Update app.py

app.py

@@ -3,6 +3,10 @@
 import pandas as pd
 import gradio as gr
 
+# ---------------------------------------------------
+# HELPERS
+# ---------------------------------------------------
+
 from helper.vintage_helpers import (
     create_booking_vintage
 )
@@ -11,11 +15,19 @@ from helper.data_merger import (
     merge_acq_perf
 )
 
+# ---------------------------------------------------
+# METRICS
+# ---------------------------------------------------
+
 from metrics.mix_metrics import (
     calculate_vintage_mix,
     calculate_limit_mix
 )
 
+# ---------------------------------------------------
+# ANALYTICS
+# ---------------------------------------------------
+
 from analytics.performance_analysis import (
     generate_metric_view
 )
@@ -33,38 +45,41 @@ perf = pd.read_csv(
 )
 
 # ---------------------------------------------------
-# CREATE VINTAGE
+# CREATE BOOKING VINTAGE
 # ---------------------------------------------------
 
-acq = create_booking_vintage(acq)
+acq = create_booking_vintage(
+    acq,
+    booking_date_col="booking_date"
+)
 
 # ---------------------------------------------------
-# MERGE DATA
+# CREATE MASTER PERFORMANCE DATASET
 # ---------------------------------------------------
 
 master_df = merge_acq_perf(
-    acq,
-    perf
+    acq_df=acq,
+    perf_df=perf
 )
 
 # ---------------------------------------------------
-# ACQUISITION ANALYSIS ENGINE
+# ACQUISITION ANALYSIS
 # ---------------------------------------------------
 
 def run_acquisition_analysis(
     analysis_type,
-    dimension
+    category
 ):
 
-    # ----------------------------------------
-    # MIX ANALYSIS
-    # ----------------------------------------
+    # -----------------------------------------
+    # PORTFOLIO MIX
+    # -----------------------------------------
 
-    if analysis_type == "Mix Analysis":
+    if analysis_type == "Portfolio Mix":
 
         result = (
             acq.groupby(
-                ["booking_vintage", dimension]
+                ["booking_vintage", category]
             )
             .agg(
                 count=("account_id", "nunique"),
@@ -82,34 +97,20 @@ def run_acquisition_analysis(
             result["count"] / vintage_total
         ) * 100
 
-        result["rate"] = result["rate"].round(2)
-
-        result = result.rename(
-            columns={
-                "booking_vintage": "vintage",
-                dimension: "segment"
-            }
+        result["rate"] = (
+            result["rate"]
+            .round(2)
         )
 
-        return result[
-            [
-                "vintage",
-                "segment",
-                "count",
-                "balance",
-                "rate"
-            ]
-        ]
+    # -----------------------------------------
+    # CREDIT LINE CONCENTRATION
+    # -----------------------------------------
 
-    # ----------------------------------------
-    # LIMIT ANALYSIS
-    # ----------------------------------------
-
-    elif analysis_type == "Limit Analysis":
+    elif analysis_type == "Credit Line Concentration":
 
         result = (
             acq.groupby(
-                ["booking_vintage", dimension]
+                ["booking_vintage", category]
             )
             .agg(
                 count=("account_id", "nunique"),
@@ -127,40 +128,53 @@ def run_acquisition_analysis(
             result["balance"] / vintage_total
         ) * 100
 
-        result["rate"] = result["rate"].round(2)
-
-        result = result.rename(
-            columns={
-                "booking_vintage": "vintage",
-                dimension: "segment"
-            }
+        result["rate"] = (
+            result["rate"]
+            .round(2)
         )
 
-        return result[
-            [
-                "vintage",
-                "segment",
-                "count",
-                "balance",
-                "rate"
-            ]
-        ]
-
     else:
 
         return pd.DataFrame()
 
+    # -----------------------------------------
+    # STANDARDIZED OUTPUT
+    # -----------------------------------------
+
+    result = result.rename(
+        columns={
+            "booking_vintage": "Vintage",
+            category: "Category",
+            "count": "Count",
+            "balance": "Balance",
+            "rate": "Rate"
+        }
+    )
+
+    return result[
+        [
+            "Vintage",
+            "Category",
+            "Count",
+            "Balance",
+            "Rate"
+        ]
+    ]
 
 # ---------------------------------------------------
-# PERFORMANCE ANALYSIS ENGINE
+# PERFORMANCE ANALYSIS
 # ---------------------------------------------------
 
 def run_performance_analysis(
     metric_name,
-    portfolio_view
+    view_level
 ):
 
-    group_mapping = {
+    # -----------------------------------------
+    # VIEW MAPPING
+    # -----------------------------------------
+
+    view_mapping = {
 
         "Overall": None,
 
@@ -177,176 +191,138 @@ def run_performance_analysis(
         "occupation_type"
     }
 
-    group_col = group_mapping[
-        portfolio_view
+    group_col = view_mapping[
+        view_level
     ]
 
+    # -----------------------------------------
+    # CALL ANALYTICS ENGINE
+    # -----------------------------------------
+
     result = generate_metric_view(
         df=master_df,
         metric_name=metric_name,
         group_col=group_col
     )
 
-    # ----------------------------------------
-    # IDENTIFY METRIC COLUMN
-    # ----------------------------------------
-
-    metric_col = [
-        col for col in result.columns
-        if "rate" in col.lower()
-    ][0]
-
-    # ----------------------------------------
-    # STANDARDIZED OUTPUT
-    # ----------------------------------------
+    # -----------------------------------------
+    # STANDARDIZE OUTPUT
+    # -----------------------------------------
 
     if group_col is not None:
 
-        final_result = result.rename(
+        result = result.rename(
             columns={
-                group_col: "segment",
-                metric_col: "rate"
+                group_col: "Category"
             }
         )
 
     else:
 
-        final_result = result.rename(
-            columns={
-                metric_col: "rate"
-            }
-        )
-
-        final_result["segment"] = "Overall"
-
-    # ----------------------------------------
-    # BALANCE LOGIC
-    # ----------------------------------------
-
-    # Delinquency metrics
-    if metric_name in [
-        "30+@3",
-        "30+@6",
-        "60+@6"
-    ]:
+        result["Category"] = "Overall"
 
-        if "bad_balance" in final_result.columns:
+    # -----------------------------------------
+    # IDENTIFY RATE COLUMN
+    # -----------------------------------------
 
-            final_result["balance"] = (
-                final_result["bad_balance"]
-            )
-
-        else:
-
-            final_result["balance"] = None
-
-    # NCL metric
-    elif metric_name == "Yr1 NCL":
-
-        if "total_balance" in final_result.columns:
-
-            final_result["balance"] = (
-                final_result["total_balance"]
-            )
-
-        else:
+    rate_col = [
+        col for col in result.columns
+        if "rate" in col.lower()
+    ][0]
 
-            final_result["balance"] = None
+    # -----------------------------------------
+    # OUTPUT FORMAT
+    # -----------------------------------------
 
-    # ----------------------------------------
-    # FINAL OUTPUT FORMAT
-    # ----------------------------------------
+    final_result = pd.DataFrame()
 
-    final_result = final_result.rename(
-        columns={
-            "booking_vintage": "vintage",
-            "total_accounts": "count"
-        }
+    final_result["Vintage"] = (
+        result["booking_vintage"]
     )
 
-    required_cols = [
-        "vintage",
-        "segment",
-        "count",
-        "balance",
-        "rate"
-    ]
-
-    for col in required_cols:
+    final_result["Category"] = (
+        result["Category"]
+    )
 
-        if col not in final_result.columns:
+    final_result["Count"] = (
+        result["total_accounts"]
+    )
 
-            final_result[col] = None
+    final_result["Balance"] = (
+        result["total_balance"]
+    )
 
-    return final_result[
-        required_cols
-    ]
+    final_result["Rate"] = (
+        result[rate_col]
+        .round(2)
+    )
 
+    return final_result
 
 # ---------------------------------------------------
-# DYNAMIC DROPDOWN LOGIC
+# DYNAMIC DROPDOWNS
 # ---------------------------------------------------
 
 def update_analysis_dropdown(
-    dataset_type
+    dataset
 ):
 
-    # ----------------------------------------
+    # -----------------------------------------
     # ACQUISITION
-    # ----------------------------------------
+    # -----------------------------------------
 
-    if dataset_type == "Acquisition":
+    if dataset == "Acquisition":
 
         return gr.update(
             choices=[
-                "Mix Analysis",
-                "Limit Analysis"
+                "Portfolio Mix",
+                "Credit Line Concentration"
             ],
-            value="Mix Analysis"
+            value="Portfolio Mix"
         )
 
-    # ----------------------------------------
+    # -----------------------------------------
     # PERFORMANCE
-    # ----------------------------------------
+    # -----------------------------------------
 
-    elif dataset_type == "Performance":
+    elif dataset == "Performance":
 
         return gr.update(
             choices=[
                 "30+@3",
                 "30+@6",
-                "60+@6",
+                "30+@9",
                 "Yr1 NCL"
             ],
             value="30+@6"
         )
 
 
-def update_dimension_dropdown(
-    dataset_type
+def update_category_dropdown(
+    dataset
 ):
 
-    # ----------------------------------------
+    # -----------------------------------------
     # ACQUISITION
-    # ----------------------------------------
+    # -----------------------------------------
 
-    if dataset_type == "Acquisition":
+    if dataset == "Acquisition":
 
         return gr.update(
             choices=[
-                "sourcing_channel",
                 "fico_band",
+                "sourcing_channel",
                 "city_tier",
                 "occupation_type"
             ],
-            value="sourcing_channel"
+            value="fico_band"
         )
 
-    # ----------------------------------------
+    # -----------------------------------------
     # PERFORMANCE
-    # ----------------------------------------
+    # -----------------------------------------
 
-    elif dataset_type == "Performance":
+    elif dataset == "Performance":
 
         return gr.update(
             choices=[
@@ -359,44 +335,42 @@ def update_dimension_dropdown(
             value="Overall"
         )
 
-
 # ---------------------------------------------------
 # MASTER ROUTER
 # ---------------------------------------------------
 
 def run_analysis(
-    dataset_type,
-    analysis_selection,
-    dimension_selection
+    dataset,
+    analysis,
+    category
 ):
 
-    # ----------------------------------------
-    # ACQUISITION ANALYSIS
-    # ----------------------------------------
+    # -----------------------------------------
+    # ACQUISITION
+    # -----------------------------------------
 
-    if dataset_type == "Acquisition":
+    if dataset == "Acquisition":
 
         return run_acquisition_analysis(
-            analysis_type=analysis_selection,
-            dimension=dimension_selection
+            analysis_type=analysis,
+            category=category
         )
 
-    # ----------------------------------------
-    # PERFORMANCE ANALYSIS
-    # ----------------------------------------
+    # -----------------------------------------
+    # PERFORMANCE
+    # -----------------------------------------
 
-    elif dataset_type == "Performance":
+    elif dataset == "Performance":
 
         return run_performance_analysis(
-            metric_name=analysis_selection,
-            portfolio_view=dimension_selection
+            metric_name=analysis,
+            view_level=category
         )
 
     else:
 
         return pd.DataFrame()
 
-
 # ---------------------------------------------------
 # GRADIO UI
 # ---------------------------------------------------
@@ -409,7 +383,7 @@ with gr.Blocks() as app:
 
     with gr.Row():
 
-        dataset_type = gr.Dropdown(
+        dataset_dropdown = gr.Dropdown(
             choices=[
                 "Acquisition",
                 "Performance"
@@ -418,45 +392,45 @@ with gr.Blocks() as app:
             label="Dataset"
         )
 
-        analysis_selection = gr.Dropdown(
+        analysis_dropdown = gr.Dropdown(
             choices=[
-                "Mix Analysis",
-                "Limit Analysis"
+                "Portfolio Mix",
+                "Credit Line Concentration"
             ],
-            value="Mix Analysis",
+            value="Portfolio Mix",
             label="Analysis"
         )
 
-        dimension_selection = gr.Dropdown(
+        category_dropdown = gr.Dropdown(
             choices=[
-                "sourcing_channel",
                 "fico_band",
+                "sourcing_channel",
                 "city_tier",
                 "occupation_type"
             ],
-            value="sourcing_channel",
-            label="Dimension"
+            value="fico_band",
+            label="Category / View"
         )
 
-    # ----------------------------------------
+    # -----------------------------------------
     # DYNAMIC DROPDOWNS
-    # ----------------------------------------
+    # -----------------------------------------
 
-    dataset_type.change(
+    dataset_dropdown.change(
         fn=update_analysis_dropdown,
-        inputs=dataset_type,
-        outputs=analysis_selection
+        inputs=dataset_dropdown,
+        outputs=analysis_dropdown
     )
 
-    dataset_type.change(
-        fn=update_dimension_dropdown,
-        inputs=dataset_type,
-        outputs=dimension_selection
+    dataset_dropdown.change(
+        fn=update_category_dropdown,
+        inputs=dataset_dropdown,
+        outputs=category_dropdown
     )
 
-    # ----------------------------------------
+    # -----------------------------------------
     # RUN BUTTON
-    # ----------------------------------------
+    # -----------------------------------------
 
     run_button = gr.Button(
         "Run Analysis"
@@ -467,9 +441,9 @@ with gr.Blocks() as app:
     run_button.click(
         fn=run_analysis,
         inputs=[
-            dataset_type,
-            analysis_selection,
-            dimension_selection
+            dataset_dropdown,
+            analysis_dropdown,
+            category_dropdown
         ],
         outputs=output_table
     )