import streamlit as st
import pandas as pd
import plotly.express as px
import plotly.graph_objects as go
from sklearn.ensemble import IsolationForest
from sklearn.model_selection import train_test_split
from sklearn.metrics import classification_report

# Streamlit app
st.title("Advanced Transaction Anomaly Detection")

# File uploader
uploaded_file = st.file_uploader("Upload your CSV file", type="csv")

if uploaded_file:
    # Load the data
    data = pd.read_csv('transaction_anomalies_dataset (1).csv')
    st.subheader("Dataset Preview")
    st.write(data.head())

    # Data Overview
    st.subheader("Dataset Overview")
    st.write("Missing Values:")
    st.write(data.isnull().sum())
    st.write("Descriptive Statistics:")
    st.write(data.describe())

    # Visualization 1: Histogram of Transaction Amount
    if 'Transaction_Amount' in data.columns:
        st.subheader("Transaction Amount Distribution")
        fig_amount = px.histogram(data, x='Transaction_Amount', nbins=30, title="Transaction Amount Distribution")
        st.plotly_chart(fig_amount)

    # Visualization 2: Box Plot of Transaction Amount by Account Type
    if 'Account_Type' in data.columns and 'Transaction_Amount' in data.columns:
        st.subheader("Box Plot: Transaction Amount by Account Type")
        fig_box = px.box(data, x='Account_Type', y='Transaction_Amount', title="Transaction Amount by Account Type")
        st.plotly_chart(fig_box)

        # Check if 'Day_of_Week' column exists
        if 'Day_of_Week' in data.columns:
            # Create bar chart for transactions by day of the week
            fig_day_of_week = px.bar(data, x='Day_of_Week', title='Count of Transactions by Day of the Week')

            # Display the chart in the Streamlit app
            st.plotly_chart(fig_day_of_week)






    # Visualization 3: Correlation Heatmap (Plotly)
    st.subheader("Correlation Heatmap")
    numeric_cols = data.select_dtypes(include=['float64', 'int64'])
    if not numeric_cols.empty:
        corr_matrix = numeric_cols.corr()
        fig_heatmap = go.Figure(data=go.Heatmap(
            z=corr_matrix.values,
            x=corr_matrix.columns,
            y=corr_matrix.columns,
            colorscale='Viridis',
            hoverongaps=False,
        ))
        fig_heatmap.update_layout(title="Correlation Heatmap", xaxis_title="Features", yaxis_title="Features")
        st.plotly_chart(fig_heatmap)

    # Visualization 4: Scatter Plot (Age vs Average Transaction Amount)
    if 'Age' in data.columns and 'Average_Transaction_Amount' in data.columns:
        st.subheader("Scatter Plot: Age vs Average Transaction Amount")
        fig_scatter = px.scatter(data, x='Age',
                                        y='Average_Transaction_Amount',
                                        color='Account_Type',
                                        title='Average Transaction Amount vs. Age',
                                        trendline='ols')
        st.plotly_chart(fig_scatter)









        

    # Anomaly Detection with Isolation Forest
    st.subheader("Anomaly Detection")
    features = ['Transaction_Amount', 'Average_Transaction_Amount', 'Frequency_of_Transactions']

    # Ensure all required features are in the dataset
    if all(feature in data.columns for feature in features):
        X = data[features]

        # Train Isolation Forest
        st.write("Training Isolation Forest model...")
        model = IsolationForest(n_estimators=100, contamination=0.1, random_state=42)
        model.fit(X)

        # Add anomaly prediction column
        data['anomaly'] = model.predict(X)
        data['anomaly'] = data['anomaly'].apply(lambda x: 1 if x == -1 else 0)

        # Display Results
        st.write("Anomaly Detection Results:")
        st.write(data[['anomaly']].value_counts())

        # Visualization: Anomalies vs Normal Transactions
        st.subheader("Anomalies vs Normal Transactions")
        fig_anomalies = px.histogram(data, x='anomaly', title="Anomalies vs Normal Transactions",
                                     labels={'anomaly': 'Anomaly (1) vs Normal (0)'})
        st.plotly_chart(fig_anomalies)

        # User Input for Prediction
        st.subheader("Predict Anomaly for a New Transaction")
        user_inputs = {}
        for feature in features:
            user_input = st.number_input(f"Enter the value for '{feature}':", value=0.0)
            user_inputs[feature] = user_input

        # Create a DataFrame from user inputs
        user_df = pd.DataFrame([user_inputs])

        # Predict anomalies using the model
        user_anomaly_pred = model.predict(user_df)
        user_anomaly_pred_binary = 1 if user_anomaly_pred[0] == -1 else 0

        if user_anomaly_pred_binary == 1:
            st.error("Anomaly detected: This transaction is flagged as an anomaly.")
        else:
            st.success("No anomaly detected: This transaction is normal.")
    else:
        st.error("Required features for anomaly detection are missing in the dataset.")