import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.ensemble import IsolationForest
from sklearn.preprocessing import StandardScaler

# Set page title and icon
st.set_page_config(page_title="Anomaly Detection App", page_icon="🔍")

# Custom CSS for better styling
st.markdown("""
<style>
    .stButton>button {
        background-color: #4CAF50;
        color: white;
        font-weight: bold;
        border-radius: 5px;
        padding: 10px 20px;
    }
    .stDownloadButton>button {
        background-color: #008CBA;
        color: white;
        font-weight: bold;
        border-radius: 5px;
        padding: 10px 20px;
    }
    .stMarkdown h1 {
        color: #4CAF50;
    }
    .stMarkdown h2 {
        color: #008CBA;
    }
</style>
""", unsafe_allow_html=True)

# Title of the app
st.title("🔍 Anomaly Detection App")
st.write("""
This app uses the **Isolation Forest** algorithm to detect anomalies in your dataset.
Upload a CSV file, and the app will identify anomalies in the data.
""")

# Upload dataset
uploaded_file = st.file_uploader("Upload your dataset (CSV file)", type=["csv"])

if uploaded_file is not None:
    # Load the dataset
    df = pd.read_csv(uploaded_file)
    
    # Show dataset preview
    st.write("### Dataset Preview")
    st.write(df.head())

    # Select features for anomaly detection
    st.write("### Select Features")
    features = st.multiselect("Choose the features to use for anomaly detection", df.columns)

    if features:
        # Allow user to adjust contamination parameter
        st.write("### Adjust Model Parameters")
        contamination = st.slider("Contamination (proportion of anomalies)", 0.01, 0.5, 0.1, 0.01)
        
        # Preprocess the data
        scaler = StandardScaler()
        df_scaled = scaler.fit_transform(df[features])

        # Train the Isolation Forest model
        with st.spinner("Training the model and detecting anomalies..."):
            model = IsolationForest(n_estimators=100, contamination=contamination, random_state=42)
            model.fit(df_scaled)

            # Predict anomalies
            predictions = model.predict(df_scaled)
            df['anomaly'] = predictions  # -1 for anomaly, 1 for normal

        # Display results
        st.write("### Anomaly Detection Results")
        st.write(df)

        # Filter and display only anomalies
        anomalies = df[df['anomaly'] == -1]
        st.write(f"### Detected Anomalies (Total: {len(anomalies)})")
        st.write(anomalies)

        # Visualize anomalies
        st.write("### Visualize Anomalies")
        if len(features) >= 2:
            fig, ax = plt.subplots()
            sns.scatterplot(data=df, x=features[0], y=features[1], hue='anomaly', palette={1: 'blue', -1: 'red'})
            st.pyplot(fig)
        else:
            st.warning("Please select at least 2 features to visualize anomalies.")

        # Download results as CSV
        st.write("### Download Results")
        st.download_button(
            label="Download Results as CSV",
            data=df.to_csv(index=False).encode('utf-8'),
            file_name='anomaly_detection_results.csv',
            mime='text/csv',
        )
else:
    st.write("Please upload a CSV file to get started.")