import streamlit as st
import pandas as pd
from sklearn.linear_model import LinearRegression
import matplotlib.pyplot as plt

# Page configuration
st.set_page_config(page_title="Crime Rate Prediction", layout="wide")
st.title("📊 Crime Rate Prediction Based on Past Data")

# CSV path (ensure the file is accessible or uploaded in cloud deployment)
csv_path = "https://huggingface.co/spaces/MLDeveloper/crime_rate_predicition/resolve/main/RS_Session_255_AS_116.1%20(2).csv"

try:
    # Load dataset
    df = pd.read_csv(csv_path)

    st.subheader("📄 Raw Dataset")
    st.dataframe(df)

    # Preprocessing
    data = df[[
        'State/UT',
        'Number of Cases Registered - 2018-19',
        'Number of Cases Registered - 2019-20',
        'Number of Cases Registered - 2020-21',
        'Number of Cases Registered - 2021-22 (up to 31.10.2021)'
    ]].copy()
    data.columns = ['State/UT', '2018', '2019', '2020', '2021']

    # Convert to numeric
    for col in ['2018', '2019', '2020', '2021']:
        data[col] = pd.to_numeric(data[col], errors='coerce').fillna(0).astype(int)

    # Sidebar input
    st.sidebar.header("🔍 Predict Future Crime")
    selected_state = st.sidebar.selectbox("Select a State/UT", data['State/UT'].unique())
    start_year = st.sidebar.slider("Select a year to predict", 2022, 2027, 2022)

    # Filter and train model
    selected_row = data[data['State/UT'] == selected_state].iloc[0]
    years = [2018, 2019, 2020, 2021]
    X_train = pd.DataFrame({'Year': years})
    y_train = selected_row[['2018', '2019', '2020', '2021']].values

    model = LinearRegression()
    model.fit(X_train, y_train)

    # Predict future crime rates
    future_years = list(range(2022, 2028))
    predictions = model.predict(pd.DataFrame({'Year': future_years}))
    result_df = pd.DataFrame({
        'Year': future_years,
        'Predicted Crime Cases': [max(0, int(pred)) for pred in predictions]
    })

    # Display single year result
    selected_year_prediction = result_df[result_df['Year'] == start_year]['Predicted Crime Cases'].values[0]
    st.success(f"📌 **Predicted Crime Rate in {selected_state} for the year {start_year}: {selected_year_prediction} cases**")

    # Show full table
    st.subheader(f"📈 Predicted Crime Rate in {selected_state} (2022–2027)")
    st.dataframe(result_df)

    # Line chart
    fig, ax = plt.subplots()
    ax.plot(result_df['Year'], result_df['Predicted Crime Cases'], marker='o', linestyle='--', color='teal')
    ax.set_xlabel("Year")
    ax.set_ylabel("Predicted Crime Cases")
    ax.set_title(f"Crime Trend Prediction for {selected_state}")
    st.pyplot(fig)

except FileNotFoundError:
    st.error(f"❌ File not found at path: {csv_path}. Please check the path.")
except Exception as e:
    st.error(f"⚠️ An error occurred: {e}")