src/streamlit_app.py

import os
import pandas as pd
import streamlit as st

from sklearn.compose import ColumnTransformer
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error, r2_score
from sklearn.model_selection import train_test_split
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import OneHotEncoder, StandardScaler


# -------------------------------------------------
# SAYFA AYARLARI
# -------------------------------------------------
st.set_page_config(
    page_title="Araç Fiyat Tahmin Sistemi",
    page_icon="🚗",
    layout="centered"
)


# -------------------------------------------------
# DOSYA YOLU
# app.py: /app/src/app.py
# cars.xlsx: /app/cars.xlsx
# -------------------------------------------------
BASE_DIR = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
DATA_PATH = os.path.join(BASE_DIR, "cars.xlsx")


# -------------------------------------------------
# VERİYİ YÜKLE
# -------------------------------------------------
@st.cache_data
def load_data():
    data = pd.read_excel(DATA_PATH)

    required_columns = [
        "Price", "Make", "Model", "Trim", "Mileage",
        "Type", "Cylinder", "Liter", "Doors",
        "Cruise", "Sound", "Leather"
    ]

    missing_columns = [
        column for column in required_columns
        if column not in data.columns
    ]

    if missing_columns:
        raise ValueError(
            f"cars.xlsx içinde şu sütunlar eksik: {', '.join(missing_columns)}"
        )

    data = data.dropna()

    return data


# -------------------------------------------------
# MODELİ EĞİT
# -------------------------------------------------
@st.cache_resource
def train_model(data):
    x = data.drop("Price", axis=1)
    y = data["Price"]

    x_train, x_test, y_train, y_test = train_test_split(
        x,
        y,
        test_size=0.20,
        random_state=42
    )

    preprocessor = ColumnTransformer(
        transformers=[
            (
                "numeric",
                StandardScaler(),
                ["Mileage", "Cylinder", "Liter", "Doors"]
            ),
            (
                "categorical",
                OneHotEncoder(handle_unknown="ignore"),
                ["Make", "Model", "Trim", "Type"]
            )
        ],
        remainder="passthrough"
    )

    pipeline = Pipeline(
        steps=[
            ("preprocessor", preprocessor),
            ("regressor", LinearRegression())
        ]
    )

    pipeline.fit(x_train, y_train)

    predictions = pipeline.predict(x_test)

    rmse = mean_squared_error(y_test, predictions) ** 0.5
    r2 = r2_score(y_test, predictions)

    return pipeline, rmse, r2


# -------------------------------------------------
# FİYAT TAHMİNİ
# -------------------------------------------------
def predict_price(
    make,
    car_model,
    trim,
    mileage,
    car_type,
    cylinder,
    liter,
    doors,
    cruise,
    sound,
    leather
):
    input_data = pd.DataFrame([{
        "Make": make,
        "Model": car_model,
        "Trim": trim,
        "Mileage": mileage,
        "Type": car_type,
        "Cylinder": cylinder,
        "Liter": liter,
        "Doors": doors,
        "Cruise": cruise,
        "Sound": sound,
        "Leather": leather
    }])

    price = pipeline.predict(input_data)[0]

    return float(price)


# -------------------------------------------------
# VERİ VE MODELİ BAŞLAT
# -------------------------------------------------
try:
    df = load_data()
    pipeline, rmse, r2 = train_model(df)

except FileNotFoundError:
    st.error("cars.xlsx dosyası ana klasörde bulunamadı.")
    st.info("Dosya yapısı şu şekilde olmalı: ana klasörde cars.xlsx, src içinde app.py.")
    st.stop()

except Exception as error:
    st.error(f"Uygulama başlatılırken hata oluştu: {error}")
    st.stop()


# -------------------------------------------------
# TASARIM
# -------------------------------------------------
st.markdown("""
<style>
.main-title {
    text-align: center;
    font-size: 40px;
    font-weight: 800;
    margin-bottom: 5px;
}

.sub-title {
    text-align: center;
    font-size: 17px;
    color: #6b7280;
    margin-bottom: 25px;
}

.result-card {
    padding: 24px;
    border-radius: 16px;
    text-align: center;
    font-size: 28px;
    font-weight: 800;
    background: #ecfdf5;
    border: 1px solid #86efac;
    color: #166534 !important;
    margin-top: 15px;
    margin-bottom: 20px;
}
}
</style>
""", unsafe_allow_html=True)

st.markdown(
    "<div class='main-title'>🚗 Araç Fiyat Tahmin Sistemi</div>",
    unsafe_allow_html=True
)

st.markdown(
    "<div class='sub-title'>Araç bilgilerini girin, model tahmini satış fiyatını hesaplasın.</div>",
    unsafe_allow_html=True
)

st.divider()


# -------------------------------------------------
# FORM
# -------------------------------------------------
left_column, right_column = st.columns(2)

with left_column:
    make = st.selectbox(
        "Marka",
        sorted(df["Make"].unique())
    )

    available_models = sorted(
        df[df["Make"] == make]["Model"].unique()
    )

    car_model = st.selectbox(
        "Model",
        available_models
    )

    available_trims = sorted(
        df[
            (df["Make"] == make) &
            (df["Model"] == car_model)
        ]["Trim"].unique()
    )

    trim = st.selectbox(
        "Donanım Paketi",
        available_trims
    )

    mileage = st.number_input(
        "Kilometre",
        min_value=0,
        max_value=500000,
        value=30000,
        step=1000
    )

    car_type = st.selectbox(
        "Araç Tipi",
        sorted(df["Type"].unique())
    )

with right_column:
    cylinder = st.selectbox(
        "Silindir Sayısı",
        sorted(df["Cylinder"].unique())
    )

    liter = st.selectbox(
        "Motor Hacmi",
        sorted(df["Liter"].unique())
    )

    doors = st.selectbox(
        "Kapı Sayısı",
        sorted(df["Doors"].unique())
    )

    cruise = st.radio(
        "Hız Sabitleyici",
        options=[0, 1],
        horizontal=True,
        format_func=lambda value: "Var" if value == 1 else "Yok"
    )

    sound = st.radio(
        "Premium Ses Sistemi",
        options=[0, 1],
        horizontal=True,
        format_func=lambda value: "Var" if value == 1 else "Yok"
    )

    leather = st.radio(
        "Deri Koltuk",
        options=[0, 1],
        horizontal=True,
        format_func=lambda value: "Var" if value == 1 else "Yok"
    )

st.divider()


# -------------------------------------------------
# TAHMİN
# -------------------------------------------------
if st.button("Tahmini Fiyatı Hesapla", use_container_width=True):
    predicted_price = predict_price(
        make=make,
        car_model=car_model,
        trim=trim,
        mileage=mileage,
        car_type=car_type,
        cylinder=cylinder,
        liter=liter,
        doors=doors,
        cruise=cruise,
        sound=sound,
        leather=leather
    )

    st.markdown(
        f"<div class='result-card'>Tahmini Araç Fiyatı: ${predicted_price:,.2f}</div>",
        unsafe_allow_html=True
    )

    metric_1, metric_2 = st.columns(2)

    with metric_1:
        st.metric(
            "R² Başarı Skoru",
            f"{r2:.2f}"
        )

    with metric_2:
        st.metric(
            "RMSE Hata Skoru",
            f"${rmse:,.0f}"
        )

    st.subheader("Girilen Araç Bilgileri")

    result_table = pd.DataFrame([{
        "Marka": make,
        "Model": car_model,
        "Donanım": trim,
        "Kilometre": mileage,
        "Araç Tipi": car_type,
        "Silindir": cylinder,
        "Motor Hacmi": liter,
        "Kapı Sayısı": doors,
        "Hız Sabitleyici": "Var" if cruise == 1 else "Yok",
        "Premium Ses": "Var" if sound == 1 else "Yok",
        "Deri Koltuk": "Var" if leather == 1 else "Yok",
        "Tahmini Fiyat": f"${predicted_price:,.2f}"
    }])

    st.dataframe(
        result_table,
        use_container_width=True,
        hide_index=True
    )

st.divider()

st.caption(
    "Bu uygulama eğitim amaçlı bir makine öğrenmesi fiyat tahmin projesidir. "
    "Tahminler gerçek piyasa fiyatını garanti etmez."
)