage_summary["ncl_rate"] = (
-        vintage_summary["total_ncl"]
-        / vintage_summary["total_balance"]
+    result["ncl_rate"] = (
+        result["total_ncl"]
+        / result["total_balance"]
     ) * 100
 
-    return vintage_summary
+    result["ncl_rate"] = (
+        result["ncl_rate"]
+        .round(2)
+    )
+
+    return result