Update post_processing.py

README.md

@@ -1 +1 @@
-**Hello world!!!**
+Series M v1 block

block.py

@@ -1,21 +1,65 @@
-@flowx_block
-def example_function(request: dict) -> dict:
+import logging
+from typing import Dict
+from pre_processing import pre_processing
+from processing import processing
+from post_processing import post_processing
+from parse_report import extract_model_variables
 
-  # 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__(application_id: str, creditBureau: dict) -> Dict:
+    """
+    Process a single application record via the scoring pipeline.
+    Returns:
+        A dict containing:
+          - prediction: float
+          - grade: str
+          - reason_codes: list or None
+    """
+    record = extract_model_variables(creditBureau)
+
+    if not record or all(v is None for v in record.values()):
+        final_result = {
+            'application_id': application_id,
+            'prediction':        0.99,
+            'grade':             'M14',
+            "reason_codes": [{
+                "code": None,
+                "rank": None,
+                "description": "Lack of account information"
+            }]
+        }
+        logger.info(
+            f"final_result (early exit due to missing or empty extracted variables): {final_result}")
+        return final_result
+
+    processed = pre_processing(record)
+
+    out = processing(processed)
+
+    final = post_processing(out, record)
+
+    final_result = {
+        'application_id': application_id,
+        'prediction':        out['prediction'],
+        'grade':             final['grade'],
+        'reason_codes': final['reason_codes'],
+    }
+
+    logger.info(f"final_result: {final_result}")
+
+    return final_result
+
+
+if __name__ == "__main__":
+    import json
+    import sys
+    with open(sys.argv[1]) as f:
+        data = json.load(f)
+    __main__(application_id=data["application_id"], creditBureau=data)

parse_report.py

@@ -0,0 +1,210 @@
+import json
+import sys
+
+VARIABLE_TO_CODE_MAP = {
+    "agg901": "00H86",
+    "agg902": "00H86",
+    "agg908": "00H86",
+    "agg911": "00H86",
+    "all231": "00ASH",
+    "at28a": "00WR3",
+    "at28b": "00WR3",
+    "at36s": "00WR3",
+    "balmag01": "00H88",
+    "balmag02": "00H88",
+    "balmag04": "00AI5",
+    "bc102s": "00WR3",
+    "bc104s": "00WR3",
+    "bc107s": "00WR3",
+    "bc20s": "00WR3",
+    "bc21s": "00WR3",
+    "bc28s": "00WR3",
+    "bc36s": "00WR3",
+    "bc97a": "00WR3",
+    "bc98a": "00WR3",
+    "bkc14": "00ASF",
+    "bkc321": "00ASI",
+    "bkc322": "00ASI",
+    "bkc323": "00ASI",
+    "bkc324": "00ASI",
+    "bkc327": "00ASI",
+    "bkc328": "00ASI",
+    "bkc84": "00ASF",
+    "br20s": "00WR3",
+    "ct319": "00AI9",
+    "ct320": "00AI9",
+    "cta11": "00AI9",
+    "cta20": "00AI9",
+    "cta21": "00AI9",
+    "cv13": "00V26",
+    "cv17": "00V26",
+    "cv21": "00V26",
+    "cv25": "00V26",
+    "duemag01": "00AI5",
+    "fi21s": "00WR3",
+    "fi34s": "00WR3",
+    "fi35s": "00WR3",
+    "g051s": "00WR3",
+    "g102s": "00WR3",
+    "g105s": "00WR3",
+    "g106s": "00WR3",
+    "g201a": "00WR3",
+    "g221d": "00WR3",
+    "g232s": "00WR3",
+    "g242b": "00WR3",
+    "g250a": "00WR3",
+    "g403s": "00WR3",
+    "g405s": "00WR3",
+    "g408s": "00WR3",
+    "g411s": "00WR3",
+    "g416s": "00WR3",
+    "g417s": "00WR3",
+    "g960s": "00WR3",
+    "g990s": "00WR3",
+    "index01": "00V53",
+    "index02": "00V53",
+    "mnpmag03": "00AI5",
+    "mt20s": "00WR3",
+    "mt34s": "00WR3",
+    "p02d": "00WBO",
+    "p02h": "00WBO",
+    "paymnt08": "00H91",
+    "pb28s": "00WR3",
+    "pb34s": "00WR3",
+    "re102s": "00WR3",
+    "re28s": "00WR3",
+    "re36s": "00WR3",
+    "ret12": "00ASF",
+    "rev12": "00ASF",
+    "rev13": "00ASF",
+    "rev14": "00ASF",
+    "rev201": "00ASG",
+    "rev202": "00ASG",
+    "rev203": "00ASG",
+    "rev223": "00ASG",
+    "rev224": "00ASG",
+    "rev225": "00ASG",
+    "rev231": "00ASH",
+    "rev232": "00ASH",
+    "rev233": "00ASH",
+    "rev252": "00ASH",
+    "rev253": "00ASH",
+    "rev321": "00ASI",
+    "rev322": "00ASI",
+    "rev54": "00ASF",
+    "rev84": "00ASF",
+    "rle904": "00ASJ",
+    "rvlr75": "00WP4",
+    "rvlr77": "00WP4",
+    "s004s": "00WR3",
+    "s114s": "00WR3",
+    "st32s": "00WR3",
+    "trv01": "00H87",
+    "trv02": "00H87",
+    "us21s": "00WR3",
+    "us34s": "00WR3",
+    "utlmag01": "00AI5",
+    "utlmag02": "00AI5",
+    "utlmag03": "00AI5",
+    "utlmag04": "00AI5",
+    "walshr02": "00H90",
+}
+
+SCORE_TO_CODE_MAP = {
+    "evtg04": "001NN",
+    "eads66": "00WDC"
+}
+
+
+def safe_get(d, *keys):
+    """Attempts both snake_case and camelCase keys in a dict, handles None safely."""
+    if not isinstance(d, dict):
+        return {}
+    for key in keys:
+        if key in d:
+            return d[key]
+    return {}
+
+
+def extract_model_variables(creditBureau: dict) -> dict:
+    if not creditBureau:
+        return {}
+    variable_code_map = VARIABLE_TO_CODE_MAP
+    score_code_map = SCORE_TO_CODE_MAP
+    # Step 0: Extract application_id
+    extracted = {}
+
+    # Step 1: Locate add-on products with fallback casing
+    credit_bureau = safe_get(creditBureau, "creditBureau", "credit_bureau")
+
+    subject_node = safe_get(
+        safe_get(
+            safe_get(credit_bureau, "product"),
+            "subject"
+        ),
+        "subject_record", "subjectRecord"
+    )
+
+    add_ons = safe_get(subject_node, "add_on_product", "addOnProduct")
+    value_map = {}
+
+
+    # Step 2: Flatten values
+    for product in add_ons if isinstance(add_ons, list) else []:
+        code = product.get("code")
+        score_model = safe_get(product, "score_model", "scoreModel")
+
+        # 2a. Flatten characteristics (already done)
+        characteristics = score_model.get("characteristic", [])
+        for char in characteristics:
+            if isinstance(char, dict) and "id" in char and "value" in char:
+                key = char["id"].lower()
+                value_map[(code, key)] = char["value"]
+
+        # 2b. Also capture score result if present
+        score = safe_get(score_model, "score")
+        if score:
+            # Capture score result
+            results = score.get("results")
+            if isinstance(results, str):
+                value_map[(code, "score")] = results
+
+            # Capture top 4 score factors as list of {code, rank}
+            raw_factors = score.get("factors", {}).get("factor", [])
+            if not isinstance(raw_factors, list):
+                raw_factors = []
+
+            top_factors = [
+                {"code": f.get("code"), "rank": f.get("rank")}
+                for f in raw_factors[:4]
+                if isinstance(f, dict) and "code" in f and "rank" in f
+            ]
+            if top_factors:
+                value_map[(code, "factors")] = top_factors
+
+
+    # Step 3.a: Use variable_to_code_map to fetch final vars
+    for var, code in variable_code_map.items():
+        extracted[var] = value_map.get((code, var))
+    
+    # Step 3.b: Use score_to_code_map to fetch score vars
+    for key, code in SCORE_TO_CODE_MAP.items():
+        value = value_map.get((code, "score"))
+        if value:
+            extracted[key] = value.lstrip("+")
+
+        # Extract factor list, if available
+        factor_list = value_map.get((code, "factors"))
+        if factor_list:
+            extracted[f"{key}_factors"] = factor_list
+
+    return extracted
+
+if __name__ == "__main__":
+    for filename in sys.argv[1:]:
+        with open(filename) as f:
+            data = json.load(f)
+        print(f"\n--- Extracting from: {filename} ---")
+        result = extract_model_variables(data)
+        for k, v in result.items():
+            print(f"{k}: {v}")

post_processing.py

@@ -0,0 +1,253 @@
+import math
+
+EPD_REASON_MAP = {
+    "000": "No Adverse Factors",
+    "001": "Available Credit On Bankcard Accounts Is Too Low",
+    "006": "Bankcard Account Balances Are Too High In Proportion To Credit Limits",
+    "010": "Too Many Delinquencies",
+    "016": "Too Few Satisfactory Accounts",
+    "020": "Length Of Time Revolving Accounts Have Been Established Is Too Short",
+    "022": "Too many inquiries",
+    "023": "Months Since Most Recent Delinquency Is Too Short",
+    "024": "Too Many Serious Delinquencies",
+    "026": "Number Of Delinquent Accounts Is Too High In Proportion To Total Number Of Accounts",
+    "029": "Retail Account Balances Are Too High In Proportion To Credit Limits",
+    "030": "Not Enough Retail Debt Experience",
+    "031": "Revolving Account Balances Are Too High In Proportion To Credit Limits",
+    "035": "Length Of Time Accounts Have Been Established Is Too Short",
+    "037": "Too Few Bankcard Accounts",
+    "043": "Too Few Open Revolving Accounts",
+    "061": "Too Many Recently Opened Accounts",
+    "066": "Too Many Serious Derogatory Items",
+    "069": "Not Enough Debt Experience",
+    "070": "Length Of Time Since Most Recent Bankcard Account Has Been Established Is Too Short",
+    "074": "Too Few Satisfactory Revolving Accounts",
+    "076": "Total Amount Past Due Is Too High",
+    "103": "Not Enough Available Credit",
+    "105": "Too Few Revolving Accounts",
+    "117": "Length Of Time Since Most Seriously Delinquent Account Has Been Established Is Too Short",
+    "132": "Too Few Open Accounts",
+    "142": "Not Enough Balance Decreases On Active Non-Mortgage Accounts",
+    "146": "Recency Of A Balance Overlimit On A Bankcard Account",
+    "154": "Insufficient Payment Activity Over The Last Year",
+    "155": "Recency Of Max Aggregate Bankcard Balance Over The Last Year",
+    "158": "Too Few Open Retail Accounts",
+    "174": "Too Few Open Bankcard Accounts",
+    "181": "High Recent Balance Range Relative To Previous Balance Range",
+    "192": "Not Enough Available Credit On Revolving Accounts",
+    "201": "Length Of Time Since Oldest Auto Account Has Been Established Is Too Short"
+}
+
+VANTAGE_REASON_MAP = {
+    "10": "Too few accounts paid as agreed",
+    "11": "Oldest account was opened too recently",
+    "12": "Delinquent or derogatory status on accounts is too recent",
+    "13": "Balances on delinquent or derogatory accounts are too high",
+    "14": "Too high proportion of accounts recently opened",
+    "15": "Lack of recently reported accounts",
+    "16": "Total of credit limits and loan amounts is too low",
+    "17": "No open accounts in your credit file",
+    "18": "Lack of account information",
+    "19": "No negative reason code",
+    "20": "Delinquent or derogatory bankcard",
+    "21": "Too many bankcards with a high balance",
+    "22": "Too few bankcards with high credit limit",
+    "23": "Too high proportion of bankcards recently opened",
+    "24": "Too many bankcards with high balance compared to credit limit",
+    "25": "Too high proportion of balances from bankcards",
+    "26": "Balances on bankcards are too high",
+    "27": "Delinquent or derogatory status on revolving accounts is too recent",
+    "28": "Average credit limit on open bankcards is too low",
+    "29": "Balances on bankcards are too high compared with credit limits",
+    "30": "Too few open revolving accounts",
+    "31": "Not enough available credit on revolving accounts",
+    "32": "Oldest bankcard was opened too recently",
+    "33": "Not enough balance paid down over time on bankcards",
+    "34": "Most recently opened revolving account is too new",
+    "35": "Lack of revolving account information",
+    "36": "Lack of recently reported revolving accounts",
+    "37": "No open bankcards in your credit file",
+    "38": "Lack of bankcard account information",
+    "39": "Balances on delinquent or derogatory bankcards are too high",
+    "4": "Balances on accts too high compared to credit limits and loan amounts",
+    "04": "Balances on accts too high compared to credit limits and loan amounts",
+    "40": "Too many delinquent or derogatory revolving accounts",
+    "41": "Average time since revolving accounts opened is too recent",
+    "42": "Total credit limits on open revolving accounts are too low",
+    "43": "Too many revolving accounts with high balance compared to credit limit",
+    "44": "Balances on revolving accts are too high compared with credit limits",
+    "45": "Not enough balance paid down over time on retail accounts",
+    "46": "Oldest revolving account was opened too recently",
+    "47": "No open retail accounts in your credit file",
+    "48": "Lack of retail account information",
+    "49": "Not enough balance paid down over time on revolving accounts",
+    "5": "Too many recent delinquencies",
+    "05": "Too many recent delinquencies",
+    "50": "Balances on personal installment accts too high compared to loan amts",
+    "51": "Too few installment accounts recently paid as agreed",
+    "52": "Delinquent or derogatory installment account",
+    "53": "Not enough balance paid down over time on installment accounts",
+    "54": "Delinquent or derogatory status on installment accounts is too recent",
+    "55": "Lack of recently reported auto accounts",
+    "56": "Lack of recently reported installment accounts",
+    "57": "No open installment accounts in your credit file",
+    "58": "Lack of installment account information",
+    "59": "Balances on retail cards are too high compared with credit limits",
+    "6": "Too many accounts recently opened",
+    "06": "Too many accounts recently opened",
+    "60": "Total delinquent or derogatory balances on real estate loans too high",
+    "61": "No open first mortgage accounts in your credit file",
+    "62": "Lack of first mortgage account information",
+    "63": "Delinquent or derogatory real estate secured loan",
+    "64": "Not enough balance paid down over time on real estate secured loans",
+    "65": "Oldest real estate secured loan was opened too recently",
+    "66": "Delinquent or derogatory status on real estate loans is too recent",
+    "67": "No open real estate secured loans in your credit file",
+    "68": "Lack of real estate secured loan information",
+    "69": "Too high proportion of balances from loans not secured by real estate",
+    "7": "You have too many delinquent or derogatory accounts",
+    "07": "You have too many delinquent or derogatory accounts",
+    "70": "Too high proportion of auto accounts are delinquent or derogatory",
+    "71": "Not enough balance paid down over time on auto accounts",
+    "72": "Too few auto accounts paid as agreed",
+    "73": "Delinquent or derogatory auto account",
+    "74": "Balances on auto accounts are too high compared with loan amounts",
+    "75": "Payments on auto accounts less than scheduled amount",
+    "76": "Delinquent or derogatory status on auto accounts is too recent",
+    "77": "No open auto accounts in your credit file",
+    "78": "Lack of auto account information",
+    "79": "No negative reason code",
+    "8": "Too few accounts recently paid as agreed",
+    "08": "Too few accounts recently paid as agreed",
+    "80": "Delinquent or derogatory student loan",
+    "81": "Not enough balance paid down over time on student loans",
+    "82": "Lack of recently reported student loans",
+    "83": "No negative reason code",
+    "84": "Number of inquiries was a factor in determining the score",
+    "85": "Too many inquiries",
+    "86": "Derogatory public records",
+    "87": "Unpaid collections",
+    "88": "Bankruptcy",
+    "89": "No negative reason code",
+    "9": "Delinquent or derogatory account",
+    "09": "Delinquent or derogatory account",
+    "90": "No open revolving accounts in your credit file",
+    "91": "Balances on delinquent or derogatory revolving accounts are too high",
+    "92": "Delinquent or derogatory first mortgage",
+    "93": "Not enough balance paid down over time on first mortgage accounts",
+    "94": "No negative reason code",
+    "95": "No negative reason code",
+    "96": "Too few open accounts",
+    "97": "Too few accounts"
+}
+
+REASON_MAP = {
+    'evtg04':   "System Generated",
+    'eads66':   "System Generated",
+    's004s':    "Length of time on file is too short",
+    'mt34s':    "Not enough balance decreases on mortgage trades in the past 12 months",
+    'ct320':    "Insufficient payment activity",
+    'us21s':    "Length of time since most recent installment account has been established is too short",
+    'utlmag02': "Revolving account balances are too high in proportion to credit limits over the last 24 months",
+    'trv01':    "Recency of a balance overlimit on a bankcard account",
+    'us34s':    "Not enough balance decreases on installment trades in the past 12 months"
+}
+
+
+def generate_reason_codes(score_key, factors):
+    # fallback to 4 null rows if no factors found
+    if not isinstance(factors, list) or len(factors) == 0:
+        return [{"code": None, "rank": None, "description": None} for _ in range(4)]
+
+    reason_map = VANTAGE_REASON_MAP if score_key == "evtg04" else EPD_REASON_MAP if score_key == "eads66" else {}
+
+    reason_codes = []
+    for f in factors[:4]:
+        code = f.get("code")
+        rank = f.get("rank")
+        description = reason_map.get(str(code), "")
+        reason_codes.append({
+            "code": code,
+            "rank": rank,
+            "description": description
+        })
+
+    # pad to 4
+    while len(reason_codes) < 4:
+        reason_codes.append({"code": None, "rank": None, "description": None})
+
+    return reason_codes
+
+
+def post_processing(processing_output, record):
+    prediction = processing_output["prediction"]
+    shape_reasoncode = processing_output["shape_reasoncode"]
+
+    # grade mapping:
+    if prediction < 0:
+        grade = "M14"
+    else:
+        m = math.ceil(prediction / 0.01)
+        m = max(m, 1)
+        m = min(m, 14)
+        grade = f"M{m}"
+
+    # if prediction ≤ 0.04, not declined
+    if prediction <= 0.04:
+        return {
+            "grade": grade,
+            "reason_codes": [{
+                "code": None,
+                "rank": None,
+                "description": None
+            }]
+        }
+
+    conditions = {
+        'evtg04': lambda x: x < 700,
+        'eads66': lambda x: x < 700,
+        's004s': lambda x: x < 12,
+        'mt34s': lambda x: x > 95,
+        'ct320': lambda x: x <= 3,
+        'us21s': lambda x: x <= 3,
+        'utlmag02': lambda x: x > 300,
+        'trv01': lambda x: x <= 3,
+        'us34s': lambda x: x > 90
+    }
+
+    for item in shape_reasoncode:
+        feat = item.get("feature")
+        val = item.get("value")
+        cond = conditions.get(feat)
+
+        if cond:
+            try:
+                if cond(val):
+                    # If score-type feature (evtg04 or eads66) → full factors-based reason
+                    if feat in ("evtg04", "eads66"):
+                        return {
+                            "grade": grade,
+                            "reason_codes": generate_reason_codes(feat, record.get(f"{feat}_factors", []))
+                        }
+                    else:
+                        # Other features → only 1 reason code based on REASON_MAP
+                        return {
+                            "grade": grade,
+                            "reason_codes": [{
+                                "code": feat,
+                                "rank": "1",
+                                "description": REASON_MAP.get(feat, "Reason not mapped")
+                            }]
+                        }
+            except Exception:
+                continue
+
+    # Default fallback
+    return {
+        "grade": grade,
+        "reason_codes": [{
+            "code": feat,
+            "rank": "1",
+            "description": "No suitable Product Offerings found"
+        }]
+    }

pre_processing.py

@@ -0,0 +1,360 @@
+import math
+
+
+lookup_dict = {
+  "balmag04": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 600.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 427.0
+  },
+  "utlmag01": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 600.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 446.0
+  },
+  "utlmag02": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 600.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 486.0
+  },
+  "utlmag03": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 600.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 492.0
+  },
+  "utlmag04": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 600.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 414.0
+  },
+  "mnpmag03": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 600.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 494.0
+  },
+  "duemag01": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 600.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 488.0
+  },
+  "trv01": {
+    "data_type": "int", "valid_min": 1.0, "valid_max": 24.0, "default_treatment_type": "unk", "observed_cap_min_value": 1.0, "observed_cap_max_value": 24.0
+  },
+  "trv02": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 12.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 12.0
+  },
+  "index01": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 999.0
+  },
+  "index02": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 999.0
+  },
+  "rvlr75": {
+    "data_type": "float", "valid_min": 0.0, "valid_max": 5.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 5.0
+  },
+  "rvlr77": {
+    "data_type": "float", "valid_min": 0.0, "valid_max": 5.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 5.0
+  },
+  "rev12": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 109.0
+  },
+  "rev13": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 82.0
+  },
+  "rev14": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 56.0
+  },
+  "rev54": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 9.0
+  },
+  "rev84": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 23.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 23.0
+  },
+  "bkc14": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 47.0
+  },
+  "bkc84": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 23.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 23.0
+  },
+  "ret12": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 42.0
+  },
+  "evtg04": {
+    "data_type": "int", "valid_min": 300.0, "valid_max": 850.0, "default_treatment_type": "unk", "observed_cap_min_value": 300.0, "observed_cap_max_value": 845.0
+  },
+  "rev201": {
+    "data_type": "float", "valid_min": 0.0, "valid_max": 50.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 50.0
+  },
+  "rev202": {
+    "data_type": "float", "valid_min": 0.0, "valid_max": 50.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 50.0
+  },
+  "rev203": {
+    "data_type": "float", "valid_min": 0.0, "valid_max": 50.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 50.0
+  },
+  "rev223": {
+    "data_type": "float", "valid_min": 0.0, "valid_max": 50.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 49.0
+  },
+  "rev224": {
+    "data_type": "float", "valid_min": 0.0, "valid_max": 50.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 47.0
+  },
+  "rev225": {
+    "data_type": "float", "valid_min": 0.0, "valid_max": 50.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 45.0
+  },
+  "walshr02": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 2.0
+  },
+  "rev231": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 18101.0
+  },
+  "rev232": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 38963.0
+  },
+  "rev233": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 68709.0
+  },
+  "rev252": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 13988.0
+  },
+  "rev253": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 12347.0
+  },
+  "all231": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 38949.0
+  },
+  "at28a": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 1330166.0
+  },
+  "at28b": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 351273.0
+  },
+  "at36s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 999.0
+  },
+  "bc20s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 510.0
+  },
+  "bc21s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 190.0
+  },
+  "bc28s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 205768.0
+  },
+  "bc36s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 999.0
+  },
+  "bc97a": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 184747.0
+  },
+  "bc98a": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 186395.0
+  },
+  "bc102s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 31000.0
+  },
+  "bc104s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 99.0
+  },
+  "bc107s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 15.0
+  },
+  "br20s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 495.0
+  },
+  "fi21s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 162.0
+  },
+  "fi34s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 100.0
+  },
+  "fi35s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 70016.0
+  },
+  "g051s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 50.0
+  },
+  "g102s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 23.0
+  },
+  "g105s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 23.0
+  },
+  "g201a": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 255142.0
+  },
+  "g221d": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 1.0
+  },
+  "g232s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 21.0
+  },
+  "g250a": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 13.0
+  },
+  "g960s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 17.0
+  },
+  "g990s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 10.0
+  },
+  "mt20s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 261.0
+  },
+  "mt34s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 101.0
+  },
+  "pb28s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 192100.0
+  },
+  "pb34s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 100.0
+  },
+  "re28s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 311518.0
+  },
+  "re36s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 999.0
+  },
+  "re102s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 41750.0
+  },
+  "s004s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 179.0
+  },
+  "s114s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 5.0
+  },
+  "st32s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 221686.0
+  },
+  "g106s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 590.0
+  },
+  "g242b": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 3.0
+  },
+  "us21s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 153.0
+  },
+  "us34s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 101.0
+  },
+  "g403s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 5.0
+  },
+  "g405s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 3.0
+  },
+  "g408s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 3.0
+  },
+  "g411s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 8.0
+  },
+  "g416s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 23.0
+  },
+  "g417s": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 23.0
+  },
+  "agg901": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 3.0
+  },
+  "agg902": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 3.0
+  },
+  "agg908": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 69376.0
+  },
+  "agg911": {
+    "data_type": "float", "valid_min": 0.0, "valid_max": 10000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 107.0
+  },
+  "rle904": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 309634.0
+  },
+  "p02d": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 99.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 3.0
+  },
+  "p02h": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 99.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 7.0
+  },
+  "balmag01": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 600.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 474.0
+  },
+  "balmag02": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 600.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 516.0
+  },
+  "cv13": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 100.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 50.0
+  },
+  "cv17": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 9.0
+  },
+  "cv21": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 171153.0
+  },
+  "cv25": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 100.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 33.0
+  },
+  "ct319": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 98.0
+  },
+  "ct320": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 52.0
+  },
+  "cta11": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 71.0
+  },
+  "cta20": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 45.0
+  },
+  "cta21": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 999.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 148.0
+  },
+  "paymnt08": {
+    "data_type": "float", "valid_min": 0.0, "valid_max": 50.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 50.0
+  },
+  "rev321": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 234408.0
+  },
+  "rev322": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 219698.0
+  },
+  "bkc321": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 171024.0
+  },
+  "bkc322": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 162590.0
+  },
+  "bkc323": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 152068.0
+  },
+  "bkc324": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 136763.0
+  },
+  "bkc327": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 171143.0
+  },
+  "bkc328": {
+    "data_type": "int", "valid_min": 0.0, "valid_max": 1000000000.0, "default_treatment_type": "unk", "observed_cap_min_value": 0.0, "observed_cap_max_value": 164352.0
+  },
+  "eads66": {
+    "data_type": "float", "valid_min": 300.0, "valid_max": 850.0, "default_treatment_type": "unk", "observed_cap_min_value": 300.0, "observed_cap_max_value": 820.0
+  }
+}
+
+
+# 1. Pre-processing: type casting and variable treatment
+def pre_processing(input_dict):
+    processed = {}
+    for var, cfg in lookup_dict.items():
+        val = input_dict.get(var)
+        # Cast to required type
+        try:
+            if cfg.get("data_type") == "int":
+                val = int(val)
+            elif cfg.get("data_type") == "float":
+                val = float(val)
+        except (ValueError, TypeError):
+            val = None
+        # Variable treatment
+        vmin = cfg.get("valid_min")
+        vmax = cfg.get("valid_max")
+        if val is not None and not (math.isnan(vmin) or math.isnan(vmax)):
+            if cfg.get("default_treatment_type") == "unk":
+                if val < vmin or val > vmax:
+                    val = float("nan")
+            cmin = cfg.get("observed_cap_min_value")
+            cmax = cfg.get("observed_cap_max_value")
+            if not math.isnan(val):
+                if cmin is not None and vmin <= val < cmin:
+                    val = cmin
+                if cmax is not None and cmax < val <= vmax:
+                    val = cmax
+        processed[var] = val
+    return processed

processing.py

@@ -0,0 +1,39 @@
+import joblib
+import xgboost as xgb
+import shap
+
+def processing(processed_dict):
+    # model_path = r"C:\Users\abinisha\citrus\centurion-dataform-mcard\flowx\model_1\artifacts\xgboost_model_v2.joblib"
+    model_path = "./xgboost_model.joblib"
+    model       = joblib.load(model_path)
+    explainer   = shap.TreeExplainer(model)
+    feature_names = model.feature_names
+
+    # build and score
+    X       = [processed_dict[f] for f in feature_names]
+    dmatrix = xgb.DMatrix([X], feature_names=feature_names)
+    pred    = float(model.predict(dmatrix)[0])
+
+    # get SHAP values as 1×n_features
+    shap_matrix = explainer.shap_values(dmatrix)
+    shap_vals   = shap_matrix[0] if getattr(shap_matrix, "ndim", 1) > 1 else shap_matrix
+
+    # rank & sort features by descending SHAP impact
+    ranked = sorted(zip(feature_names, shap_vals),
+                    key=lambda fv: fv[1],
+                    reverse=True)
+
+    # build a *list* of pre-sorted reason-codes
+    shape_reasoncode = [
+        {
+            "feature":   feat,
+            "value":     processed_dict[feat],
+            "shap_rank": i+1
+        }
+        for i, (feat, _) in enumerate(ranked)
+    ]
+
+    return {
+        "prediction":       pred,
+        "shape_reasoncode": shape_reasoncode
+    }

request_schema.json

@@ -1 +1,14 @@
-{}
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "type": "object",
+  "properties": {
+    "application_id": {
+      "type": ["string", "null"],
+      "description": "Unique identifier for the application."
+    },
+    "creditBureau": {
+      "type": ["object", "null"],
+      "description": "Credit bureau details"
+    }
+  }
+}

requirements.txt

@@ -1 +1,3 @@
-{}
+joblib == 1.3.2
+xgboost == 2.1.4
+shap == 0.46.0

response_schema.json

@@ -1 +1,48 @@
-{}
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "type": "object",
+  "properties": {
+    "application_id": {
+      "type": "string",
+      "description": "Unique identifier for the application"
+    },
+    "prediction": {
+      "type": "number",
+      "description": "Model's predicted probability score"
+    },
+    "grade": {
+      "type": "string",
+      "description": "HD Model Grade"
+    },
+    "reason_codes": {
+      "type": "array",
+      "description": "List of reason codes explaining the model decision",
+      "items": {
+        "type": "object",
+        "properties": {
+          "code": {
+            "type": ["string", "null"],
+            "description": "Feature or score reason code"
+          },
+          "rank": {
+            "type": ["string", "null"],
+            "description": "Rank of importance (1 to 4)"
+          },
+          "description": {
+            "type": ["string", "null"],
+            "description": "Human-readable explanation for the reason"
+          }
+        },
+        "required": ["code", "rank", "description"]
+      },
+      "minItems": 1,
+      "maxItems": 4
+    }
+  },
+  "required": [
+    "application_id",
+    "prediction",
+    "grade",
+    "reason_codes"
+  ]
+}