From 2412af6abe293ead46536a4d72a7698c1db6d154 Mon Sep 17 00:00:00 2001
From: wangchunlin <wangchunin666@gmail.com>
Date: Fri, 29 Nov 2024 16:31:11 +0800
Subject: [PATCH] =?UTF-8?q?=E6=9C=AA=E8=AE=AD=E7=BB=83=E6=97=B6=E4=BD=BF?=
 =?UTF-8?q?=E7=94=A8=E9=BB=98=E8=AE=A4=E6=A8=A1=E5=9E=8B?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 config/config.yaml         | 13 ++++++-------
 utils/feature_processor.py | 17 +++++++++++++++++
 utils/model_trainer.py     | 20 ++++++++++++++++++--
 3 files changed, 41 insertions(+), 9 deletions(-)

diff --git a/config/config.yaml b/config/config.yaml
index 618cf18..ecccbd2 100644
--- a/config/config.yaml
+++ b/config/config.yaml
@@ -8,7 +8,7 @@ system:
     - "evaluate_api"
   host: "0.0.0.0"
   port: 8088
-  device: "cuda"  # 可选: "cpu", "cuda"
+  device: "cpu"  # 可选: "cpu", "cuda"
 
 #---路径配置---#
 paths:
@@ -24,10 +24,10 @@ paths:
 #---训练配置---#
 training:
   n_epochs: 300
-  # batch_size: 16 # CPU
-  # learning_rate: 0.001 # CPU
-  batch_size: 128 # GPU
-  learning_rate: 0.001 # GPU
+  batch_size: 16 # CPU
+  learning_rate: 0.001 # CPU
+  # batch_size: 128 # GPU
+  # learning_rate: 0.001 # GPU
   early_stop_patience: 50
   scheduler:
     gamma: 0.98
@@ -36,7 +36,7 @@ training:
     l1_lambda: 1e-5
     l2_lambda: 1e-4
   dropout_rate: 0.2
-  experimental_mode: true
+  experimental_mode: false
   experiments_count: 50
   replace_model: true
   data_mode: "train_val"  # 可选: "train", "train_val", "all"
@@ -59,7 +59,6 @@ model:
 #---特征配置---#
 features:
   label_name: "类别"
-  correlation_threshold: 0.7
   groups:
     核心症状:
       - "强迫症状数字化"
diff --git a/utils/feature_processor.py b/utils/feature_processor.py
index 91dc623..52896e7 100644
--- a/utils/feature_processor.py
+++ b/utils/feature_processor.py
@@ -10,6 +10,7 @@ from sklearn.preprocessing import StandardScaler
 import pickle
 from pydantic import BaseModel, Field
 import logging
+import shutil
 
 class Features(BaseModel):
     """
@@ -189,6 +190,7 @@ class FeatureWeightApplier:
 class FeatureNormalizer:
     """特征归一化处理类"""
     def __init__(self, config: Dict[str, Any]):
+        self.config = config
         self.feature_groups = {}
         self.scalers = {}
         
@@ -232,6 +234,21 @@ class FeatureNormalizer:
         Args:
             path: 加载路径
         """
+        # 如果主路径不存在，尝试从默认路径复制
+        if not os.path.exists(path):
+            default_path = os.path.join(
+                os.path.dirname(__file__), 
+                '..', 
+                self.config['paths']['normalizer']['default']
+            )
+            if os.path.exists(default_path):
+                # 确保目标目录存在
+                os.makedirs(os.path.dirname(path), exist_ok=True)
+                shutil.copyfile(default_path, path)
+                logging.info(f"Copied normalizer from default path: {default_path} to {path}")
+            else:
+                raise FileNotFoundError(f"No normalizer found in either path: {path} or {default_path}")
+            
         with open(path, 'rb') as f:
             self.scalers = pickle.load(f)
 
diff --git a/utils/model_trainer.py b/utils/model_trainer.py
index c7a33ff..62ce7b9 100644
--- a/utils/model_trainer.py
+++ b/utils/model_trainer.py
@@ -93,8 +93,24 @@ class MLModel:
     def load_model(self) -> None:
         """加载已训练的模型"""
         self.create_model()
-        self.model.load_state_dict(torch.load(self.config['paths']['model']['train'], 
-                                            map_location=self.device))
+        model_path = self.config['paths']['model']['train']
+        
+        # 如果主路径不存在，尝试从默认路径复制
+        if not os.path.exists(model_path):
+            default_model_path = os.path.join(
+                os.path.dirname(__file__), 
+                '..', 
+                self.config['paths']['model']['default']
+            )
+            if os.path.exists(default_model_path):
+                # 确保目标目录存在
+                os.makedirs(os.path.dirname(model_path), exist_ok=True)
+                shutil.copyfile(default_model_path, model_path)
+                logging.info(f"Copied model from default path: {default_model_path} to {model_path}")
+            else:
+                raise FileNotFoundError(f"No model found in either path: {model_path} or {default_model_path}")
+        
+        self.model.load_state_dict(torch.load(model_path, map_location=self.device))
 
     def plot_correlation(self, X: np.ndarray, feature_names: List[str]) -> None:
         """绘制特征相关性矩阵"""