video added

app.py

@@ -71,21 +71,10 @@ def process_data(file_obj, model_col: str, train_col: str, data_source: str) ->
         if not file_obj:
             return "Error: No file uploaded.", None, None, None, None, None, None
 
-        file_path = file_obj.name
-        if file_path.endswith(".csv"):
-            df = pd.read_csv(file_path)
-        elif file_path.endswith(".json"):
-            df = pd.read_json(file_path)
-        else:
-            return (
-                "Error: Invalid file type. Please upload a CSV or JSON file.",
-                None,
-                None,
-                None,
-                None,
-                None,
-                None,
-            )
+        global uploaded_df
+        df = uploaded_df.get("data")
+        if df is None:
+            return "Error: File not yet processed.", None, None, None, None, None, None
 
         if model_col not in df.columns or train_col not in df.columns:
             return (
@@ -113,7 +102,7 @@ def process_data(file_obj, model_col: str, train_col: str, data_source: str) ->
         evaluation_markdown = summary + details + explanation
 
         return (
-            None,  # No error
+            None,
             render_pipeline_graph(data_source),
             df.head().to_markdown(index=False, numalign="left", stralign="left"),
             evaluation_markdown,
@@ -121,16 +110,25 @@ def process_data(file_obj, model_col: str, train_col: str, data_source: str) ->
             ttr_train,
             similarity,
         )
-
     except Exception as e:
         return f"An error occurred: {str(e)}", None, None, None, None, None, None
 
 
-def update_dropdowns(file_obj) -> Tuple[list, list, str]:
-    """Updates dropdown choices based on the uploaded file."""
+# Store uploaded DataFrame globally for later access
+uploaded_df = {}
+
+
+def update_dropdowns(file_obj) -> Tuple[gr.Dropdown, gr.Dropdown, str]:
+    global uploaded_df
     if not file_obj:
-        return [], [], "No file uploaded."
+        uploaded_df["data"] = None  # Clear cached file
+        return (
+            gr.update(choices=[], value=None),
+            gr.update(choices=[], value=None),
+            "No file uploaded.",
+        )
 
+    # Read the file and extract columns
     try:
         file_name = getattr(file_obj, "name", "")
         if file_name.endswith(".csv"):
@@ -138,14 +136,45 @@ def update_dropdowns(file_obj) -> Tuple[list, list, str]:
         elif file_name.endswith(".json"):
             df = pd.read_json(file_obj)
         else:
-            return [], [], "Invalid file type. Only .csv and .json are supported."
+            return (
+                gr.update(choices=[], value=None),
+                gr.update(choices=[], value=None),
+                "Invalid file type.",
+            )
 
+        uploaded_df["data"] = df
         columns = df.columns.tolist()
         preview = df.head().to_markdown(index=False, numalign="left", stralign="left")
-        return columns, columns, preview  # For two dropdowns + preview
+
+        return (
+            gr.update(choices=columns, value=None),
+            gr.update(choices=columns, value=None),
+            preview,
+        )
 
     except Exception as e:
-        return [], [], f"Error reading file: {e}"
+        return (
+            gr.update(choices=[], value=None),
+            gr.update(choices=[], value=None),
+            f"Error reading file: {e}",
+        )
+
+
+def clear_all_fields():
+    global uploaded_df
+    uploaded_df.clear()  # Clear stored DataFrame
+    return (
+        None,  # file_input
+        gr.update(choices=[], value=None),  # model_col_input
+        gr.update(choices=[], value=None),  # train_col_input
+        "",  # file_preview
+        "",  # output_markdown
+        "",  # warning_output
+        None,  # ttr_output_metric
+        None,  # ttr_train_metric
+        None,  # similarity_metric
+        render_pipeline_graph("Synthetic Generated Data"),  # pipeline_output default
+    )
 
 
 def main_interface():
@@ -178,9 +207,22 @@ def main_interface():
 
         with gr.Accordion("📚 Research Reference", open=False):
             gr.HTML(render_core_reference())
+        gr.HTML(
+            """
+            <div style="display: flex; flex-direction: column; align-items: center; margin-bottom: 20px;">
+                <h3 style="text-align: center;">📽️ How to Use MADGuard AI Explorer</h3>
+                <iframe width="720" height="405"
+                        src="https://www.youtube.com/embed/qjMwvaBXQeY"
+                        title="MADGuard AI Tutorial" frameborder="0"
+                        allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture"
+                        allowfullscreen></iframe>
+            </div>
+            """
+        )
 
         gr.Markdown("## 1. Pipeline Simulation")
-        data_source, description = render_pipeline()
+        data_source, description = render_pipeline(default="Synthetic Generated Data")
+
         gr.HTML(description)
         pipeline_output = gr.Image(type="filepath", label="Pipeline Graph")
         warning_output = gr.HTML()
@@ -190,11 +232,18 @@ def main_interface():
         data_source.change(
             fn=render_pipeline_graph, inputs=data_source, outputs=pipeline_output
         )
+        interface.load(
+            fn=render_pipeline_graph,
+            inputs=[data_source],
+            outputs=[pipeline_output],
+        )
 
         gr.Markdown("## 2. Upload CSV or JSON File")
         file_input = gr.File(
             file_types=[".csv", ".json"], label="Upload a CSV or JSON file"
         )
+        clear_btn = gr.Button("🧹 Clear All")
+
         gr.Markdown(
             """
     📝 **Note:**  
@@ -206,12 +255,17 @@ def main_interface():
 
         with gr.Row():
             model_col_input = gr.Dropdown(
-                choices=[], label="Select column for model output"
+                choices=[],
+                value=None,
+                label="Select column for model output",
+                interactive=True,
             )
             train_col_input = gr.Dropdown(
-                choices=[], label="Select column for future training data"
+                choices=[],
+                value=None,
+                label="Select column for future training data",
+                interactive=True,
             )
-
         file_preview = gr.Markdown(label="📄 File Preview")
 
         output_markdown = gr.Markdown(label="🔍 Evaluation Summary")
@@ -224,10 +278,15 @@ def main_interface():
             ttr_train_metric = gr.Number(label="Lexical Diversity (Training Set)")
             similarity_metric = gr.Number(label="Semantic Similarity (Cosine)")
 
+        def handle_file_upload(file_obj, data_source_val):
+            dropdowns = update_dropdowns(file_obj)
+            graph = render_pipeline_graph(data_source_val)
+            return *dropdowns, graph
+
         file_input.change(
-            update_dropdowns,
-            inputs=file_input,
-            outputs=[model_col_input, train_col_input, file_preview],
+            fn=handle_file_upload,
+            inputs=[file_input, data_source],
+            outputs=[model_col_input, train_col_input, file_preview, pipeline_output],
         )
 
         def process_and_generate(
@@ -260,6 +319,23 @@ def main_interface():
             ttr_train_metric,
             similarity_metric,
         ]
+        clear_btn.click(
+            fn=clear_all_fields,
+            inputs=[],
+            outputs=[
+                file_input,
+                model_col_input,
+                train_col_input,
+                file_preview,
+                output_markdown,
+                warning_output,
+                ttr_output_metric,
+                ttr_train_metric,
+                similarity_metric,
+                pipeline_output,
+            ],
+        )
+
         for input_component in inputs:
             input_component.change(
                 fn=process_and_generate, inputs=inputs, outputs=outputs