Multikolinieritas
& Implikasinya

import pandas as pd
import numpy as np
from statsmodels.stats.outliers_influence import variance_inflation_factor
from sklearn.linear_model import LinearRegression

np.random.seed(42)
n = 100
jam_belajar = np.random.normal(4, 1.5, n)
# Buat variabel yang sangat berkorelasi dengan jam_belajar
catatan     = jam_belajar * 1.2 + np.random.normal(0, 0.5, n)
les_tambahan= jam_belajar * 0.8 + np.random.normal(0, 0.3, n)
kehadiran   = np.random.normal(80, 10, n)  # independen
nilai_uts   = 40 + 7*jam_belajar + 0.3*kehadiran + np.random.normal(0,5,n)

df = pd.DataFrame({
    'jam_belajar'  : jam_belajar,
    'catatan'      : catatan,
    'les_tambahan' : les_tambahan,
    'kehadiran'    : kehadiran,
    'nilai_uts'    : nilai_uts
})

features = ['jam_belajar', 'catatan', 'les_tambahan', 'kehadiran']
X = df[features]

# ── Matriks korelasi antar prediktor ─────────────
print("=== KORELASI ANTAR PREDIKTOR ===")
print(X.corr().round(3))

# ── Hitung VIF ────────────────────────────────────
X_const = pd.DataFrame({'const': 1, **{c: X[c] for c in features}})
vif_data = pd.DataFrame({
    'variabel': features,
    'VIF': [variance_inflation_factor(X_const.values, i+1)
            for i in range(len(features))]
}).round(2)

print("\n=== VARIANCE INFLATION FACTOR ===")
print(vif_data)
print("\nVariabel dengan VIF > 5:")
print(vif_data[vif_data['VIF'] > 5])

from sklearn.linear_model import Ridge, Lasso, LinearRegression
from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error

X_all = df[features].values
y = df['nilai_uts'].values

X_train, X_test, y_train, y_test = train_test_split(
    X_all, y, test_size=0.2, random_state=42)

# Scaling wajib sebelum regularisasi
sc = StandardScaler()
X_train_s = sc.fit_transform(X_train)
X_test_s  = sc.transform(X_test)

models = {
    'OLS biasa': LinearRegression(),
    'Ridge (λ=1)': Ridge(alpha=1.0),
    'Lasso (λ=0.5)': Lasso(alpha=0.5)
}

print(f"{'Model':<18} {'RMSE Test':>12} {'Koefisien'}")
print('-'*70)
for name, model in models.items():
    model.fit(X_train_s, y_train)
    y_pred = model.predict(X_test_s)
    rmse = mean_squared_error(y_test, y_pred, squared=False)
    coef_str = ' | '.join([f"{features[i]}={c:.2f}"
                           for i,c in enumerate(model.coef_)])
    print(f"{name:<18} {rmse:>10.3f}  {coef_str}")