app.py

import streamlit as st
import pandas as pd
import numpy as np
import pickle
import os
import subprocess

# Ensure scikit-learn is installed
try:
    import sklearn
except ImportError:
    subprocess.run(["pip", "install", "scikit-learn", "--quiet"])
    import sklearn

from sklearn.linear_model import LinearRegression  # Now it should work
from sklearn.preprocessing import LabelEncoder

# Load dataset
@st.cache_data
def load_data():
    df = pd.read_csv("car_data.csv")  # Replace with your dataset
    return df

def preprocess_data(df):
    label_encoders = {}
    for column in ["brand", "model", "fuel_type"]:  # Encode categorical features
        le = LabelEncoder()
        df[column] = le.fit_transform(df[column])
        label_encoders[column] = le
    return df, label_encoders

# Train the model (Manual Data Split)
def train_model(df):
    X = df[["year", "mileage", "brand", "model", "fuel_type"]].values
    y = df["price"].values

    # Manual 80-20 split
    split_index = int(0.8 * len(X))  # 80% for training, 20% for testing
    X_train, X_test = X[:split_index], X[split_index:]
    y_train, y_test = y[:split_index], y[split_index:]

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

df = load_data()
df, label_encoders = preprocess_data(df)
model = train_model(df)

# Streamlit UI
st.title("Car Price Prediction App")
st.sidebar.header("Enter Car Details")

year = st.sidebar.slider("Year", int(df["year"].min()), int(df["year"].max()), 2015)
mileage = st.sidebar.number_input("Mileage (in km)", min_value=0, max_value=500000, value=50000)

brand = st.sidebar.selectbox("Brand", label_encoders["brand"].classes_)
filtered_models = df[df["brand"] == label_encoders["brand"].transform([brand])[0]]["model"].unique()
model_name = st.sidebar.selectbox("Model", label_encoders["model"].inverse_transform(filtered_models))
fuel_type = st.sidebar.selectbox("Fuel Type", label_encoders["fuel_type"].classes_)

if st.sidebar.button("Predict Price"):
    brand_encoded = label_encoders["brand"].transform([brand])[0]
    model_encoded = label_encoders["model"].transform([model_name])[0]
    fuel_encoded = label_encoders["fuel_type"].transform([fuel_type])[0]
    
    input_data = np.array([[year, mileage, brand_encoded, model_encoded, fuel_encoded]])
    predicted_price = model.predict(input_data)[0]
    
    st.write(f"### Predicted Car Price: ${predicted_price:,.2f}")