Update app.py

app.py

@@ -1,8 +1,9 @@
 import streamlit as st
 import joblib
 import numpy as np
+import pandas as pd
 
-# Load model and scaler
+# Load model and scaler once, cached for performance
 @st.cache_resource
 def load_artifacts():
     model = joblib.load("bug_predictor_model.pkl")
@@ -12,7 +13,10 @@ def load_artifacts():
 model, scaler = load_artifacts()
 
 st.title("🔍 Software Bug Prediction System")
-st.write("Predict whether a software module is likely to be **defective** based on its metrics (NASA KC1 dataset).")
+st.write(
+    "Predict whether a software module is likely to be **defective** based on metrics "
+    "from the NASA KC1 dataset."
+)
 
 # List of feature names in the same order as training
 feature_names = [
@@ -21,26 +25,106 @@ feature_names = [
     'uniq_Opnd', 'total_Op', 'total_Opnd', 'branchCount'
 ]
 
-st.subheader("📥 Enter Module Metrics")
+tab_single, tab_bulk = st.tabs(["🧮 Single module input", "📂 Bulk prediction via CSV"])
 
-inputs = []
-cols = st.columns(3)
+# =========================
+# TAB 1: SINGLE ROW INPUT
+# =========================
+with tab_single:
+    st.subheader("📥 Enter Module Metrics Manually")
 
-for idx, name in enumerate(feature_names):
-    with cols[idx % 3]:
-        val = st.number_input(name, value=0.0)
-        inputs.append(val)
+    inputs = []
+    cols = st.columns(3)  # 3-column layout for nicer UI
 
-if st.button("Predict Defect Risk"):
-    input_array = np.array(inputs).reshape(1, -1)
-    scaled = scaler.transform(input_array)
-    pred = model.predict(scaled)[0]
-    proba = getattr(model, "predict_proba", lambda x: None)(scaled)
+    for idx, name in enumerate(feature_names):
+        with cols[idx % 3]:
+            val = st.number_input(name, value=0.0)
+            inputs.append(val)
 
-    if pred == 1:
-        st.error("⚠️ Defect Likely")
-    else:
-        st.success("✅ No Defect Predicted")
+    if st.button("Predict Defect Risk", key="single_predict"):
+        # Convert inputs to 2D array
+        input_array = np.array(inputs).reshape(1, -1)
 
-    if proba is not None:
-        st.write(f"Probability of Defect: **{proba[0][1]:.2f}**")
+        # Scale using same scaler from training
+        scaled = scaler.transform(input_array)
+
+        # Predict with loaded model
+        pred = model.predict(scaled)[0]
+
+        # Probability of defect (if supported)
+        proba = model.predict_proba(scaled)[0][1] if hasattr(model, "predict_proba") else None
+
+        if pred == 1:
+            st.error("⚠️ Defect Likely")
+        else:
+            st.success("✅ No Defect Predicted")
+
+        if proba is not None:
+            st.write(f"Probability of Defect: **{proba:.2f}**")
+
+
+# =========================
+# TAB 2: BULK CSV PREDICTION
+# =========================
+with tab_bulk:
+    st.subheader("📂 Upload CSV for Bulk Prediction")
+    st.write(
+        "Upload a CSV file containing the following columns (no target column needed):"
+    )
+    st.code(", ".join(feature_names), language="text")
+
+    uploaded_file = st.file_uploader("Choose a CSV file", type=["csv"])
+
+    if uploaded_file is not None:
+        try:
+            df = pd.read_csv(uploaded_file)
+
+            st.write("📄 Preview of uploaded data:")
+            st.dataframe(df.head())
+
+            # Check if all required columns exist
+            missing_cols = [col for col in feature_names if col not in df.columns]
+            if missing_cols:
+                st.error(
+                    "The following required columns are missing from the uploaded file:\n"
+                    + ", ".join(missing_cols)
+                )
+            else:
+                # Keep only the required columns in correct order
+                X = df[feature_names].copy()
+
+                # Scale features
+                X_scaled = scaler.transform(X)
+
+                # Predict
+                preds = model.predict(X_scaled)
+
+                # Probabilities (if available)
+                if hasattr(model, "predict_proba"):
+                    probas = model.predict_proba(X_scaled)[:, 1]
+                else:
+                    probas = None
+
+                # Add predictions to dataframe
+                df["Defect_Prediction"] = np.where(
+                    preds == 1, "Defect Likely", "No Defect Predicted"
+                )
+
+                if probas is not None:
+                    df["Defect_Probability"] = probas
+
+                st.success("✅ Predictions generated!")
+                st.write("📊 Results:")
+                st.dataframe(df.head())
+
+                # Allow user to download results
+                csv_data = df.to_csv(index=False).encode("utf-8")
+                st.download_button(
+                    label="⬇️ Download Predictions as CSV",
+                    data=csv_data,
+                    file_name="bug_predictions.csv",
+                    mime="text/csv",
+                )
+
+        except Exception as e:
+            st.error(f"❌ Error reading file: {e}")