init

app.py

@@ -62,20 +62,30 @@ from tensorflow.keras.layers import LSTM
 with custom_object_scope({'Orthogonal': Orthogonal}):
     model = load_model('models/lstm-combinedmodel.h5')
 
-# Function to parse a text file into a DataFrame
+# Function to parse the uploaded file from Gradio
 def parse_text_file(file):
-    # Read content from file-like object
-    content = file.read()
-    if isinstance(content, bytes):  # Check if the content is bytes, if so, decode it
+    # Gradio uploads files as a 'werkzeug.datastructures.FileStorage' object
+    # which acts like a file object. Reading it directly if .read() doesn't work.
+    # We need to check the type and handle it accordingly.
+    if hasattr(file, 'read'):
+        content = file.read()
+    elif hasattr(file, 'file'):
+        # In case Gradio passes the FileStorage object where the file is accessible via '.file'
+        content = file.file.read()
+    else:
+        raise ValueError("The file format provided is not supported.")
+        
+    if isinstance(content, bytes):  # Check if the content is bytes, decode it
         content = content.decode('utf-8')
-    file_like = StringIO(content)  # Use StringIO to create a file-like object from the string
 
-    # Dictionary to hold data extracted from the text file
+    file_like = StringIO(content)  # Convert string to a file-like object
+
+    # Dictionary to hold data extracted from text file
     dfdict = {}
 
-    # Read file lines and process them
+    # Process each line from file-like object
     for line in file_like:
-        line = line.strip().split()  # Split line into parts, also stripping newlines
+        line = line.strip().split()  # Split line into parts, removing newlines and extra spaces
         if 'Timestamp:' in line:
             line.remove('Timestamp:')  # Remove specific label
         if 'ID:' in line:
@@ -83,8 +93,8 @@ def parse_text_file(file):
         if 'DLC:' in line:
             line.remove('DLC:')
         
-        # Extract timestamp as key and remaining data as value
-        if line:
+        # Ensure line has enough elements to extract a key and value
+        if len(line) > 2:
             key = float(line[0])
             value = line[1:]
             dfdict[key] = value  # Add to dictionary
@@ -94,48 +104,19 @@ def parse_text_file(file):
     df.index.name = 'Timestamp'  # Set index name
     return df
 
-# Function to use the model to make predictions from the DataFrame
-def make_predictions(dataframe):
-    # Ensure DataFrame has the necessary columns
-    required_columns = ['CAN ID', 'RTR', 'DLC', 'Data1', 'Data2', 'Data3', 'Data4', 'Data5', 'Data6', 'Data7', 'Data8']
-    if not all(col in dataframe.columns for col in required_columns):
-        raise ValueError("Missing required columns in the DataFrame.")
-    
-    # Convert to the format expected by the model
-    input_data = dataframe[required_columns].values
-    
-    # Predict using the model
-    predictions = model.predict(input_data)
-    
-    # Determine the predicted class and confidence
-    predicted_class = np.argmax(predictions, axis=1)
-    confidence = np.max(predictions, axis=1)
-    
-    # Map numeric class to label
-    class_labels = {0: "Normal", 1: "Anomaly"}
-    
-    results = []
-    for cls, conf in zip(predicted_class, confidence):
-        results.append(f"Predicted Class: {class_labels[cls]}, Confidence: {conf:.4f}")
-    
-    return results
-
-# Gradio interface function
 def interface_func(uploaded_file):
     # Parse the text file into a DataFrame
     df = parse_text_file(uploaded_file)
     
-    # Get predictions
-    predictions = make_predictions(df)
-    
-    return "\n".join(predictions)  # Return predictions as text
+    # Optionally process the DataFrame further or directly show it
+    return df.to_html()
 
 # Set up the Gradio interface
 iface = gr.Interface(
     fn=interface_func,
     inputs=gr.File(label="Upload a text file"),
-    outputs="text",
-    description="Upload a text file with CAN data to receive predictions."
+    outputs="html",
+    description="Upload a text file with CAN data to convert it to a DataFrame."
 )
 
 # Launch the interface