Update src/streamlit_app.py

src/streamlit_app.py

@@ -4,360 +4,250 @@ import numpy as np
 from sklearn.preprocessing import StandardScaler
 from sklearn.neighbors import KNeighborsRegressor
 
-# ---------------------------
-# Page config & macOS-like styling
-# ---------------------------
-st.set_page_config(
-    page_title="Car Mod Performance Estimator — macOS Edition",
-    page_icon="🚘",
-    layout="wide"
-)
-
-st.markdown(
-    """
-    <style>
-    /* macOS-like system font and gentle background */
-    html, body, #root, .main {
-        font-family: -apple-system, "SF Pro Text", "Segoe UI", Roboto, "Helvetica Neue", Arial;
-        background: linear-gradient(180deg, #f7f8fa 0%, #eef1f6 100%);
-        color: #0b1220;
+# ---------------------------------------------------------
+# 1. macOS STYLING (CSS Injection)
+# ---------------------------------------------------------
+st.set_page_config(page_title="Mac-Mod Tuner", page_icon="🖥️", layout="wide")
+
+st.markdown("""
+<style>
+    /* Main Background - Apple Light Grey */
+    .stApp {
+        background-color: #f5f5f7;
+        font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif;
     }
-
-    /* Frosted panels */
-    .frost {
-        background: rgba(255,255,255,0.72);
-        border-radius: 14px;
-        padding: 18px;
-        box-shadow: 0 8px 24px rgba(14, 21, 47, 0.06);
-        border: 1px solid rgba(13, 22, 39, 0.04);
+    
+    /* Sidebar Styling */
+    section[data-testid="stSidebar"] {
+        background-color: #e8e8ed; /* Slightly darker grey */
+        border-right: 1px solid #d1d1d6;
     }
-
-    /* Headline */
-    .headline {
-        display:flex;
-        align-items:center;
-        gap:12px;
-        margin-bottom:6px;
+    
+    /* Card/Container Styling (The "Window" look) */
+    div[data-testid="stVerticalBlock"] > div[style*="background-color"] {
+        background-color: white;
+        border-radius: 18px;
+        padding: 20px;
+        box-shadow: 0 4px 20px rgba(0,0,0,0.05);
+        border: 1px solid #e5e5ea;
     }
-    .headline h1 { margin: 0; font-size: 1.6rem; }
-    .headline p { margin: 0; color:#6b7280; font-size:0.95rem; }
 
-    /* Small chips */
-    .chip {
-        display:inline-block;
-        padding:6px 10px;
-        margin:4px 6px 4px 0;
-        border-radius:999px;
-        background: rgba(14,165,233,0.10);
-        color:#0369a1;
-        border: 1px solid rgba(14,165,233,0.18);
-        font-size:0.85rem;
+    /* Input Widgets - Rounded pills */
+    .stSelectbox > div > div {
+        border-radius: 12px;
+        border: 1px solid #d1d1d6;
     }
-
-    /* subtle footer text */
-    .muted { color:#6b7280; font-size:0.9rem; }
-
-    /* metric cards adapt */
-    .metric-card {
-        background: linear-gradient(180deg, rgba(255,255,255,0.85), rgba(250,250,250,0.75));
+    .stNumberInput > div > div > input {
         border-radius: 12px;
-        padding: 12px;
-        box-shadow: 0 6px 20px rgba(14,21,47,0.04);
-        border: 1px solid rgba(13,22,39,0.03);
     }
-
-    </style>
-    """,
-    unsafe_allow_html=True
-)
-
-# ---------------------------
-# Synthetic data & model (unchanged logic)
-# ---------------------------
-def generate_dataset(n=400):
+    
+    /* Metrics Styling */
+    div[data-testid="stMetricValue"] {
+        font-weight: 600;
+        color: #1d1d1f;
+    }
+    
+    /* Headers */
+    h1, h2, h3 {
+        color: #1d1d1f;
+        font-weight: 700;
+        letter-spacing: -0.5px;
+    }
+    
+    /* Custom "Traffic Lights" for visual flair */
+    .traffic-lights {
+        display: flex;
+        gap: 8px;
+        margin-bottom: 20px;
+    }
+    .dot { width: 12px; height: 12px; border-radius: 50%; }
+    .red { background-color: #ff5f57; border: 1px solid #e0443e; }
+    .yellow { background-color: #febc2e; border: 1px solid #d89e24; }
+    .green { background-color: #28c840; border: 1px solid #1aab29; }
+</style>
+""", unsafe_allow_html=True)
+
+# ---------------------------------------------------------
+# 2. DATA & MODEL (Expanded for V10/V12 & Nitrous)
+# ---------------------------------------------------------
+
+@st.cache_resource
+def train_model():
     np.random.seed(42)
+    n = 1000 # More data for better accuracy with V12s
     data = []
+
     for _ in range(n):
-        engine = np.random.choice([1.6, 2.0, 2.5, 3.0, 3.5, 5.0])
-        cyl = np.random.choice([4, 6, 8])
-        base_hp = int(engine * cyl * np.random.uniform(18, 22))
-        intake = np.random.choice([0, 1, 2])          # stock/CAI/perf
-        exhaust = np.random.choice([0, 1, 2])         # stock/catback/straight
-        induction = np.random.choice([0, 1, 2])       # none/turbo/super
-        fuel = np.random.choice([0, 1, 2, 3])         # 87/91/93/E85
-        tune = np.random.choice([0, 1, 2])            # none/mild/aggressive
-        altitude = np.random.uniform(0, 2000)
-        hp_gain = (
-            intake * np.random.uniform(3, 10) +
-            exhaust * np.random.uniform(5, 20) +
-            induction * np.random.uniform(25, 100) +
-            tune * np.random.uniform(10, 35) +
-            fuel * np.random.uniform(2, 8) -
-            altitude * 0.01 +
-            np.random.uniform(-3, 3)
+        # Expanded Engine Options
+        engine = np.random.choice([1.6, 2.0, 2.4, 3.0, 3.8, 4.0, 5.0, 5.2, 6.0, 6.5])
+        cyl = np.random.choice([4, 6, 8, 10, 12])
+        
+        # Base HP logic (roughly 80-120 hp per liter depending on tech)
+        base_hp = int(engine * np.random.uniform(60, 110)) 
+        
+        # Mods
+        intake = np.random.choice([0, 1, 2])       # Stock/Sport/Race
+        exhaust = np.random.choice([0, 1, 2, 3])   # Stock/Catback/Headers/Straight
+        induction = np.random.choice([0, 1, 2, 3]) # None/Turbo/Twin-Turbo/Super
+        cams = np.random.choice([0, 1, 2])         # Stock/Street/Track
+        nitrous = np.random.choice([0, 1])         # No/Yes (50 shot equivalent)
+        fuel = np.random.choice([0, 1, 2, 3])      # 87/91/93/E85
+        tune = np.random.choice([0, 1, 2])         # None/Stage 1/Stage 2
+        
+        # Gain Logic
+        gain = (
+            (intake * 3) + 
+            (exhaust * 5) + 
+            (induction * (base_hp * 0.35)) +  # Forced induction is % based
+            (cams * (base_hp * 0.10)) + 
+            (nitrous * 50) + 
+            (tune * (base_hp * 0.08)) + 
+            (fuel * 4) +
+            np.random.uniform(-5, 5)
         )
-        data.append([engine, cyl, base_hp, intake, exhaust, induction, fuel, tune, altitude, hp_gain])
-    columns = ["engine", "cyl", "base_hp", "intake", "exhaust", "induction", "fuel", "tune", "altitude", "hp_gain"]
-    return pd.DataFrame(data, columns=columns)
-
-df = generate_dataset()
-
-X = df.drop("hp_gain", axis=1)
-y = df["hp_gain"]
-scaler = StandardScaler()
-X_scaled = scaler.fit_transform(X)
-
-model = KNeighborsRegressor(n_neighbors=5, weights='distance')
-model.fit(X_scaled, y)
-
-# ---------------------------
-# UI - Header
-# ---------------------------
-st.markdown(
-    """
-    <div class="headline">
-        <div style="font-size:1.6rem;">🚘</div>
-        <div>
-            <h1>Car Mod Performance — macOS UI</h1>
-            <p>Interactive estimator with extra tuning parameters — realistic, lightweight, and stylish.</p>
-        </div>
-    </div>
-    """,
-    unsafe_allow_html=True
-)
-
-# ---------------------------
-# Main layout: left (controls) and right (dashboard)
-# ---------------------------
-left, right = st.columns([1.05, 1])
-
-with left:
-    st.markdown('<div class="frost">', unsafe_allow_html=True)
-    st.subheader("🔧 Build & Tune")
-    col1, col2 = st.columns(2)
-
-    # Engine choices: displacement + configuration (I4..V12)
-    engine_disp_options = [0.8,1.0,1.2,1.4,1.6,1.8,2.0,2.2,2.4,2.5,2.8,3.0,3.2,3.5,4.0,4.4,5.0,6.0,8.0]
-    engine_disp = col1.selectbox("Engine Displacement (L)", engine_disp_options, index=engine_disp_options.index(2.0) if 2.0 in engine_disp_options else 0)
-
-    engine_config = col2.selectbox("Engine Layout", ["I4","I6","V6","V8","V10","V12"], index=0)
-    # Map layout -> cylinders
-    layout_to_cyl = {"I4":4,"I6":6,"V6":6,"V8":8,"V10":10,"V12":12}
-    cyl = layout_to_cyl[engine_config]
-
-    base_hp = st.number_input("Base Horsepower (stock)", min_value=60, max_value=1200, value=int(max(90, round(engine_disp * cyl * 20))), step=1)
-
-    # Vehicle weight & perception
-    weight_kg = st.number_input("Vehicle Weight (kg)", min_value=700, max_value=4000, value=1500)
-    weight_reduction = st.slider("Weight Reduction (%) — (mods / lightening)", 0, 40, 0)
-
-    st.markdown("---")
-    st.markdown("#### 🧩 Bolt-ons & Induction")
-
-    intake = st.selectbox("Intake", ["Stock", "Cold Air", "Performance"])
-    exhaust = st.selectbox("Exhaust", ["Stock", "Cat-back", "Straight Pipe"])
-    exhaust_dia = st.slider("Exhaust Diameter (mm)", 40, 120, 60)
-
-    induction = st.selectbox("Forced Induction", ["None", "Turbo", "Twin-Turbo", "Supercharger", "Twincharged"])
-    # boost only relevant if induction present
-    boost_psi = st.slider("Target Boost (psi) — (only if turbo/super)", 0, 40, 8)
-
-    turbo_size = st.selectbox("Turbo Size (mm) — (if applicable)", ["N/A","Shr. 45-50","Small 50-60","Medium 60-70","Large 70+"])
-    intercooler = st.selectbox("Intercooler", ["None","Air-to-Air","Air-to-Water"])
-    meth = st.checkbox("Methanol Injection (Wet Kit)", value=False)
-
-    st.markdown("---")
-    st.markdown("#### ⚙️ Internal & ECU")
-    cam = st.selectbox("Cam Profile", ["Stock", "Road", "Race"])
-    headers = st.selectbox("Headers", ["Stock", "Performance"])
-    intake_manifold = st.selectbox("Intake Manifold", ["Stock", "High-flow"])
-    tune = st.selectbox("ECU Tune Level", ["None", "Mild", "Aggressive"])
-
-    st.markdown("---")
-    st.markdown("#### ⛽ Fuel & Environment")
-    fuel = st.selectbox("Fuel Octane / Type", ["87","91","93","E85"])
-    altitude = st.slider("Altitude (meters)", 0, 3000, 200)
-
-    st.markdown("---")
-    st.markdown('<div class="muted">Tip: The ML model uses a synthetic dataset — results are approximate and for educational/estimation use only.</div>', unsafe_allow_html=True)
-    st.markdown('</div>', unsafe_allow_html=True)
-
-# ---------------------------
-# Map categorical to numeric (for model input)
-# ---------------------------
-intake_map = {"Stock":0, "Cold Air":1, "Performance":2}
-exhaust_map = {"Stock":0, "Cat-back":1, "Straight Pipe":2}
-# collapse twin-turbo and twincharged to turbo/supercharger categories for model
-induction_model_map = {"None":0, "Turbo":1, "Twin-Turbo":1, "Supercharger":2, "Twincharged":2}
-fuel_map = {"87":0, "91":1, "93":2, "E85":3}
-tune_map = {"None":0, "Mild":1, "Aggressive":2}
-
-input_for_model = np.array([[
-    engine_disp,
-    cyl,
-    base_hp,
-    intake_map.get(intake,0),
-    exhaust_map.get(exhaust,0),
-    induction_model_map.get(induction,0),
-    fuel_map.get(fuel,0),
-    tune_map.get(tune,0),
-    altitude
+        
+        # Diminishing returns for small engines with big mods
+        if cyl < 6 and gain > 200: 
+            gain *= 0.8
+
+        data.append([engine, cyl, base_hp, intake, exhaust, induction, cams, nitrous, fuel, tune, gain])
+
+    columns = ["engine", "cyl", "base_hp", "intake", "exhaust", "induction", "cams", "nitrous", "fuel", "tune", "hp_gain"]
+    df = pd.DataFrame(data, columns=columns)
+
+    X = df.drop("hp_gain", axis=1)
+    y = df["hp_gain"]
+    
+    scaler = StandardScaler()
+    X_scaled = scaler.fit_transform(X)
+    
+    model = KNeighborsRegressor(n_neighbors=7, weights='distance')
+    model.fit(X_scaled, y)
+    
+    return scaler, model
+
+scaler, model = train_model()
+
+# ---------------------------------------------------------
+# 3. SIDEBAR CONTROLS (The "Settings Pane")
+# ---------------------------------------------------------
+
+with st.sidebar:
+    st.header("⚙️ Configuration")
+    
+    with st.expander("🚙 Base Vehicle Stats", expanded=True):
+        col_eng_1, col_eng_2 = st.columns(2)
+        with col_eng_1:
+            cyl = st.selectbox("Cylinders", [4, 5, 6, 8, 10, 12])
+        with col_eng_2:
+            engine = st.selectbox("Size (L)", [1.6, 2.0, 2.5, 3.0, 3.5, 4.0, 5.0, 5.2, 6.0, 6.5, 8.4])
+            
+        base_hp = st.number_input("Factory HP", 100, 1000, 300)
+
+    with st.expander("🔧 Bolt-on Modifications", expanded=True):
+        intake = st.selectbox("Intake", ["Stock", "High Flow Filter", "Cold Air Intake"])
+        exhaust = st.selectbox("Exhaust", ["Stock", "Cat-back", "Long Tube Headers", "Full Straight Pipe"])
+
+    with st.expander("🔥 Internals & Boost", expanded=True):
+        induction = st.selectbox("Forced Induction", ["Naturally Aspirated", "Single Turbo", "Twin Turbo", "Supercharger"])
+        cams = st.selectbox("Camshafts", ["Stock", "Street Profile", "Track/Race Profile"])
+        nitrous = st.checkbox("Nitrous Oxide System (NOS)", value=False)
+
+    with st.expander("💻 Tuning & Fuel", expanded=True):
+        fuel = st.selectbox("Fuel Type", ["87 Octane", "91 Octane", "93 Octane", "E85 (Ethanol)"])
+        tune = st.selectbox("ECU Map", ["Stock Map", "Stage 1", "Stage 2"])
+
+# ---------------------------------------------------------
+# 4. MAIN DASHBOARD (The "App View")
+# ---------------------------------------------------------
+
+# Prediction Mappings
+intake_map = {"Stock":0, "High Flow Filter":1, "Cold Air Intake":2}
+exhaust_map = {"Stock":0, "Cat-back":1, "Long Tube Headers":2, "Full Straight Pipe":3}
+induction_map = {"Naturally Aspirated":0, "Single Turbo":1, "Twin Turbo":2, "Supercharger":3}
+cams_map = {"Stock":0, "Street Profile":1, "Track/Race Profile":2}
+nitrous_map = {False:0, True:1}
+fuel_map = {"87 Octane":0, "91 Octane":1, "93 Octane":2, "E85 (Ethanol)":3}
+tune_map = {"Stock Map":0, "Stage 1":1, "Stage 2":2}
+
+# Calculation
+input_data = np.array([[
+    engine, cyl, base_hp,
+    intake_map[intake], exhaust_map[exhaust], induction_map[induction],
+    cams_map[cams], nitrous_map[nitrous], fuel_map[fuel], tune_map[tune]
 ]])
 
-# Predict base gain from the trained model
-input_scaled = scaler.transform(input_for_model)
-pred_base = float(model.predict(input_scaled)[0])
-
-# ---------------------------
-# Heuristic extra gains from new advanced params
-# (We add these on top of model prediction to reflect advanced bolt-ons)
-# ---------------------------
-# cam, headers, intake manifold contributions
-cam_gain_map = {"Stock":0.0, "Road":5.0, "Race":12.0}
-headers_gain_map = {"Stock":0.0, "Performance":6.0}
-intake_manifold_gain = {"Stock":0.0, "High-flow":4.0}
-intercooler_gain_map = {"None":0.0, "Air-to-Air":4.0, "Air-to-Water":6.5}
-turbo_size_map = {"N/A":0.0, "Shr. 45-50":2.5, "Small 50-60":6.0, "Medium 60-70":12.0, "Large 70+":20.0}
-
-cam_gain = cam_gain_map.get(cam, 0.0)
-headers_gain = headers_gain_map.get(headers, 0.0)
-intake_man_gain = intake_manifold_gain.get(intake_manifold, 0.0)
-intercooler_gain = intercooler_gain_map.get(intercooler, 0.0)
-turbo_size_gain = turbo_size_map.get(turbo_size, 0.0)
-
-# boost contribution: if induction present, boost * factor; twin turbo gives more but is captured by induction choice
-if induction in ["Turbo", "Twin-Turbo"]:
-    boost_gain = boost_psi * 1.9  # psi-to-hp rough factor for turbo setups (heuristic)
-elif induction in ["Supercharger", "Twincharged"]:
-    boost_gain = boost_psi * 1.4  # superchargers typically have different curve
-else:
-    boost_gain = 0.0
-
-# methanol injection
-meth_gain = 10.0 if meth else 0.0
-
-# exhaust diameter small effect (larger diameter -> small gain if engine can flow)
-exhaust_dia_gain = max(0.0, (exhaust_dia - 55) * 0.08)
-
-# combined extra heuristic
-extra_gain = (
-    cam_gain +
-    headers_gain +
-    intake_man_gain +
-    intercooler_gain +
-    turbo_size_gain +
-    boost_gain * 0.9 +   # scale down a bit to avoid massive overestimates
-    meth_gain +
-    exhaust_dia_gain
-)
-
-# final predicted gain: model base + heuristic extra
-pred_total_gain = pred_base + extra_gain
-new_hp = base_hp + pred_total_gain
-
-# account for realistic minimums
-pred_total_gain = max(pred_total_gain, -5.0)  # avoid negative crazy values
-new_hp = max(new_hp, 30.0)
-
-# power-to-weight (hp per ton)
-effective_weight = weight_kg * (1 - weight_reduction / 100.0)
-hp_per_ton = new_hp / (effective_weight / 1000.0)
-
-# ---------------------------
-# Right panel: Dashboard (macOS style cards)
-# ---------------------------
-with right:
-    st.markdown('<div class="frost">', unsafe_allow_html=True)
-    st.subheader("📊 Performance Dashboard")
+pred = model.predict(scaler.transform(input_data))[0]
+new_hp = base_hp + pred
+pct_gain = (pred / base_hp) * 100
 
-    # Top metrics
-    k1, k2, k3 = st.columns(3)
-    k1.metric("Estimated HP Gain", f"{pred_total_gain:.1f} HP")
-    k2.metric("New Estimated Horsepower", f"{new_hp:.1f} HP")
-    k3.metric("HP / Ton", f"{hp_per_ton:.1f}")
+# --- UI LAYOUT ---
 
-    # Build quick summary chips
-    st.markdown(
-        f"""
-        <div style="margin-top:10px;">
-            <span class="chip">Engine: {engine_disp}L • {engine_config}</span>
-            <span class="chip">Intake: {intake}</span>
-            <span class="chip">Exhaust: {exhaust} ({exhaust_dia}mm)</span>
-            <span class="chip">Induction: {induction} • Boost: {boost_psi}psi</span>
-            <span class="chip">Tune: {tune}</span>
-            <span class="chip">Fuel: {fuel}</span>
-        </div>
-        """,
-        unsafe_allow_html=True
-    )
-
-    st.markdown("---")
-
-    # small two-column charts & info
-    left_panel, right_panel = st.columns([1, 1])
-
-    with left_panel:
-        st.markdown('<div class="metric-card">', unsafe_allow_html=True)
-        st.write("Power Curve Snapshot")
-        st.bar_chart(
-            pd.DataFrame(
-                {"HP": [base_hp, base_hp + pred_base, new_hp]},
-                index=["Stock", "Model Gain", "Final"]
-            )
-        )
-        st.markdown('</div>', unsafe_allow_html=True)
-
-    with right_panel:
-        st.markdown('<div class="metric-card">', unsafe_allow_html=True)
-        st.write("Tuning Contribution Breakdown")
-        breakdown = pd.DataFrame({
-            "component": ["Model base", "Cam", "Headers", "Intake Manifold", "Intercooler", "Turbo Size", "Boost", "Meth", "Exhaust Dia"],
-            "hp": [pred_base, cam_gain, headers_gain, intake_man_gain, intercooler_gain, turbo_size_gain, boost_gain * 0.9, meth_gain, exhaust_dia_gain]
-        })
-        st.dataframe(
-    breakdown.style.format("{:.1f}"),
-    height=220,
-    use_container_width=True
-    )
-
-
-    st.markdown("---")
-    # performance badge description
-    perf_text = "Balanced cruiser"
-    if pred_total_gain < 15:
-        perf_text = "Mild improvement — street friendly"
-    elif pred_total_gain < 50:
-        perf_text = "Noticeable power — spirited driving"
-    else:
-        perf_text = "Serious power — track-capable build"
-
-    st.markdown(f"### ⚡ Build verdict: **{perf_text}**")
-    st.write("Power-to-weight and HP gain are quick indicators — actual drivability depends on gearing, cooling, and reliability.")
-
-    st.markdown("<div class='muted'>Tip: This estimator combines a synthetic ML model plus simple heuristics for extra bolt-ons. For precise dyno numbers consult a professional tuner.</div>", unsafe_allow_html=True)
-
-    st.markdown("</div>", unsafe_allow_html=True)
-
-# ---------------------------
-# Optional: show dataset / debug
-# ---------------------------
-with st.expander("🔎 Peek training data & model info"):
-    st.write("Sample synthetic dataset (used to train the toy estimator):")
-    st.dataframe(df.head(10))
-    st.write("- Model: KNeighborsRegressor (distance weighted)")
-    st.write("- Additional bolt-on gains computed with lightweight heuristics")
-    st.write("- Use this for estimation and learning, not as a dyno replacement.")
-
-# ---------------------------
-# Footer
-# ---------------------------
-st.markdown(
-    """
-    <div style="margin-top:14px; text-align:center;">
-        <span class="muted">Made for learning — treat numbers as estimates. Enjoy tuning! 🍏</span>
+# 1. Header with Fake Window Controls
+st.markdown("""
+    <div class="traffic-lights">
+        <div class="dot red"></div>
+        <div class="dot yellow"></div>
+        <div class="dot green"></div>
     </div>
-    """,
-    unsafe_allow_html=True
-)
+""", unsafe_allow_html=True)
+
+st.title("Performance Estimator Pro")
+st.markdown(f"Analysis for **{cyl}-Cylinder {engine}L Engine**")
+
+st.divider()
+
+# 2. Main Hero Section (Glass Cards)
+col1, col2 = st.columns([1.5, 1])
+
+with col1:
+    st.markdown("### 📈 Dyno Projection")
+    # Area chart looks cleaner/more modern than bar
+    chart_data = pd.DataFrame({
+        "Horsepower": [base_hp, new_hp],
+        "Stage": ["Factory", "Modified"]
+    })
+    
+    # Custom color for the chart to match Apple blue
+    st.vega_lite_chart(chart_data, {
+        "mark": {"type": "bar", "cornerRadiusEnd": 4, "color": "#007AFF"},
+        "encoding": {
+            "x": {"field": "Stage", "type": "nominal", "axis": {"labelAngle": 0}},
+            "y": {"field": "Horsepower", "type": "quantitative"},
+            "tooltip": ["Stage", "Horsepower"]
+        }
+    }, use_container_width=True)
+
+with col2:
+    st.markdown("### ⚡ Results")
+    
+    # Container for the metrics
+    with st.container():
+        st.markdown(f"""
+        <div style="padding: 10px;">
+            <span style="font-size: 14px; color: #86868b; text-transform: uppercase; letter-spacing: 1px; font-weight: 600;">Total Power</span>
+            <div style="font-size: 48px; font-weight: 700; color: #1d1d1f; line-height: 1.2;">{int(new_hp)} HP</div>
+        </div>
+        """, unsafe_allow_html=True)
+        
+        st.divider()
+        
+        c1, c2 = st.columns(2)
+        with c1:
+            st.metric("Gain", f"+{int(pred)} HP")
+        with c2:
+            st.metric("Improvement", f"{pct_gain:.1f}%")
+
+# 3. Spec Summary Row
+st.markdown("### 📋 Build Summary")
+with st.container():
+    c1, c2, c3, c4 = st.columns(4)
+    c1.info(f"**Induction:** {induction}")
+    c2.info(f"**Fuel:** {fuel}")
+    c3.info(f"**Camshafts:** {cams}")
+    c4.info(f"**Nitrous:** {'Enabled' if nitrous else 'Disabled'}")
+
+if pct_gain > 50:
+    st.toast("🚀 Massive gains detected! Check transmission limits.", icon="⚠️")