app.py

import streamlit as st
import pandas as pd
import plotly.express as px
import sys
import os

sys.path.append(os.path.abspath('src'))
import db

st.set_page_config(page_title="Student Performance Tracker", page_icon="🎓", layout="wide")

# Session State for Login
if 'logged_in' not in st.session_state:
    st.session_state['logged_in'] = False
    st.session_state['role'] = None
    st.session_state['department'] = None
    st.session_state['username'] = None

def login():
    st.title("🔒 Login")
    with st.form("login_form"):
        username = st.text_input("Username")
        password = st.text_input("Password", type="password")
        submit = st.form_submit_button("Login")
        
        if submit:
            try:
                user = db.authenticate_user(username, password)
                if user:
                    st.session_state['logged_in'] = True
                    st.session_state['role'] = user[0]
                    st.session_state['department'] = user[1]
                    st.session_state['username'] = username
                    st.rerun()
                else:
                    st.error("Invalid username or password")
            except Exception as e:
                st.error(f"Database error: {e}. Please run generate_data.py and db.py.")

def logout():
    st.session_state['logged_in'] = False
    st.session_state['role'] = None
    st.session_state['department'] = None
    st.session_state['username'] = None
    st.rerun()

if not st.session_state['logged_in']:
    login()
else:
    st.sidebar.markdown(f"**User:** {st.session_state['username']}")
    st.sidebar.markdown(f"**Role:** {st.session_state['role']}")
    st.sidebar.button("Logout", on_click=logout)
    
    st.title("🎓 Student Performance Predictor Dashboard")
    
    # Load Data
    try:
        df = db.get_all_students()
        
        # Filter data based on role
        if st.session_state['role'] == 'HOD' and st.session_state['department'] != 'All':
            df = df[df['department'] == st.session_state['department']]
            
    except Exception as e:
        st.error(f"Database not found. Please run the setup scripts. Error: {e}")
        df = pd.DataFrame()
        
    if not df.empty:
        # Priority Queue
        st.subheader("🚨 Priority Queue: Top Students Needing Urgent Attention")
        critical_df = df[df['risk_label'] == 'Critical Risk'].sort_values(by='risk_score', ascending=False).head(20)
        
        if not critical_df.empty:
            st.dataframe(critical_df[['student_id', 'name', 'department', 'semester', 'risk_score', 'consecutive_absences']], width='stretch')
        else:
            st.success("No students in critical risk category.")
            
        st.divider()
        
        st.header("Overview Metrics")
        total_students = len(df)
        high_risk_students = len(df[df['risk_label'].isin(['High Risk', 'Critical Risk'])])
        avg_attendance = df['attendance_percentage'].mean()
        avg_marks = df['internal_marks'].mean()
        
        col1, col2, col3, col4 = st.columns(4)
        with col1:
            st.metric("Total Students", f"{total_students}")
        with col2:
            st.metric("High/Critical Risk", f"{high_risk_students}")
        with col3:
            st.metric("Average Attendance", f"{avg_attendance:.1f}%")
        with col4:
            st.metric("Average Internal Marks", f"{avg_marks:.1f}")
            
        st.divider()
        
        col_chart1, col_chart2 = st.columns(2)
        with col_chart1:
            st.subheader("Risk Distribution")
            risk_counts = df['risk_label'].value_counts().reset_index()
            risk_counts.columns = ['Risk Level', 'Count']
            color_map = {"Low Risk": "#28a745", "Moderate Risk": "#ffc107", "High Risk": "#fd7e14", "Critical Risk": "#dc3545"}
            fig = px.pie(risk_counts, values='Count', names='Risk Level', color='Risk Level', color_discrete_map=color_map, hole=0.4)
            st.plotly_chart(fig, width='stretch')
            
        with col_chart2:
            st.subheader("Risk by Department")
            dept_risk = df.groupby(['department', 'risk_label']).size().reset_index(name='Count')
            fig2 = px.bar(dept_risk, x='department', y='Count', color='risk_label', color_discrete_map=color_map, barmode='stack')
            st.plotly_chart(fig2, width='stretch')
            
        st.divider()
        
        # ERP Upload
        st.subheader("📤 College ERP Upload")
        uploaded_file = st.file_uploader("Upload CSV containing new student data", type=['csv'])
        if uploaded_file is not None:
            new_df = pd.read_csv(uploaded_file)
            st.write(f"Uploaded {len(new_df)} records. In production, this would parse and append to the database.")
            st.dataframe(new_df.head(), width='stretch')
            
    else:
        st.info("No data available.")