Update app.py

app.py

@@ -193,6 +193,7 @@
 #         st.subheader("📉 Study vs Performance Trend")
 #         st.line_chart(df.set_index("Subject"))
 # Smart Study Planner - FINAL VERSION
+# Smart Study Planner - FINAL STABLE VERSION
 
 import streamlit as st
 import numpy as np
@@ -240,16 +241,22 @@ class Student:
         marks = [s.marks for s in self.subjects]
         return np.max(marks), np.min(marks)
 
+    # -------------------------------
+    # SAFE TREND CALCULATION
+    # -------------------------------
     def performance_trend(self):
-        hours = [s.hours for s in self.subjects]
-        marks = [s.marks for s in self.subjects]
+        hours = np.array([s.hours for s in self.subjects])
+        marks = np.array([s.marks for s in self.subjects])
 
         if len(hours) > 1:
-            return np.corrcoef(hours, marks)[0][1]
+            trend = np.corrcoef(hours, marks)[0][1]
+            if np.isnan(trend):
+                trend = 0
+            return np.clip(trend, -1, 1)
         return 0
 
     # -------------------------------
-    # AI-STYLE SMART PREDICTION
+    # AI-STYLE SMART PREDICTION (FIXED)
     # -------------------------------
     def predict_performance(self):
         hours = np.array([s.hours for s in self.subjects])
@@ -260,23 +267,17 @@ class Student:
 
         avg_marks = np.mean(marks)
         avg_hours = np.mean(hours)
+        trend = self.performance_trend()
 
-        # Trend (relationship between hours and marks)
-        if len(hours) > 1:
-            trend = np.corrcoef(hours, marks)[0][1]
-        else:
-            trend = 0
-
-        # Consistency (low variance = better)
-        consistency = 1 / (1 + np.var(marks))
+        variance = np.var(marks)
+        consistency = 1 / (1 + variance)
 
-        # Smart prediction formula
         predicted = avg_marks \
                     + (avg_hours * 1.5) \
                     + (trend * 10) \
                     + (consistency * 5)
 
-        return min(predicted, 100)
+        return float(min(predicted, 100))
 
     def get_grade(self, marks):
         if marks >= 85:
@@ -296,12 +297,11 @@ class Student:
         return min(self.subjects, key=lambda x: x.marks).name
 
     def suggest_improvement(self):
-        weak = self.weak_subject()
-        return f"Focus more on {weak} and increase study hours by at least 2 hours."
+        return f"Focus more on {self.weak_subject()} and increase study time by 2 hours."
 
 
 # -------------------------------
-# SESSION STATE (FIXED LOGIC)
+# SESSION STATE (FIXED)
 # -------------------------------
 
 if "student" not in st.session_state:
@@ -389,17 +389,22 @@ if st.button("📊 Analyze Performance"):
 
         # ---------------- PREDICTION ----------------
         st.subheader("🔮 Smart Prediction")
-        st.success(f"Expected Marks: {predicted:.2f}")
-        st.success(f"Predicted Grade: {predicted_grade}")
 
-        # ---------------- EXTRA INSIGHTS ----------------
+        if np.isnan(predicted):
+            st.error("Prediction error due to insufficient variation in data")
+        else:
+            st.success(f"Expected Marks: {predicted:.2f}")
+            st.success(f"Predicted Grade: {predicted_grade}")
+
+        # ---------------- INSIGHTS ----------------
         st.subheader("📌 Insights")
+
         st.write(f"🏆 Best Subject: **{student.best_subject()}**")
         st.write(f"⚠ Weak Subject: **{student.weak_subject()}**")
 
         st.info(student.suggest_improvement())
 
-        # ---------------- CONSISTENCY SCORE ----------------
+        # ---------------- CONSISTENCY ----------------
         marks_list = [s.marks for s in student.subjects]
         consistency_score = 1 / (1 + np.var(marks_list))
         st.write(f"📊 Consistency Score: {round(consistency_score, 2)}")