Update app.py

app.py

@@ -13,6 +13,17 @@ from torch.utils.data import Dataset, DataLoader
 from torch.optim.lr_scheduler import ReduceLROnPlateau
 from sklearn.model_selection import train_test_split
 import glob
+import sys
+import csv
+
+maxInt = sys.maxsize
+
+while True:
+    try:
+        csv.field_size_limit(maxInt)
+        break
+    except OverflowError:
+        maxInt = int(maxInt/10)
 
 class ClimateNet(nn.Module):
     def __init__(self, input_size=(256, 256), output_size=(64, 64)):
@@ -307,7 +318,7 @@ class ClimatePredictor:
             raise e
 
 def load_examples_from_directory(base_dir):
-    """폴더에서 예제 데이터 로드"""
+    """폴더에서 예제 데이터 로드 - CSV 처리 개선"""
     examples = []
     sample_dirs = sorted(glob.glob(os.path.join(base_dir, "sample_*")))
     
@@ -319,7 +330,13 @@ def load_examples_from_directory(base_dir):
             elevation_path = os.path.join(sample_dir, "terrain", "elevation_data.npy")
             weather_path = os.path.join(sample_dir, "weather", "weather_data.csv")
             
-            weather_data = pd.read_csv(weather_path)
+            # CSV 파일 읽기 개선
+            try:
+                weather_data = pd.read_csv(weather_path, engine='python')
+            except Exception as e:
+                print(f"Error reading CSV file {weather_path}: {str(e)}")
+                continue
+            
             wind_speed = weather_data['wind_speed'].mean()
             wind_direction = weather_data['wind_direction'].mean()
             temperature = weather_data['temperature'].mean()
@@ -335,12 +352,15 @@ def load_examples_from_directory(base_dir):
                 float(temperature),
                 float(humidity)
             ])
+            print(f"Successfully loaded example from {sample_dir}")
         except Exception as e:
             print(f"Error loading example from {sample_dir}: {str(e)}")
             continue
     
+    print(f"Total examples loaded: {len(examples)}")
     return examples
 
+
 def create_gradio_interface():
     predictor = ClimatePredictor('best_model.pth')