Update app.py

app.py

@@ -76,29 +76,76 @@ async def visualize(
 
 
     return StreamingResponse(img_stream, media_type="image/png")'''
-from fastapi import FastAPI
 import gradio as gr
-import uvicorn
+import pandas as pd
+import matplotlib.pyplot as plt
+import seaborn as sns
+import openai
+from fastapi import FastAPI
+from gradio.routes import mount_gradio_app
+from transformers import pipeline
+
+# Load Hugging Face model for NLP interpretation
+generator = pipeline("text2text-generation", model="HuggingFaceH4/t5-small")
 
-# FastAPI instance
 app = FastAPI()
 
-# Example FastAPI route
-@app.get("/")
-def read_root():
-    return {"message": "Welcome to FastAPI & Gradio App!"}
+def generate_code(natural_language_request, df):
+    """Convert NL request to Python visualization code using a Hugging Face model."""
+    prompt = f"Generate Python code to visualize data using matplotlib or seaborn: {natural_language_request}"
+    response = generator(prompt, max_length=100)[0]['generated_text']
+    
+    # Ensure the generated code runs with df
+    code = f"""
+import matplotlib.pyplot as plt
+import seaborn as sns
 
-# Gradio Interface
-def greet(name):
-    return f"Hello, {name}!"
+df = {df.to_dict()}  # Convert dataframe to dictionary for execution
+{response}
 
-gradio_app = gr.Interface(fn=greet, inputs="text", outputs="text")
+plt.show()
+"""
+    return code
 
-# Mount Gradio inside FastAPI
-@app.get("/gradio")
-def gradio_ui():
-    return gradio_app.launch(share=False, inline=True)
+def visualize(file, request):
+    """Process the Excel file and generate a visualization based on the user's request."""
+    if not file:
+        return "Please upload an Excel file."
+    
+    # Load the first sheet of the Excel file
+    df = pd.read_excel(file, sheet_name=0)
+
+    # Generate Python visualization code
+    code = generate_code(request, df)
+
+    # Execute the generated code safely
+    try:
+        exec_globals = {"df": df, "plt": plt, "sns": sns}
+        exec(code, exec_globals)
+        
+        # Save plot to image file
+        plt.savefig("output.png")
+        return "output.png"
+    except Exception as e:
+        return f"Error executing code: {str(e)}"
+
+# Gradio UI
+gr_interface = gr.Interface(
+    fn=visualize,
+    inputs=[
+        gr.File(label="Upload Excel File"),
+        gr.Textbox(label="Describe your visualization (e.g., 'Show a histogram of column A')")
+    ],
+    outputs=gr.Image(label="Generated Visualization"),
+    title="AI-Powered Data Visualization",
+    description="Upload an Excel file and describe your visualization in natural language."
+)
+
+# Mount Gradio app to FastAPI
+app = mount_gradio_app(app, gr_interface, path="/")
 
 if __name__ == "__main__":
+    import uvicorn
     uvicorn.run(app, host="0.0.0.0", port=7860)
 
+