import pandas as pd
import numpy as np
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import classification_report
import io

def predict_engagement(file):
    """
    Predicts engagement levels from uploaded student logs CSV.
    Assumes a binary 'Engaged' column: 1 = Engaged, 0 = Not Engaged.
    
    Parameters:
        file (file-like): CSV file uploaded by user
    
    Returns:
        str: Prediction summary and performance metrics (if labeled)
    """
    try:
        df = pd.read_csv(file)

        if 'Engaged' not in df.columns:
            return "❌ CSV must include a binary column named 'Engaged' (1 or 0)."

        # Separate features and labels
        X = df.drop(columns=['Engaged'])
        y = df['Engaged']

        # Clean non-numeric columns
        X = X.select_dtypes(include=[np.number])

        # Train/test split
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

        # Train model
        model = RandomForestClassifier()
        model.fit(X_train, y_train)

        # Predict
        predictions = model.predict(X_test)
        report = classification_report(y_test, predictions, target_names=["Not Engaged", "Engaged"])

        return "📊 Engagement Prediction Report:\n\n" + report

    except Exception as e:
        return f"Error during prediction: {str(e)}"