Atualizar API

api/app.py

@@ -1,4 +1,4 @@
-# python -m uvicorn app:app --reload
+# python -m uvicorn main:app --reload
 
 from __future__ import annotations
 from typing import Any, Dict, List, Optional
@@ -18,16 +18,9 @@ FEATHER_DATASET = os.path.join(ROOT_DIR, "data", "final_dataset.feather")
 PARQUET_DATASET = os.path.join(ROOT_DIR, "data", "final_dataset.parquet")
 
 DROP_COLS = {
-	"tx_year",
-	"tx_month",
-	"periodo",
-	"terminal_soft_descriptor",
 	"card_hash",
-	"card_bin",
-	"is_transactional_fraud",
+	"terminal_id",
 	"is_fraud",
-	"cluster",
-	"fraude_tipo_extendido",
 }
 
 
@@ -43,28 +36,20 @@ _MODEL = None
 _FEATURES: Optional[List[str]] = None
 _CARD_MEDIANS: Dict[str, Dict[str, float]] = {}
 _TERM_MEDIANS: Dict[str, Dict[str, float]] = {}
+_CARD_IDS: Optional[set] = None
+_TERM_IDS: Optional[set] = None
 
 
-FRAUD_TYPE_MAP = {
-	0: ("c0", "não é fraude"),
-	1: ("c1", "fraude em cartão"),
-	2: ("c2", "desacordo comercial"),
-	3: ("c3", "fraude no terminal"),
-	4: ("c4", "conluio"),
-}
-
 CARD_FEATURES = {
 	"dias_desde_primeira_transacao_do_cartao",
 	"qtas_transacoes_cartao_dia",
 	"qtas_fraudes_cartao",
 	"valor_medio_cartao",
-	"valor_medio_cartao_3_transacoes",
 	"desvio_padrao_valor_cartao",
 	"entropia_geografica_cartao",
-	"frequencia_transacoes_24h",
-	"media_tempo_entre_transacoes",
 	"fraude_ratio_cartao",
 	"tempo_medio_denuncia_cartao",
+	"velocidade_ultima_transacao",
 }
 
 TERMINAL_FEATURES = {
@@ -72,7 +57,7 @@ TERMINAL_FEATURES = {
 	"qtas_transacoes_terminal_dia",
 	"qtas_fraudes_terminal",
 	"valor_medio_terminal",
-	"media_valor_terminal_semana",
+	"desvio_padrao_valor_terminal",
 	"fraude_ratio_terminal",
 	"tempo_medio_denuncia_terminal",
 }
@@ -84,23 +69,17 @@ def _predict(ensemble, X: pd.DataFrame) -> Dict[str, Any]:
 	items: List[Dict[str, Any]] = []
 	for i in range(len(X)):
 		pred_class = int(y_pred[i])
-		is_fraud = bool(pred_class != 0)
-		probs = None
+		is_fraud = bool(pred_class == 1)
+		prob_fraud = None
 		if y_prob is not None:
 			pp = y_prob[i]
-			probs = [float(p) for p in pp]
-		code_name = FRAUD_TYPE_MAP.get(pred_class)
-		fraud_code = code_name[0] if code_name else None
-		fraud_label = code_name[1] if code_name else None
-		row = X.iloc[i]
-		debug = {c: (float(row[c]) if pd.notna(row[c]) else None) for c in X.columns}
+			if isinstance(pp, (list, np.ndarray)) and len(pp) >= 2:
+				prob_fraud = float(pp[1])
+			else:
+				prob_fraud = float(pp)
 		items.append({
-			"predicted_class": pred_class,
 			"is_fraud": bool(is_fraud),
-			"fraud_type": fraud_code if is_fraud else None,
-			"fraud_type_name": fraud_label if is_fraud else None,
-			"class_probabilities": probs,
-			"_debug_processed_features": debug,
+			"fraud_probability": prob_fraud,
 		})
 	return {"items": items}
 
@@ -198,20 +177,26 @@ def _ensure_dataframe(records: List[Dict[str, Any]], feature_order: List[str]) -
 	return df
 
 
+def _ensure_id_sets():
+	global _CARD_IDS, _TERM_IDS
+	if _CARD_IDS is not None and _TERM_IDS is not None:
+		return
+	df_ids = _load_dataset(["card_hash", "terminal_id"])
+	if "card_hash" in df_ids.columns:
+		_CARD_IDS = set(df_ids["card_hash"].astype(str).dropna().unique())
+	else:
+		_CARD_IDS = set()
+	if "terminal_id" in df_ids.columns:
+		_TERM_IDS = set(df_ids["terminal_id"].astype(str).dropna().unique())
+	else:
+		_TERM_IDS = set()
+
+
 @app.get("/health")
 def health():
 	return {"status": "ok"}
 
 
-@app.get("/")
-def root():
-	return {
-		"status": "ok",
-		"health": "/health",
-		"docs": "/docs",
-	}
-
-
 @app.post("/predict")
 def predict_one(body: TransactionBody, request: Request):
 	try:
@@ -242,3 +227,19 @@ def predict_batch(body: BatchBody, request: Request):
 	except Exception as e:
 		traceback.print_exc()
 		raise HTTPException(status_code=500, detail=str(e))
+
+
+@app.get("/ids/exists")
+def ids_exists(terminal_id: Optional[str] = None, card_hash: Optional[str] = None):
+	try:
+		_ensure_id_sets()
+		term_ok = False
+		card_ok = False
+		if terminal_id is not None and _TERM_IDS is not None:
+			term_ok = str(terminal_id) in _TERM_IDS
+		if card_hash is not None and _CARD_IDS is not None:
+			card_ok = str(card_hash) in _CARD_IDS
+		return {"terminal_id_exists": term_ok, "card_hash_exists": card_ok}
+	except Exception as e:
+		traceback.print_exc()
+		raise HTTPException(status_code=500, detail=str(e))

data/final_dataset.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:880f13b2180b1ebc1845cc090f6be5fc8c4a5f08c478791a71f2af9b6315ad90
+size 246272480