update

app.py

@@ -111,8 +111,15 @@ def create_app():
         
         # Analysis Tab
         with gr.Tab("Analysis"):
-            # Use create_analysis_screen to get UI components including visualization container
-            analysis_options, analysis_params, run_analysis_btn, analysis_output, bow_top_slider, ngram_n, ngram_top, topic_count = create_analysis_screen()
+            # Fix the value unpacking to match the actual return values from create_analysis_screen()
+            analysis_components = create_analysis_screen()
+            analysis_options = analysis_components[0]
+            analysis_params = analysis_components[1]
+            run_analysis_btn = analysis_components[2]
+            analysis_output = analysis_components[3]
+            ngram_n = analysis_components[4]
+            ngram_top = analysis_components[5]
+            topic_count = analysis_components[6]
             
             # Pre-create visualization components (initially hidden)
             visualization_area_visible = gr.Checkbox(value=False, visible=False, label="Visualization Visible")
@@ -137,7 +144,7 @@ def create_app():
             status_message = gr.Markdown(visible=False)
             
             # Define a helper function to extract parameter values and run the analysis
-            def run_analysis(dataset, selected_analysis, bow_top, ngram_n, ngram_top, topic_count):
+            def run_analysis(dataset, selected_analysis, ngram_n, ngram_top, topic_count):
                 try:
                     if not dataset or "entries" not in dataset or not dataset["entries"]:
                         return (
@@ -158,7 +165,6 @@ def create_app():
                         )
                     
                     parameters = {
-                        "bow_top": bow_top,
                         "ngram_n": ngram_n,
                         "ngram_top": ngram_top,
                         "topic_count": topic_count
@@ -558,7 +564,7 @@ def create_app():
             # Connect the run button to the analysis function
             run_analysis_btn.click(
                 fn=run_analysis,
-                inputs=[dataset_state, analysis_options, bow_top_slider, ngram_n, ngram_top, topic_count],
+                inputs=[dataset_state, analysis_options, ngram_n, ngram_top, topic_count],
                 outputs=[
                     analysis_results_state,
                     analysis_output,

ui/analysis_screen.py

@@ -14,7 +14,7 @@ def create_analysis_screen():
     Create the analysis options screen
     
     Returns:
-        tuple: (analysis_options, analysis_params, run_analysis_btn, analysis_output, bow_top_slider, ngram_n, ngram_top, topic_count)
+        tuple: (analysis_options, analysis_params, run_analysis_btn, analysis_output, ngram_n, ngram_top, topic_count)
     """
     with gr.Column() as analysis_screen:
         gr.Markdown("## Analysis Options")
@@ -35,14 +35,6 @@ def create_analysis_screen():
                 label="Select Analysis Type"
             )
         
-        # Create slider directly here for easier access
-        gr.Markdown("### Bag of Words Parameters")
-        bow_top_slider = gr.Slider(
-            minimum=10, maximum=100, value=25, step=5, 
-            label="Top Words to Compare", 
-            elem_id="bow_top_slider"
-        )
-        
         # Create N-gram parameters accessible at top level
         ngram_n = gr.Radio(
             choices=["1", "2", "3"], value="2", 
@@ -61,15 +53,6 @@ def create_analysis_screen():
             label="Number of Topics",
             visible=False
         )
-
-        bias_methods = gr.CheckboxGroup(
-            choices=["sentiment", "partisan", "framing"],
-            label="Bias Detection Methods",
-            value=["sentiment", "partisan"],
-            visible=False,  # Initially hidden, will be shown when Bias Detection is selected
-            interactive=True
-    )
-    
         
         # Parameters for each analysis type
         with gr.Group() as analysis_params:
@@ -83,15 +66,10 @@ def create_analysis_screen():
                 gr.Markdown("### N-gram Parameters")
                 # We're already using ngram_n and ngram_top defined above
                 
-            # Bias detection parameters
+            # Bias detection parameters - simplified with no checkboxes
             with gr.Group(visible=False) as bias_params:
                 gr.Markdown("### Bias Detection Parameters")
-                bias_methods = gr.CheckboxGroup(
-                    choices=["Sentiment Analysis", "Partisan Leaning", "Framing Analysis"],
-                    value=["Sentiment Analysis", "Partisan Leaning"],
-                    label="Bias Detection Methods",
-                    interactive=True  # Ensure this is interactive
-                )
+                gr.Markdown("Using partisan leaning bias detection")
             
             # Classifier parameters
             with gr.Group(visible=False) as classifier_params:
@@ -108,7 +86,6 @@ def create_analysis_screen():
                     ngram_n: gr.update(visible=selected == "N-gram Analysis"),
                     ngram_top: gr.update(visible=selected == "N-gram Analysis"),
                     topic_count: gr.update(visible=selected == "Topic Modeling"),
-                    bow_top_slider: gr.update(visible=selected == "Bag of Words")
                 }
                 
             # Set up event handler for analysis selection
@@ -123,7 +100,6 @@ def create_analysis_screen():
                     ngram_n, 
                     ngram_top, 
                     topic_count,
-                    bow_top_slider
                 ]
             )
         
@@ -133,8 +109,8 @@ def create_analysis_screen():
         # Analysis output area - hidden JSON component to store raw results
         analysis_output = gr.JSON(label="Analysis Results", visible=False)
     
-    # Return the components needed by app.py, now including bias_methods
-    return analysis_options, analysis_params, run_analysis_btn, analysis_output, bow_top_slider, ngram_n, ngram_top, topic_count, bias_methods
+    # Return the components needed by app.py, with bow_top_slider removed
+    return analysis_options, analysis_params, run_analysis_btn, analysis_output, ngram_n, ngram_top, topic_count
 
 # Add the implementation of these helper functions
 def extract_important_words(text, top_n=20):
@@ -370,12 +346,8 @@ def process_analysis_request(dataset, selected_analysis, parameters):
     
     # Process based on the selected analysis type
     if selected_analysis == "Bag of Words":
-        # Get the top_n parameter and ensure it's an integer
-        top_n = parameters.get("bow_top", 25)
-        if isinstance(top_n, str):
-            top_n = int(top_n)
-        
-        print(f"Using top_n value: {top_n}")  # Debug print
+        # Use fixed default value of 25 for top_n
+        top_n = 25
         
         # Perform Bag of Words analysis using the processor
         bow_results = compare_bow(
@@ -448,10 +420,7 @@ def process_analysis_request(dataset, selected_analysis, parameters):
         }
 
     elif selected_analysis == "Bias Detection":
-        # Get the bias detection methods from parameters
-        bias_methods = parameters.get("bias_methods", 
-                                   ["Sentiment Analysis", "Partisan Leaning", "Framing Analysis"])
-        
+        # Use partisan leaning bias detection by default
         try:
             # Perform bias detection analysis
             bias_results = compare_bias(
@@ -460,25 +429,19 @@ def process_analysis_request(dataset, selected_analysis, parameters):
                 model_names=[model1_name, model2_name]
             )
             
-            # Filter results based on selected methods
+            # Filter results to only include partisan leaning
             filtered_results = {"models": [model1_name, model2_name]}
             
             # Always include comparative data
             if "comparative" in bias_results:
                 filtered_results["comparative"] = bias_results["comparative"]
             
-            # Include individual model results based on selected methods
+            # Include only partisan leaning for each model
             for model in [model1_name, model2_name]:
                 filtered_results[model] = {}
                 
-                if "Sentiment Analysis" in bias_methods and model in bias_results:
-                    filtered_results[model]["sentiment"] = bias_results[model]["sentiment"]
-                
-                if "Partisan Leaning" in bias_methods and model in bias_results:
+                if model in bias_results:
                     filtered_results[model]["partisan"] = bias_results[model]["partisan"]
-                
-                if "Framing Analysis" in bias_methods and model in bias_results:
-                    filtered_results[model]["framing"] = bias_results[model]["framing"]
             
             results["analyses"][prompt_text]["bias_detection"] = filtered_results