Update app.py

app.py

@@ -1,6 +1,3 @@
-# Install dependencies (if needed)
-# !pip install gradio pandas scikit-learn
-
 import pandas as pd
 import gradio as gr
 from sklearn.feature_extraction.text import TfidfVectorizer
@@ -9,16 +6,15 @@ from sklearn.metrics.pairwise import cosine_similarity
 # Load dataset
 df = pd.read_csv("mcq_dataset.csv")
 
-# Clean domain/subdomain for consistent matching
+# Normalize domain and subdomain values
 df['domain'] = df['domain'].astype(str).str.strip().str.lower()
 df['subdomain'] = df['subdomain'].astype(str).str.strip()
 
-# Function to get top N MCQs using cosine similarity
+# Function to fetch top MCQs
 def get_top_mcqs(user_input, domain, subdomain, top_n=10):
     if not domain or not subdomain:
         return pd.DataFrame()
 
-    # Normalize input for comparison
     domain = domain.strip().lower()
     subdomain = subdomain.strip()
 
@@ -41,7 +37,7 @@ def get_top_mcqs(user_input, domain, subdomain, top_n=10):
 
     return top_questions.reset_index(drop=True)
 
-# Quiz execution logic
+# Function to format quiz output
 def run_quiz(domain, subdomain, keyword_input):
     mcq_df = get_top_mcqs(keyword_input, domain, subdomain)
 
@@ -59,26 +55,31 @@ def run_quiz(domain, subdomain, keyword_input):
 
     return quiz_output
 
-# Update subdomains dynamically
+# Dynamic subdomain update
 def update_subdomains(domain):
     if not domain:
         return gr.Dropdown.update(choices=[], value=None)
 
     domain = domain.strip().lower()
     subdomains = df[df["domain"] == domain]["subdomain"].dropna().unique().tolist()
-    
     if not subdomains:
         subdomains = ["No Subdomains Available"]
 
     return gr.Dropdown.update(choices=sorted(subdomains), value=subdomains[0])
 
-# Gradio UI
+# Gradio Interface
 with gr.Blocks() as demo:
     gr.Markdown("## 🧠 Domain-Based MCQ Quiz System")
 
-    with gr.Row():
-        domain_dropdown = gr.Dropdown(label="Select Domain", choices=sorted(df['domain'].unique().tolist()))
-        subdomain_dropdown = gr.Dropdown(label="Select Subdomain")
+    with gr.Column():
+        domain_dropdown = gr.Dropdown(
+            label="Select Domain", 
+            choices=sorted(df['domain'].unique().tolist())
+        )
+        subdomain_dropdown = gr.Dropdown(
+            label="Select Subdomain", 
+            filterable=True  # 👈 Enables searching subdomains
+        )
 
     domain_dropdown.change(fn=update_subdomains, inputs=domain_dropdown, outputs=subdomain_dropdown)