修改--将读取文件集成到模型训练中

api/model_api.py

@@ -51,7 +51,7 @@ async def train_model(request: TrainRequest):
     """模型训练"""
     result = model_manager.train_model(
         train_data=request.dataset['train'],
-        val_data=request.dataset.get('val'),
+        val_data=request.dataset['val'],
         model_config={
             'model_name': request.model,
             'parameters': request.parameters,

doc/安装文档.md

@@ -26,6 +26,28 @@ GET http://10.0.0.202:8992/data/feature/method/{method_name}
 ### 3.5 处理数据集
 POST http://10.0.0.202:8992/data/process
 
+传递参数
+    {
+        "input_path": "dataset/dataset_raw/breast_cancer.csv",
+        "output_dir": "dataset/dataset_processed",
+        "process_methods": [
+            {
+                "method_name": "IsolationForest",
+                "params": {
+                    "contamination": 0.1,
+                    "random_state": 42
+                }
+            }
+        ],
+        "feature_methods": [
+            
+        ],
+        "split_params": {
+            "test_size": 0.1,
+            "val_size": 0.2
+        }
+    }
+
 ### 3.6 获取可用数据集列表
 GET http://10.0.0.202:8992/data/datasets
 

example_model_manager.py

@@ -71,15 +71,7 @@ print("--------------------------------------------获取所有已训练模型 e
 
 
 print("--------------------------------------------模型训练---------------------------------------------------")
-# 加载数据
-train_data = pd.read_csv('/home/admin-root/haotian/MLPlatform/dataset/dataset_processed/breast_cancer_20250219_144629/train_breast_cancer_20250219_144629.csv')
-val_data = pd.read_csv('/home/admin-root/haotian/MLPlatform/dataset/dataset_processed/breast_cancer_20250219_144629/val_breast_cancer_20250219_144629.csv')
 
-# 准备特征和标签
-X_train = train_data.drop('target', axis=1)
-y_train = train_data['target']
-X_val = val_data.drop('target', axis=1)
-y_val = val_data['target']
 
 # 模型配置
 model_config = {
@@ -98,14 +90,8 @@ model_config = {
 # 模型文件 直接在 mlruns/文件夹下
 for i in range(3, 4):
     result = manager.train_model(
-        {
-            'features': X_train,
-            'labels': y_train
-        },
-        {
-            'features': X_val,
-            'labels': y_val
-        },
+        '/home/admin-root/haotian/MLPlatform/dataset/dataset_processed/breast_cancer_20250224_170615/train_breast_cancer_20250224_170615.csv',
+        '/home/admin-root/haotian/MLPlatform/dataset/dataset_processed/breast_cancer_20250224_170615/val_breast_cancer_20250224_170615.csv',
         model_config,
         f'breast_cancer_classification_{i}'
     )

function/model_manager.py

@@ -771,8 +771,8 @@ class ModelManager:
             }
     def train_model(
         self,
-        train_data: Dict,
-        val_data: Dict,
+        train_path: str,
+        val_path: str,
         model_config: Dict,
         experiment_name: str
     ) -> Dict:
@@ -789,6 +789,11 @@ class ModelManager:
             训练结果字典
         """
         try:
+
+
+
+
+
             # 检查实验是否存在且被删除
             experiment = mlflow.get_experiment_by_name(experiment_name)
             if experiment and experiment.lifecycle_stage == 'deleted':
@@ -800,6 +805,28 @@ class ModelManager:
             
             # 设置MLflow实验
             mlflow.set_experiment(experiment_name)
+
+            if os.path.exists(train_path):
+                # 加载数据
+                train_data = pd.read_csv(train_path)
+            else:
+                return {
+                    'status': 'error',
+                    'message': '找不到训练集路径'
+                }
+            if os.path.exists(val_path):
+                val_data = pd.read_csv(val_path)
+            else:
+                return{
+                    'status': 'error',
+                    'message': '找不到验证集路径'
+                }
+
+            # 准备特征和标签
+            X_train = train_data.drop('target', axis=1)
+            y_train = train_data['target']
+            X_val = val_data.drop('target', axis=1)
+            y_val = val_data['target']
             
             with mlflow.start_run() as run:
                 # 记录基本信息
@@ -833,15 +860,15 @@ class ModelManager:
                 
                 # 训练模型
                 self.logger.info(f"Starting training {model_config['algorithm']}")
-                model.fit(train_data['features'], train_data['labels'])
+                model.fit(X_train, y_train)
 
                 # timestamp = datetime.datetime.now().strftime('%Y%m%d_%H%M%S')
                 # mlflow.log_param('end_time', timestamp)
                 
                 # 在验证集上评估
-                val_predictions = model.predict(val_data['features'])
+                val_predictions = model.predict(X_val)
                 metrics = self._calculate_metrics(
-                    val_data['labels'],
+                    y_val,
                     val_predictions,
                     model_config['task_type']
                 )