import streamlit as st
import pandas as pd
import numpy as np
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler, OrdinalEncoder, OneHotEncoder
from sklearn.impute import SimpleImputer
from sklearn.compose import ColumnTransformer
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeClassifier
from sklearn.svm import SVC
from sklearn.linear_model import LogisticRegression
from sklearn.neighbors import KNeighborsClassifier

# Set dark theme and page config
st.set_page_config(page_title="Cancer Prediction", page_icon="🩺", layout="centered")
st.markdown("""
    <style>
    body { background-color: #121212; color: white; }
    .stButton>button { background-color: #ff4b4b; color: white; }
    .stSelectbox, .stRadio, .stNumberInput, .stSlider { color: white; }
    </style>
""", unsafe_allow_html=True)

# Load dataset
def load_data():
    df = pd.read_csv('cancer_prediction_data (2).csv')
    return df

# Data Preprocessing
def preprocess_data(df):
    numeric = ['Age', 'Tumor_Size']
    ordinal = ['Tumor_Grade', 'Symptoms_Severity', 'Alcohol_Consumption', 'Exercise_Frequency']
    nominal = ['Gender', 'Family_History', 'Smoking_History']
    
    numeric_preprocess = Pipeline([
        ('imputer', SimpleImputer(strategy='mean')),
        ('scaler', StandardScaler())
    ])
    ordinal_preprocess = Pipeline([
        ('imputer', SimpleImputer(strategy='most_frequent')),
        ('encoder', OrdinalEncoder(handle_unknown="use_encoded_value", unknown_value=-1))
    ])
    nominal_preprocess = Pipeline([
        ('imputer', SimpleImputer(strategy='most_frequent')),
        ('encoder', OneHotEncoder(sparse_output=False, handle_unknown='ignore'))
    ])
    
    preprocess = ColumnTransformer([
        ('num', numeric_preprocess, numeric),
        ('ord', ordinal_preprocess, ordinal),
        ('nom', nominal_preprocess, nominal)
    ], remainder='passthrough')
    
    X = df.drop('Cancer_Present', axis=1)
    y = df['Cancer_Present']
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=23)
    return X_train, X_test, y_train, y_test, preprocess

# Train Models
def train_model(X_train, y_train, preprocess, model_name):
    models = {
        'Decision Tree': DecisionTreeClassifier(),
        'SVM': SVC(),
        'Logistic Regression': LogisticRegression(),
        'KNN': KNeighborsClassifier()
    }
    
    model = models[model_name]
    pipeline = Pipeline([
        ('preprocessor', preprocess),
        ('classifier', model)
    ])
    pipeline.fit(X_train, y_train)
    return pipeline

# Streamlit UI
st.markdown("<h1 style='text-align: center; color: red;'>🩺 Cancer Prediction 🩺</h1>", unsafe_allow_html=True)
st.write("An intelligent system for early cancer detection using machine learning.")

df = load_data()
X_train, X_test, y_train, y_test, preprocess = preprocess_data(df)

st.subheader("🔬 Choose a Machine Learning Model")
model_name = st.selectbox("", ['Decision Tree', 'SVM', 'Logistic Regression', 'KNN'])

if st.button("🚀 Train & Evaluate Model"):
    model = train_model(X_train, y_train, preprocess, model_name)
    accuracy = model.score(X_test, y_test)
    st.success(f"Model trained successfully! Accuracy: {accuracy:.2f}")
    st.session_state['trained_model'] = model

# Prediction Section
st.markdown("<h2 style='color: red;'>🔍 Make a Prediction</h2>", unsafe_allow_html=True)
age = st.number_input("📅 Age", min_value=18, max_value=100, value=30)
tumor_size = st.number_input("🧬 Tumor Size (cm)", min_value=1.0, max_value=10.0, value=5.0)
smoking_history = st.radio("🚬 Smoking History", ['Non-Smoker', 'Former Smoker', 'Current Smoker'])
alcohol_consumption = st.selectbox("🍷 Alcohol Consumption", ['None', 'Low', 'Moderate', 'High'])

tumor_grade = st.selectbox("Tumor Grade", [1, 2, 3])
symptoms_severity = st.selectbox("Symptoms Severity", [1, 2, 3])
exercise_frequency = st.selectbox("Exercise Frequency", ['Never', 'Rarely', 'Occasionally', 'Regularly'])
gender = st.radio("Gender", ['Male', 'Female'])
family_history = st.radio("Family History", ['No', 'Yes'])

input_data = [[age, tumor_size, tumor_grade, symptoms_severity, smoking_history, 
               alcohol_consumption, exercise_frequency, gender, family_history]]

if st.button("🔮 Predict Cancer Presence"):
    if 'trained_model' in st.session_state:
        model = st.session_state['trained_model']
        input_df = pd.DataFrame(input_data, columns=X_train.columns)
        input_transformed = model.named_steps['preprocessor'].transform(input_df)
        prediction = model.named_steps['classifier'].predict(input_transformed)
        
        st.write("Cancer Prediction:", "✅ Positive" if prediction[0] == 1 else "❌ Negative")
    else:
        st.error("Please train a model first!")