# ====== K-Means Dasar (Data Sintetis) ======
# Jika perlu: !pip install scikit-learn
import numpy as np
import matplotlib.pyplot as plt
from sklearn.datasets import make_blobs
from sklearn.cluster import KMeans
from sklearn.preprocessing import StandardScaler

X, y_true = make_blobs(n_samples=600, centers=4, cluster_std=0.80, random_state=0)
sc = StandardScaler()
Xs = sc.fit_transform(X)

kmeans = KMeans(n_clusters=4, n_init=10, random_state=0)
labels = kmeans.fit_predict(Xs)

plt.figure(figsize=(5,4))
plt.scatter(Xs[:,0], Xs[:,1], c=labels, s=12)
plt.title('K-Means (K=4) pada Data Terskala')
plt.show()

print('Inersia:', kmeans.inertia_)
print('Centroid (scaled space):\n', kmeans.cluster_centers_)

# ====== Elbow Method & Silhouette Score ======
import numpy as np
import matplotlib.pyplot as plt
from sklearn.cluster import KMeans
from sklearn.metrics import silhouette_score
from sklearn.preprocessing import StandardScaler
from sklearn.datasets import make_blobs

X, _ = make_blobs(n_samples=800, centers=5, cluster_std=0.9, random_state=42)
Xs = StandardScaler().fit_transform(X)

Ks = range(2, 10)
inertias = []
sils = []
for k in Ks:
    km = KMeans(n_clusters=k, n_init=10, random_state=0).fit(Xs)
    inertias.append(km.inertia_)
    sils.append(silhouette_score(Xs, km.labels_))

plt.figure(figsize=(5,3))
plt.plot(list(Ks), inertias, marker='o')
plt.title('Elbow Method (Inertia vs K)')
plt.xlabel('K'); plt.ylabel('Inertia'); plt.grid(True, alpha=0.3)
plt.show()

plt.figure(figsize=(5,3))
plt.plot(list(Ks), sils, marker='o')
plt.title('Silhouette Score vs K')
plt.xlabel('K'); plt.ylabel('Silhouette'); plt.grid(True, alpha=0.3)
plt.show()

best_k = Ks[int(np.argmax(sils))]
print('K terbaik (menurut Silhouette) ≈', best_k)

# ====== PCA: Visualisasi 2D & Variance Explained ======
from sklearn.decomposition import PCA
from sklearn.preprocessing import StandardScaler
from sklearn.datasets import load_wine
import numpy as np
import matplotlib.pyplot as plt

X, y = load_wine(return_X_y=True)
Xs = StandardScaler().fit_transform(X)

pca = PCA(n_components=2)
X2 = pca.fit_transform(Xs)

print('Explained variance ratio (2 komponen):', pca.explained_variance_ratio_)

plt.figure(figsize=(5,4))
plt.scatter(X2[:,0], X2[:,1], c=y, s=12)
plt.title('PCA 2D pada Wine Dataset')
plt.xlabel('PC1'); plt.ylabel('PC2')
plt.show()

# Scree plot untuk beberapa komponen
pca_full = PCA().fit(Xs)
plt.figure(figsize=(5,3))
plt.plot(np.cumsum(pca_full.explained_variance_ratio_), marker='o')
plt.title('Cumulative Explained Variance')
plt.xlabel('Jumlah Komponen'); plt.ylabel('Var Cum.')
plt.grid(True, alpha=0.3)
plt.show()

# ====== Pipeline: Scaling → PCA (2D) → K-Means ======
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler
from sklearn.decomposition import PCA
from sklearn.cluster import KMeans
from sklearn.metrics import silhouette_score
from sklearn.datasets import load_iris
import matplotlib.pyplot as plt

X, _ = load_iris(return_X_y=True)
pipe = Pipeline([
    ('sc', StandardScaler()),
    ('pca', PCA(n_components=2)),
    ('km', KMeans(n_clusters=3, n_init=10, random_state=0))
])
Xt = pipe.fit_transform(X)  # output setelah PCA
labels = pipe.named_steps['km'].labels_

print('Silhouette (PCA 2D + KMeans K=3):', silhouette_score(Xt, labels))

plt.figure(figsize=(5,4))
plt.scatter(Xt[:,0], Xt[:,1], c=labels, s=14)
plt.title('Iris: PCA 2D + K-Means (K=3)')
plt.xlabel('PC1'); plt.ylabel('PC2')
plt.show()

# ====== Silhouette Plot (Visual) untuk K-Means ======
import numpy as np
import matplotlib.pyplot as plt
from sklearn.datasets import make_blobs
from sklearn.preprocessing import StandardScaler
from sklearn.cluster import KMeans
from sklearn.metrics import silhouette_samples, silhouette_score

X, _ = make_blobs(n_samples=700, centers=4, cluster_std=0.9, random_state=10)
Xs = StandardScaler().fit_transform(X)

k = 4
km = KMeans(n_clusters=k, n_init=10, random_state=0).fit(Xs)
labels = km.labels_

silhouette_avg = silhouette_score(Xs, labels)
sample_silhouette_values = silhouette_samples(Xs, labels)

print('Rata-rata Silhouette:', silhouette_avg)

y_lower = 10
plt.figure(figsize=(6,4))
for i in range(k):
    ith_cluster_silhouette_values = sample_silhouette_values[labels == i]
    ith_cluster_silhouette_values.sort()
    size_cluster_i = ith_cluster_silhouette_values.shape[0]
    y_upper = y_lower + size_cluster_i
    plt.fill_betweenx(np.arange(y_lower, y_upper), 0, ith_cluster_silhouette_values)
    y_lower = y_upper + 10
plt.axvline(x=silhouette_avg, color='k', linestyle='--')
plt.title('Silhouette Plot (K=4)')
plt.xlabel('Nilai Silhouette'); plt.ylabel('Sampel')
plt.show()

# ====== Kuis 13 (cek mandiri) ======
qs=[
  ("Tujuan K-Means adalah...",{"a":"Maksimalkan jarak antar-centroid","b":"Minimalkan inersia (sum of squared distances)","c":"Maksimalkan Silhouette","d":"Minimalkan MAE"},"b"),
  ("Mengapa scaling penting sebelum K-Means?",{"a":"Untuk mempercepat GPU","b":"Agar fitur berskala besar tidak mendominasi jarak","c":"Mengubah jumlah klaster","d":"Supaya PCA tidak perlu"},"b"),
  ("PCA memilih komponen berdasarkan...",{"a":"Korelasi dengan label","b":"Varians proyeksi","c":"Jumlah sampel","d":"Random"},"b"),
  ("Silhouette score tinggi menunjukkan...",{"a":"Klaster tumpang tindih","b":"Klaster kompak & terpisah baik","c":"Banyak outlier","d":"Tidak bisa ditentukan"},"b"),
]
print('Kunci jawaban:')
for i,(_,__,ans) in enumerate(qs,1):
    print(f'Q{i}: {ans}')