PD v2 Processing block

README.md

@@ -1 +1,10 @@
-**Hello world!!!**
+## Overview
+This block (`block.py`) is responsible for loading and scoring the model.
+
+## Key Inputs & Outputs
+- **Request**: Refer to `request_schema.json` for detailed input fields and validation rules.  
+- **Response**: Refer to `response_schema.json` for the returned structure and data types.
+
+## Implementation Details
+- All core logic resides in `block.py` within the `__main__` function.
+- Example usage and validation are demonstrated in `test_block.py`.

block.py

@@ -1,21 +1,65 @@
-@flowx_block
-def example_function(request: dict) -> dict:
+import logging
+import xgboost as xgb
+import joblib
+import pandas as pd
 
-  # Processing logic here...
+# Configure logging
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(name)s - %(message)s",
+)
+logger = logging.getLogger(__name__)
 
-  return {
-    "meta_info": [
-      {
-        "name": "created_date",
-        "type": "string",
-        "value": "2024-11-05"
-      }
-    ],
-    "fields": [
-      {
-        "name": "",
-        "type": "",
-        "value": ""
-      }
-    ]
+def __main__(pti : float, score_results : float, BALMAG01 : float, revolving_amount_monthly_payment : float, closed_with_balance_amount_current_balance : float, AT31S : int, AT20S : int, BC21S : int, record_counts_revolving_trade_count : int, record_counts_total_trade_count : int, PAYMNT10 : float, AGG102 : float, total_amount_high_credit : float, revolving_amount_current_balance : float, total_amount_current_balance : float, REV83 : float, revolving_amount_high_credit : float, closed_with_balance_amount_monthly_payment : float, revolving_amount_percent_available_credit : float, AGG101 : float, revolving_amount_credit_limit : float, AT09S : int, US01S : int)->dict:
+
+    input_data = {
+        "pti": pti, "score_results": score_results, "BALMAG01": BALMAG01,
+        "revolving_amount_monthly_payment": revolving_amount_monthly_payment,
+        "closed_with_balance_amount_current_balance": closed_with_balance_amount_current_balance,
+        "AT31S": AT31S, "AT20S": AT20S, "BC21S": BC21S,
+        "record_counts_revolving_trade_count": record_counts_revolving_trade_count,
+        "record_counts_total_trade_count": record_counts_total_trade_count, "PAYMNT10": PAYMNT10,
+        "AGG102": AGG102, "total_amount_high_credit": total_amount_high_credit,
+        "revolving_amount_current_balance": revolving_amount_current_balance,
+        "total_amount_current_balance": total_amount_current_balance, "REV83": REV83,
+        "revolving_amount_high_credit": revolving_amount_high_credit,
+        "closed_with_balance_amount_monthly_payment": closed_with_balance_amount_monthly_payment,
+        "revolving_amount_percent_available_credit": revolving_amount_percent_available_credit,
+        "AGG101": AGG101, "revolving_amount_credit_limit": revolving_amount_credit_limit,
+        "AT09S": AT09S, "US01S": US01S, "has_mortgage": None
     }
+
+    # Load model
+    try:
+        model = joblib.load("./xgboost_model.joblib") 
+        # model = joblib.load("C:/Users/cbollu/Downloads/test_blocks/test_blocks/sequence-1/pd_v2_processing/xgboost_model.joblib") 
+    except Exception as e:
+        logger.exception("An unexpected error occurred while loading the model.")
+        raise e
+    df_pre_processed = pd.DataFrame(input_data, index=[0])
+
+    if df_pre_processed.empty:
+        print("PD V2 Pre Processed DataFrame is empty.")
+
+    expected_features = model.feature_names
+    actual_features = df_pre_processed.columns.tolist()
+    missing_features = [feature for feature in expected_features if feature not in actual_features]
+    
+    # Add missing features as None (NaN) values
+    for feature in missing_features:
+        df_pre_processed[feature] = None
+
+    # Convert object columns to categorical
+    for col in df_pre_processed.columns:
+        if df_pre_processed[col].dtype == 'object':
+            df_pre_processed[col] = pd.Categorical(df_pre_processed[col])
+
+    # Prepare data for prediction
+    dmatrix = xgb.DMatrix(df_pre_processed[expected_features], enable_categorical=True)
+    
+    # Make prediction
+    prediction = model.predict(dmatrix)[0]
+
+    logger.info(f"PD V2 Predicted Score: {prediction}")
+
+    return {"probability": float(prediction)}

request_schema.json

@@ -1 +1,99 @@
-{}
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "type": "object",
+  "properties": {
+    "pti": {
+      "type": ["number", "null"],
+      "description": "external + internal monthly payment to income ratio"
+    },
+    "score_results": {
+      "type": ["number", "null"],
+      "description": "TransUnion score"
+    },
+    "BALMAG01": {
+      "type": ["number", "null"],
+      "description": "Non-mortgage balance magnitude"
+    },
+    "revolving_amount_monthly_payment": {
+      "type": ["number", "null"],
+      "description": "Minimum amount the borrower is required to pay each month to maintain the account in good standing"
+    },
+    "closed_with_balance_amount_current_balance": {
+      "type": ["number", "null"],
+      "description": "The current balance of closed credit accounts"
+    },
+    "AT31S": {
+      "type": ["integer", "null"],
+      "description": "Percentage of open trades > 75% of credit line verified in past 12 months"
+    },
+    "AT20S": {
+      "type": ["integer", "null"],
+      "description": "Months since oldest trade opened"
+    },
+    "BC21S": {
+      "type": ["integer", "null"],
+      "description": "Months since most recent credit card trade opened"
+    },
+    "record_counts_revolving_trade_count": {
+      "type": ["integer", "null"],
+      "description": "Records in the database related to revolving trade accounts (a credit card account)"
+    },
+    "record_counts_total_trade_count": {
+      "type": ["integer", "null"],
+      "description": "Total number of trade-related (transaction) records"
+    },
+    "PAYMNT10": {
+      "type": ["number", "null"],
+      "description": "Number of payments in the last quarter"
+    },
+    "AGG102": {
+      "type": ["number", "null"],
+      "description": "Aggregate non-mortgage balances for month 2"
+    },
+    "total_amount_high_credit": {
+      "type": ["number", "null"],
+      "description": "The highest credit amount extended across all credit accounts"
+    },
+    "revolving_amount_current_balance": {
+      "type": ["number", "null"],
+      "description": "The current owed balance on revolving credit accounts"
+    },
+    "total_amount_current_balance": {
+      "type": ["number", "null"],
+      "description": "The total current balance across all credit accounts"
+    },
+    "REV83": {
+      "type": ["number", "null"],
+      "description": "Months since a revolving account last exceeded 75% utilization"
+    },
+    "revolving_amount_high_credit": {
+      "type": ["number", "null"],
+      "description": "The highest credit amount that has been extended to the borrower in revolving credit accounts"
+    },
+    "closed_with_balance_amount_monthly_payment": {
+      "type": ["number", "null"],
+      "description": "The monthly payment amount for closed credit accounts (loans)"
+    },
+    "revolving_amount_percent_available_credit": {
+      "type": ["number", "null"],
+      "description": "The percentage of available credit that has been utilized in revolving credit accounts"
+    },
+    "AGG101": {
+      "type": ["number", "null"],
+      "description": "Aggregate non-mortgage balances for month 1"
+    },
+    "revolving_amount_credit_limit": {
+      "type": ["number", "null"],
+      "description": "The total credit limit on revolving credit accounts"
+    },
+    "AT09S": {
+      "type": ["integer", "null"],
+      "description": "Number of trades opened in past 24 months"
+    },
+    "US01S": {
+      "type": ["integer", "null"],
+      "description": "Number of unsecured installment trades"
+    }
+  },
+  "required": []
+}

requirements.txt

@@ -1 +1,5 @@
-{}
+jsonschema==4.23.0
+xgboost==1.7.5
+joblib==1.3.2
+pandas==2.2.2
+numpy==1.23.5

response_schema.json

@@ -1 +1,10 @@
-{}
+{
+    "$schema": "http://json-schema.org/draft-07/schema#",
+    "type": "object",
+    "properties": {
+      "probability": {
+        "type": "number",
+        "description": "Model predicted score."
+      }
+    }
+  }

test_block.py

@@ -0,0 +1,15 @@
+import unittest
+from block import __main__
+
+class TestBlock(unittest.TestCase):
+
+    def test_main_success(self):
+        result = __main__(pti= 0.3277136364,score_results= 600.0,BALMAG01= 196.0,revolving_amount_monthly_payment= 56.0,closed_with_balance_amount_current_balance= 8411.0,AT31S= 71.0,AT20S= 166.0,BC21S= 4.0,record_counts_revolving_trade_count= 9.0,record_counts_total_trade_count= 18.0,PAYMNT10= 4.0,AGG102= 24994.0,total_amount_high_credit= 53807.0,revolving_amount_current_balance= 1635.0,total_amount_current_balance= 38353.0,REV83= 0.0,revolving_amount_high_credit= 1720.0,closed_with_balance_amount_monthly_payment= 0.0,revolving_amount_percent_available_credit= 18.0,AGG101= 11043.0,revolving_amount_credit_limit= 2000.0,AT09S= 4.0,US01S= 0.0)
+        self.assertAlmostEqual(result['probability'], 0.33663413, places=7)
+
+    # def test_main_invalid_input(self):
+    #     with self.assertRaises(TypeError):
+    #         __main__(pti= 231,score_results= 600.0,BALMAG01= 196.0,revolving_amount_monthly_payment= 56.0,closed_with_balance_amount_current_balance= 8411.0,AT31S= 71.0,AT20S= 166.0,BC21S= 4.0,record_counts_revolving_trade_count= 9.0,record_counts_total_trade_count= 18.0,PAYMNT10= 4.0,AGG102= 24994.0,total_amount_high_credit= 53807.0,revolving_amount_current_balance= 1635.0,total_amount_current_balance= 38353.0,REV83= 0.0,revolving_amount_high_credit= 1720.0,closed_with_balance_amount_monthly_payment= 0.0,revolving_amount_percent_available_credit= 18.0,AGG101= 11043.0,revolving_amount_credit_limit= 2000.0,AT09S= 4.0,US01S= 0.0)
+
+if __name__ == "__main__":
+    unittest.main()