Unsupervised Learning:
Clustering & PCA

import pandas as pd
import numpy as np
from sklearn.cluster import KMeans, DBSCAN
from sklearn.decomposition import PCA
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import silhouette_score

np.random.seed(42)
n = 300
# Simulasi data pelanggan: frekuensi, nilai transaksi, recency
frekuensi = np.concatenate([
    np.random.normal(12, 2, 60),   # premium
    np.random.normal(4,  1, 160),   # reguler
    np.random.normal(0.5,0.3, 80)   # dormant
])
nilai = np.concatenate([
    np.random.normal(850, 100, 60),
    np.random.normal(210, 50,  160),
    np.random.normal(75,  20,  80)
])

df = pd.DataFrame({'frekuensi': frekuensi.clip(0), 'nilai_rb': nilai.clip(10)})

# ── Wajib: scaling sebelum clustering ────────────
sc = StandardScaler()
X = sc.fit_transform(df)

# ── K-Means: Elbow method ─────────────────────────
inertia = []
sil_scores = []
K_range = range(2, 8)
for k in K_range:
    km = KMeans(n_clusters=k, random_state=42, n_init=10).fit(X)
    inertia.append(km.inertia_)
    sil_scores.append(silhouette_score(X, km.labels_))

print("=== ELBOW METHOD ===")
for k, iner, sil in zip(K_range, inertia, sil_scores):
    print(f"K={k}: Inertia={iner:7.1f} | Silhouette={sil:.3f}")

# ── K-Means dengan K=3 ───────────────────────────
km3 = KMeans(n_clusters=3, random_state=42, n_init=10)
df['cluster'] = km3.fit_predict(X)

print("\n=== PROFIL CLUSTER K=3 ===")
profil = df.groupby('cluster').agg({
    'frekuensi': ['mean', 'count'],
    'nilai_rb': 'mean'
}).round(1)
print(profil)

# ── DBSCAN ───────────────────────────────────────
dbs = DBSCAN(eps=0.4, min_samples=5)
labels_dbs = dbs.fit_predict(X)
n_clusters = len(set(labels_dbs)) - (1 if -1 in labels_dbs else 0)
n_noise    = list(labels_dbs).count(-1)
print(f"\nDBSCAN: {n_clusters} cluster, {n_noise} noise points")

# ── PCA: visualisasi 2D ───────────────────────────
# Tambahkan fitur recency untuk demo PCA (3 dimensi → 2)
df['recency'] = np.where(df.cluster==0,
    np.random.normal(3,1,n), np.where(df.cluster==1,
    np.random.normal(18,5,n), np.random.normal(90,20,n)))
X3 = StandardScaler().fit_transform(df[['frekuensi','nilai_rb','recency']])

pca = PCA(n_components=2)
X_pca = pca.fit_transform(X3)
print(f"\n=== PCA: 3D → 2D ===")
print(f"Explained variance ratio: {pca.explained_variance_ratio_.round(3)}")
print(f"Total variansi tertangkap: {pca.explained_variance_ratio_.sum():.1%}")