Merge pull request #30 from Diaure/feature/db_schema_setup

App/database.py

@@ -1,20 +1,30 @@
 import os
 from dotenv import load_dotenv
 from sqlalchemy import create_engine
-from sqlalchemy.orm import sessionmaker
+from sqlalchemy.orm import sessionmaker, declarative_base
 
 load_dotenv()
 
-DB_USER = os.getenv("DB_USER")
-DB_PASSWORD = os.getenv("DB_PASSWORD")
-DB_HOST = os.getenv("DB_HOST")
-DB_PORT = os.getenv("DB_PORT")
-DB_NAME = os.getenv("DB_NAME")
+# Détection si on est en CI (GitHub Actions) ou en test 
+IS_CI = os.getenv("CI") == "true"
+IS_PYTEST = "pytest" in os.getenv("PYTHONPATH", "") or os.getenv("PYTEST_CURRENT_TEST") is not None 
 
-DATABASE_URL = (
-    f"postgresql+psycopg2://{DB_USER}:{DB_PASSWORD}"
-    f"@{DB_HOST}:{DB_PORT}/{DB_NAME}"
-)
+SKIP_DB = IS_CI or IS_PYTEST
 
-engine = create_engine(DATABASE_URL)
-SessionLocal = sessionmaker(bind=engine)
+DB_USER = os.getenv("DB_USER", "postgres")
+DB_PASSWORD = os.getenv("DB_PASSWORD", "password")
+DB_HOST = os.getenv("DB_HOST", "localhost")
+DB_PORT = os.getenv("DB_PORT", "5432")
+DB_NAME = os.getenv("DB_NAME", "test_db")
+
+DATABASE_URL = (f"postgresql+psycopg2://{DB_USER}:{DB_PASSWORD}"f"@{DB_HOST}:{DB_PORT}/{DB_NAME}")
+
+Base = declarative_base()
+
+if not SKIP_DB:
+    engine = create_engine(DATABASE_URL)
+    SessionLocal = sessionmaker(autocommit = False, autoflush = False, bind = engine)
+
+else: 
+    engine = None 
+    SessionLocal = None

App/model.py

@@ -0,0 +1,51 @@
+from sqlalchemy import Column, Integer, String, Float, Boolean, DateTime, ForeignKey
+from sqlalchemy.sql import func
+from App.database import Base
+
+class Input(Base):
+    __tablename__ = "inputs"
+
+    id = Column(Integer, primary_key=True, index=True)
+    genre = Column(String)
+    statut_marital = Column(String)
+    departement = Column(String)
+    poste = Column(String)
+    domaine_etude = Column(String)
+    frequence_deplacement = Column(String)
+    heure_supplementaires = Column(Boolean)
+    evolution_cat_evol = Column(String)
+    categorie_employe = Column(String)
+    satisfaction_employee_nature_travail = Column(Integer)
+    nombre_participation_pee = Column(Integer)
+    ecart_note_evaluation = Column(Integer)
+    revenu_mensuel = Column(Integer)
+    distance_domicile_travail = Column(Integer)
+    satisfaction_globale = Column(Float)
+    niveau_education = Column(Integer)
+    note_evaluation_actuelle = Column(Integer)
+    satisfaction_employee_equipe = Column(Integer)
+    age = Column(Integer)
+    revenu_par_annee_experience_interne = Column(Integer)
+    satisfaction_employee_equilibre_pro_perso = Column(Integer)
+    nombre_experiences_precedentes = Column(Integer)
+    annees_dans_l_entreprise = Column(Integer)
+    nb_formations_suivies = Column(Integer)
+    revenu_par_annee_experience_totale = Column(Integer)
+    ratio_sans_promotion = Column(Integer)
+    satisfaction_employee_environnement = Column(Integer)
+    exp_hors_entreprise = Column(Integer)
+    mobilite_promotion = Column(Integer)
+    annees_depuis_la_derniere_promotion = Column(Integer)
+
+    created_at = Column(DateTime(timezone=True), server_default=func.now())
+
+class Predictions(Base):
+    __tablename__ = "predictions"
+    id = Column(Integer, primary_key=True, index=True)
+    input_id = Column(Integer, ForeignKey("inputs.id"))
+
+    prediction_label = Column(String)
+    prediction_proba = Column(Float)
+    model_version = Column(String)
+
+    created_at = Column(DateTime(timezone=True), server_default=func.now())

App/predict.py

@@ -4,6 +4,9 @@ from App.schemas import EmployeeFeatures
 import json
 from pathlib import Path
 from huggingface_hub import hf_hub_download
+from sqlalchemy.orm import Session
+from App.database import SessionLocal
+from App.model import Input, Predictions
 
 MODEL_REPO = "Diaure/xgb_model"
 
@@ -43,7 +46,29 @@ def predict_employee(data: dict):
     pred = model.predict(df)[0]
     proba = model.predict_proba(df)[0][1]
 
+    db: Session = SessionLocal() if SessionLocal is not None else None
+
+    if db is not None:
+        try:
+            # enregistrer les inputs: à chaque appel de POST/predict, on stocke d'abord les entrées de l'utilisateur
+            input_row = Input(**data) 
+            db.add(input_row) 
+            db.commit() 
+            db.refresh(input_row)
+
+            # puis on récupère les ids générés automatiquement et enregistre les prédictions liés aux ids
+            pred_row = Predictions(input_id = input_row.id, prediction_label = classes_mapping[str(pred)], prediction_proba = float(proba), model_version = "v1") 
+            db.add(pred_row) 
+            db.commit()
+        
+        except Exception as e: 
+            print("🔥 ERREUR DB :", e) 
+            raise e
+
+        finally:
+            db.close()
+
+    # puis on renvoie la réponse API
     return {
         "Prediction": classes_mapping[str(pred)],
-        "Probabilite_depart": float(proba)
-    }
+        "Probabilite_depart": float(proba)}

App/schemas.py

@@ -10,7 +10,6 @@ class EmployeeFeatures(BaseModel):
     heure_supplementaires: bool
     evolution_cat_evol: str
     categorie_employe: str
-
     satisfaction_employee_nature_travail: int
     nombre_participation_pee: int
     ecart_note_evaluation: int

README.md

@@ -15,9 +15,7 @@ pinned: false
 opérationnels et accessibles via une API performante.
 
 Ce projet correspond à un **Proof of Concept (POC)** visant à déployer un modèle de machine
-learning en production en appliquant les bonnes pratiques d’ingénierie logicielle :
-versionnage, tests, base de données et automatisation.
-
+learning en production en appliquant les bonnes pratiques d’ingénierie logicielle: versionnage, tests, base de données et automatisation.
 
 
 ## Objectifs du projet
@@ -29,7 +27,7 @@ versionnage, tests, base de données et automatisation.
 
 
 ## Périmètre fonctionnel
-Le projet inclut :
+Le projet inclut:
 - Une API développée avec **FastAPI**
 - L’exposition d’un modèle de machine learning via des endpoints REST
 - Une base de données **PostgreSQL** pour stocker les entrées/sorties du modèle
@@ -91,7 +89,7 @@ L’API est déployée publiquement sur Hugging Face Spaces.
 
 - URL de l’API :
 https://diaure-futurisys-ml-api.hf.space
-- Documentation interactive (Swagger UI) :
+- Documentation interactive (Swagger UI):
 https://diaure-futurisys-ml-api.hf.space/docs. Ele permet de:
   - visualiser les endpoints
   - tester directement l’endpoint `/predict`
@@ -100,12 +98,12 @@ https://diaure-futurisys-ml-api.hf.space/docs. Ele permet de:
 ### `Endpoint principal`
 `POST /predict`
 
-Cet endpoint reçoit les caractéristiques d’un employé et retourne :
+Cet endpoint reçoit les caractéristiques d’un employé et retourne:
 
 - une prédiction lisible ("Reste" ou "Part")
 - la probabilité associée au départ
 
-Exemple de réponse :
+Exemple de réponse:
 ```json
 {
   "Prediction": "Part",
@@ -121,20 +119,115 @@ L’API expose un endpoint principal de prédiction.
 
 **POST /predict**
   - Description : retourne une prédiction de départ d’un employé
-  - Validation des données : Pydantic
-  - Réponses possibles :
-    - 200 : prédiction valide
-    - 422 : données invalides
+  - Validation des données: Pydantic
+  - Réponses possibles:
+    - 200: prédiction valide
+    - 422: données invalides
 
-## Stack technique
-- **Langage** : Python
-- **API** : FastAPI
-- **Machine Learning** : scikit-learn
-- **Base de données** : PostgreSQL
-- **Tests** : Pytest, pytest-cov
-- **CI/CD** : GitHub Actions, Hugging Face
-- **Versionnage** : Git / GitHub
+## Base de données et traçabilité des prédictions
+### `Objectifs`
+
+L’intégration d’une base de données PostgreSQL permet d’inscrire le projet dans une logique MLOps et de répondre à plusieurs objectifs clés:
+- assurer la traçabilité complète des prédictions du modèle
+- conserver l’historique des données d’entrée utilisateur
+- stocker les résultats de prédiction (label, probabilité, version du modèle)
+- préparer une architecture compatible avec un déploiement en production.
+
+### `Méthodologie utilisée`
+- **PostgreSQL** a été retenu pour:
+  - sa robustesse et sa fiabilité
+  - sa compatibilité native avec SQLAlchemy
+  - son usage courant en environnement professionnel
+
+- **SQLAlchemy** est utilisé comme couche d’abstraction:
+  - gestion centralisée de la connexion à la base
+  - cohérence entre le schéma Python et la base SQL
+
+Les identifiants de connexion sont stockés dans des variables d’environnement (`.env`) afin d’éviter toute exposition de secrets dans le dépôt Git.
+
+### `Modélisation de la base de données`
+La base de données repose sur trois tables distinctes, chacune ayant un rôle précis.
+1. `employees_dataset - Dataset de référence`
+Il contient le dataset final nettoyé et préparé lors de l'entraînement du modèle en incluant l'ensemble des **32 deatures** du modèle. Il sert de:
+  - référence de schéma
+  - source de validation
+  - base documentaire du modèle
+
+C'est une table qui n'est jamais alimentée par l'utilisateur.
+
+```python
+load_dotenv()
+
+BASE_DIR = os.path.dirname(os.path.abspath(__file__))
+csv_path = os.path.join(BASE_DIR, "dataset_final.csv")
+
+df = pd.read_csv(csv_path, encoding="latin-1")
 
+DB_USER = os.getenv("DB_USER")
+DB_PASSWORD = os.getenv("DB_PASSWORD")
+DB_HOST = os.getenv("DB_HOST")
+DB_PORT = os.getenv("DB_PORT")
+DB_NAME = os.getenv("DB_NAME")
+
+DATABASE_URL = (f"postgresql+psycopg2://{DB_USER}:{DB_PASSWORD}"f"@{DB_HOST}:{DB_PORT}/{DB_NAME}")
+
+engine = create_engine(DATABASE_URL)
+
+df.to_sql("employees_dataset", engine, if_exists="replace", index=False)
+```
+
+2. `inputs - Entrées utilisateur`
+  - Enregistre chaque requête utilisateur envoyée à l'endpoint `/predict`
+  - Contient exactement les features attendues par le modèle
+  - Structure strictement alignée avec le schéma Pydandic(`EmployeeFeatures`)
+  - Permet:
+    - l'audit des predictions
+    - l'analyse à posteriori
+    - la reproductibilité des résultats.
+```python
+class Input(Base):
+    __tablename__ = "inputs"
+
+    id = Column(Integer, primary_key=True, index=True)
+    genre = Column(String)
+    statut_marital = Column(String)
+    departement = Column(String)
+    poste = Column(String)
+```
+
+3. `predictions - Résultats du modèle`
+  - Continet:
+    - le label de prédiction
+    - la probabilité associée
+  - Reliée à `inputs` via une clé étrangère
+  - Garantit une trçabilité complète.
+```python
+class Predictions(Base):
+    __tablename__ = "predictions"
+    id = Column(Integer, primary_key=True, index=True)
+    input_id = Column(Integer, ForeignKey("inputs.id"))
+
+    prediction_label = Column(String)
+    prediction_proba = Column(Float)
+    model_version = Column(String)
+```
+
+### `Interaction API <> Base de données`
+Lors d’un appel à l’endpoint `POST /predict`:
+- les données utilisateur sont validées via **Pydantic**
+- les entrées sont enregistrées dans la table **inputs**
+- le modèle est exécuté
+- la prédiction est enregistrée dans la table **predictions**
+- la réponse est retournée à l’utilisateur.
+
+## Stack technique
+- **Langage**: Python
+- **API**: FastAPI
+- **Machine Learning**: scikit-learn
+- **Base de données**: PostgreSQL
+- **Tests**: Pytest, pytest-cov
+- **CI/CD**: GitHub Actions
+- **Versionnage**: Git / GitHub
 
 
 ## Structure du projet
@@ -143,7 +236,9 @@ futurisys_ml-api/
 ├── github/workflows
 │   ├── ci.yml       # Description des évènement déclenchants des tests
 ├── app/             # Code applicatif principal
+│   ├── database.py  # Point de connexion à la base PostgreSQL
 │   ├── main.py      # Point d’entrée de l’API
+│   ├── model.py     # Définition des tables de la database
 │   ├── predict.py   # Application du modèle
 │   ├── schemas.py   # Validation des données (Pydantic)
 │   ── model/                            # Elements du modèle
@@ -151,10 +246,14 @@ futurisys_ml-api/
 │   ├── modele_final_xgb.joblib          # Modèle final avec hyperparamètres
 │   ├── preprocesseur_fitted.joblib      # Pipeline entrainé
 |
-├── scripts/         # Scripts bd (BD, données)
-├── tests/           # Tests unitaires, fonctionnels
-│   ├── test_api.py  # Test automatisé API Pytest
+├── scripts/                   # Scripts bd (BD, données)
+│   ├── create_tables.py       # Créaton des tables définies dans model.py
+│   ├── dataset_final.csv      # Data final
+│   ├── insert_dataset.py      # Code chargement de la table dataset_final
+├── tests/               # Tests unitaires, fonctionnels
+│   ├── test_api.py      # Test automatisé de l'API via Pytest
 |
+├── .env             # Stockage des variables sensibles et de configuration
 ├── .gitignore       # Nettoyage du dépôt
 ├── Dockerfile       # Reproduction du dépôt
 ├── poetry.lock      # Nettoyage du dépôt

poetry.lock

@@ -1882,6 +1882,83 @@ files = [
 dev = ["abi3audit", "black", "check-manifest", "coverage", "packaging", "psleak", "pylint", "pyperf", "pypinfo", "pytest", "pytest-cov", "pytest-instafail", "pytest-xdist", "requests", "rstcheck", "ruff", "setuptools", "sphinx", "sphinx_rtd_theme", "toml-sort", "twine", "validate-pyproject[all]", "virtualenv", "vulture", "wheel"]
 test = ["psleak", "pytest", "pytest-instafail", "pytest-xdist", "setuptools"]
 
+[[package]]
+name = "psycopg2-binary"
+version = "2.9.11"
+description = "psycopg2 - Python-PostgreSQL Database Adapter"
+optional = false
+python-versions = ">=3.9"
+groups = ["main"]
+files = [
+    {file = "psycopg2-binary-2.9.11.tar.gz", hash = "sha256:b6aed9e096bf63f9e75edf2581aa9a7e7186d97ab5c177aa6c87797cd591236c"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:d6fe6b47d0b42ce1c9f1fa3e35bb365011ca22e39db37074458f27921dca40f2"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:a6c0e4262e089516603a09474ee13eabf09cb65c332277e39af68f6233911087"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:c47676e5b485393f069b4d7a811267d3168ce46f988fa602658b8bb901e9e64d"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-manylinux2014_ppc64le.manylinux_2_17_ppc64le.whl", hash = "sha256:a28d8c01a7b27a1e3265b11250ba7557e5f72b5ee9e5f3a2fa8d2949c29bf5d2"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:5f3f2732cf504a1aa9e9609d02f79bea1067d99edf844ab92c247bbca143303b"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-manylinux_2_38_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:865f9945ed1b3950d968ec4690ce68c55019d79e4497366d36e090327ce7db14"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:91537a8df2bde69b1c1db01d6d944c831ca793952e4f57892600e96cee95f2cd"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-musllinux_1_2_ppc64le.whl", hash = "sha256:4dca1f356a67ecb68c81a7bc7809f1569ad9e152ce7fd02c2f2036862ca9f66b"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-musllinux_1_2_riscv64.whl", hash = "sha256:0da4de5c1ac69d94ed4364b6cbe7190c1a70d325f112ba783d83f8440285f152"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:37d8412565a7267f7d79e29ab66876e55cb5e8e7b3bbf94f8206f6795f8f7e7e"},
+    {file = "psycopg2_binary-2.9.11-cp310-cp310-win_amd64.whl", hash = "sha256:c665f01ec8ab273a61c62beeb8cce3014c214429ced8a308ca1fc410ecac3a39"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:0e8480afd62362d0a6a27dd09e4ca2def6fa50ed3a4e7c09165266106b2ffa10"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:763c93ef1df3da6d1a90f86ea7f3f806dc06b21c198fa87c3c25504abec9404a"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:2e164359396576a3cc701ba8af4751ae68a07235d7a380c631184a611220d9a4"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-manylinux2014_ppc64le.manylinux_2_17_ppc64le.whl", hash = "sha256:d57c9c387660b8893093459738b6abddbb30a7eab058b77b0d0d1c7d521ddfd7"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:2c226ef95eb2250974bf6fa7a842082b31f68385c4f3268370e3f3870e7859ee"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-manylinux_2_38_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:a311f1edc9967723d3511ea7d2708e2c3592e3405677bf53d5c7246753591fbb"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:ebb415404821b6d1c47353ebe9c8645967a5235e6d88f914147e7fd411419e6f"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-musllinux_1_2_ppc64le.whl", hash = "sha256:f07c9c4a5093258a03b28fab9b4f151aa376989e7f35f855088234e656ee6a94"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-musllinux_1_2_riscv64.whl", hash = "sha256:00ce1830d971f43b667abe4a56e42c1e2d594b32da4802e44a73bacacb25535f"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:cffe9d7697ae7456649617e8bb8d7a45afb71cd13f7ab22af3e5c61f04840908"},
+    {file = "psycopg2_binary-2.9.11-cp311-cp311-win_amd64.whl", hash = "sha256:304fd7b7f97eef30e91b8f7e720b3db75fee010b520e434ea35ed1ff22501d03"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:be9b840ac0525a283a96b556616f5b4820e0526addb8dcf6525a0fa162730be4"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:f090b7ddd13ca842ebfe301cd587a76a4cf0913b1e429eb92c1be5dbeb1a19bc"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:ab8905b5dcb05bf3fb22e0cf90e10f469563486ffb6a96569e51f897c750a76a"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-manylinux2014_ppc64le.manylinux_2_17_ppc64le.whl", hash = "sha256:bf940cd7e7fec19181fdbc29d76911741153d51cab52e5c21165f3262125685e"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:fa0f693d3c68ae925966f0b14b8edda71696608039f4ed61b1fe9ffa468d16db"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-manylinux_2_38_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:a1cf393f1cdaf6a9b57c0a719a1068ba1069f022a59b8b1fe44b006745b59757"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:ef7a6beb4beaa62f88592ccc65df20328029d721db309cb3250b0aae0fa146c3"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-musllinux_1_2_ppc64le.whl", hash = "sha256:31b32c457a6025e74d233957cc9736742ac5a6cb196c6b68499f6bb51390bd6a"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-musllinux_1_2_riscv64.whl", hash = "sha256:edcb3aeb11cb4bf13a2af3c53a15b3d612edeb6409047ea0b5d6a21a9d744b34"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:62b6d93d7c0b61a1dd6197d208ab613eb7dcfdcca0a49c42ceb082257991de9d"},
+    {file = "psycopg2_binary-2.9.11-cp312-cp312-win_amd64.whl", hash = "sha256:b33fabeb1fde21180479b2d4667e994de7bbf0eec22832ba5d9b5e4cf65b6c6d"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:b8fb3db325435d34235b044b199e56cdf9ff41223a4b9752e8576465170bb38c"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:366df99e710a2acd90efed3764bb1e28df6c675d33a7fb40df9b7281694432ee"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:8c55b385daa2f92cb64b12ec4536c66954ac53654c7f15a203578da4e78105c0"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-manylinux2014_ppc64le.manylinux_2_17_ppc64le.whl", hash = "sha256:c0377174bf1dd416993d16edc15357f6eb17ac998244cca19bc67cdc0e2e5766"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:5c6ff3335ce08c75afaed19e08699e8aacf95d4a260b495a4a8545244fe2ceb3"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-manylinux_2_38_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:84011ba3109e06ac412f95399b704d3d6950e386b7994475b231cf61eec2fc1f"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:ba34475ceb08cccbdd98f6b46916917ae6eeb92b5ae111df10b544c3a4621dc4"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-musllinux_1_2_ppc64le.whl", hash = "sha256:b31e90fdd0f968c2de3b26ab014314fe814225b6c324f770952f7d38abf17e3c"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-musllinux_1_2_riscv64.whl", hash = "sha256:d526864e0f67f74937a8fce859bd56c979f5e2ec57ca7c627f5f1071ef7fee60"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:04195548662fa544626c8ea0f06561eb6203f1984ba5b4562764fbeb4c3d14b1"},
+    {file = "psycopg2_binary-2.9.11-cp313-cp313-win_amd64.whl", hash = "sha256:efff12b432179443f54e230fdf60de1f6cc726b6c832db8701227d089310e8aa"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-macosx_10_13_x86_64.whl", hash = "sha256:92e3b669236327083a2e33ccfa0d320dd01b9803b3e14dd986a4fc54aa00f4e1"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-macosx_11_0_arm64.whl", hash = "sha256:e0deeb03da539fa3577fcb0b3f2554a97f7e5477c246098dbb18091a4a01c16f"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:9b52a3f9bb540a3e4ec0f6ba6d31339727b2950c9772850d6545b7eae0b9d7c5"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-manylinux2014_ppc64le.manylinux_2_17_ppc64le.whl", hash = "sha256:db4fd476874ccfdbb630a54426964959e58da4c61c9feba73e6094d51303d7d8"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:47f212c1d3be608a12937cc131bd85502954398aaa1320cb4c14421a0ffccf4c"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-manylinux_2_38_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:e35b7abae2b0adab776add56111df1735ccc71406e56203515e228a8dc07089f"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-musllinux_1_2_aarch64.whl", hash = "sha256:fcf21be3ce5f5659daefd2b3b3b6e4727b028221ddc94e6c1523425579664747"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-musllinux_1_2_ppc64le.whl", hash = "sha256:9bd81e64e8de111237737b29d68039b9c813bdf520156af36d26819c9a979e5f"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-musllinux_1_2_riscv64.whl", hash = "sha256:32770a4d666fbdafab017086655bcddab791d7cb260a16679cc5a7338b64343b"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-musllinux_1_2_x86_64.whl", hash = "sha256:c3cb3a676873d7506825221045bd70e0427c905b9c8ee8d6acd70cfcbd6e576d"},
+    {file = "psycopg2_binary-2.9.11-cp314-cp314-win_amd64.whl", hash = "sha256:4012c9c954dfaccd28f94e84ab9f94e12df76b4afb22331b1f0d3154893a6316"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:20e7fb94e20b03dcc783f76c0865f9da39559dcc0c28dd1a3fce0d01902a6b9c"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:4bdab48575b6f870f465b397c38f1b415520e9879fdf10a53ee4f49dcbdf8a21"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-manylinux2014_aarch64.manylinux_2_17_aarch64.whl", hash = "sha256:9d3a9edcfbe77a3ed4bc72836d466dfce4174beb79eda79ea155cc77237ed9e8"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-manylinux2014_ppc64le.manylinux_2_17_ppc64le.whl", hash = "sha256:44fc5c2b8fa871ce7f0023f619f1349a0aa03a0857f2c96fbc01c657dcbbdb49"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-manylinux2014_x86_64.manylinux_2_17_x86_64.whl", hash = "sha256:9c55460033867b4622cda1b6872edf445809535144152e5d14941ef591980edf"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-manylinux_2_38_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:2d11098a83cca92deaeaed3d58cfd150d49b3b06ee0d0852be466bf87596899e"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-musllinux_1_2_aarch64.whl", hash = "sha256:691c807d94aecfbc76a14e1408847d59ff5b5906a04a23e12a89007672b9e819"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-musllinux_1_2_ppc64le.whl", hash = "sha256:8b81627b691f29c4c30a8f322546ad039c40c328373b11dff7490a3e1b517855"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-musllinux_1_2_riscv64.whl", hash = "sha256:b637d6d941209e8d96a072d7977238eea128046effbf37d1d8b2c0764750017d"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:41360b01c140c2a03d346cec3280cf8a71aa07d94f3b1509fa0161c366af66b4"},
+    {file = "psycopg2_binary-2.9.11-cp39-cp39-win_amd64.whl", hash = "sha256:875039274f8a2361e5207857899706da840768e2a775bf8c65e82f60b197df02"},
+]
+
 [[package]]
 name = "ptyprocess"
 version = "0.7.0"
@@ -2146,6 +2223,21 @@ files = [
 [package.dependencies]
 six = ">=1.5"
 
+[[package]]
+name = "python-dotenv"
+version = "1.2.1"
+description = "Read key-value pairs from a .env file and set them as environment variables"
+optional = false
+python-versions = ">=3.9"
+groups = ["main"]
+files = [
+    {file = "python_dotenv-1.2.1-py3-none-any.whl", hash = "sha256:b81ee9561e9ca4004139c6cbba3a238c32b03e4894671e181b671e8cb8425d61"},
+    {file = "python_dotenv-1.2.1.tar.gz", hash = "sha256:42667e897e16ab0d66954af0e60a9caa94f0fd4ecf3aaf6d2d260eec1aa36ad6"},
+]
+
+[package.extras]
+cli = ["click (>=5.0)"]
+
 [[package]]
 name = "pytz"
 version = "2025.2"
@@ -2925,4 +3017,4 @@ scikit-learn = ["scikit-learn"]
 [metadata]
 lock-version = "2.1"
 python-versions = "3.11.9"
-content-hash = "3d637a5079e784dd4465ba9f6b8da1b9325e24dcab11362b3de0fcb746a94440"
+content-hash = "4e865f02d97ed8361a2eae0ad8e4f5e6e91092ffd20634c88693ba4ae861215e"

pyproject.toml

@@ -24,7 +24,9 @@ dependencies = [
     "ipykernel>=6.25,<7.0",
     "huggingface-hub ==1.3.1",
     "fastapi ==0.115.0",
-    "uvicorn ==0.30.1"
+    "uvicorn ==0.30.1",
+    "python-dotenv ==1.2.1",
+    "psycopg2-binary ==2.9.11"
 ]
 
 [build-system]

requirements.txt

@@ -13,3 +13,4 @@ pandas==2.2.2
 scikit-learn==1.4.2
 xgboost ==2.0.3
 huggingface-hub ==1.3.1
+python-dotenv ==1.2.1

scripts/create_tables.py

@@ -0,0 +1,6 @@
+from App.database import engine
+from App.database import Base
+
+Base.metadata.create_all(bind=engine)
+
+print("Tables créées avec succès")

scripts/insert_dataset.py

@@ -16,10 +16,7 @@ DB_HOST = os.getenv("DB_HOST")
 DB_PORT = os.getenv("DB_PORT")
 DB_NAME = os.getenv("DB_NAME")
 
-DATABASE_URL = (
-    f"postgresql+psycopg2://{DB_USER}:{DB_PASSWORD}"
-    f"@{DB_HOST}:{DB_PORT}/{DB_NAME}"
-)
+DATABASE_URL = (f"postgresql+psycopg2://{DB_USER}:{DB_PASSWORD}"f"@{DB_HOST}:{DB_PORT}/{DB_NAME}")
 
 engine = create_engine(DATABASE_URL)