ANOVA -testi ja Silhouette Score -testi

Raakadatan klusterointi.py

@@ -3,6 +3,8 @@ import pandas as pd
 from sklearn.cluster import KMeans 
 from sklearn.preprocessing import StandardScaler #StandardScaler datan normalisointi, eli skaalaus 
 import matplotlib.pyplot as plt #matplotlib visualisointiin
+from sklearn.metrics import silhouette_score
+from scipy.stats import f_oneway
 
 # Ladataan raakadata
 data = pd.read_csv("raakadata.csv")
@@ -44,3 +46,19 @@ print(indicators)
 summary = data[["Height", "Circumference", "BarkThickness", "PineNo", "NeedleNo"]].describe()
 
 print(summary)
+
+# Lasketaan Silhouette Score raakadataklusteroinnille
+silhouette_raw = silhouette_score(X_scaled, data["Cluster_Raw"])
+print(f"Raakadatan klusteroinnin Silhouette Score: {silhouette_raw:.3f}")
+
+# ANOVA-testit raakadatan klusteroinnille
+print("\nANOVA-testit raakadatan klusteroinnille:")
+
+for feature in features:
+    groups = []
+    for cluster_label in sorted(data["Cluster_Raw"].unique()):
+        group = data[data["Cluster_Raw"] == cluster_label][feature]
+        groups.append(group)
+    
+    stat, p_value = f_oneway(*groups)
+    print(f"{feature}: p-arvo = {p_value:.2e}")
\ No newline at end of file

Uusien ominaisuuksien klusterointi.py

@@ -4,6 +4,8 @@ from sklearn.cluster import KMeans
 from sklearn.preprocessing import StandardScaler # StandardScaler datan normalisointi, eli skaalaus 
 import matplotlib.pyplot as plt # matplotlib visualisointiin, lataa gitillä "pip install matplotlib"
 import numpy as np # Liian kaukaisten pisteiden poistamiseksi kuvaajasta
+from sklearn.metrics import silhouette_score
+from scipy.stats import f_oneway
 
 # Ladataan raakadata
 data = pd.read_csv("raakadata.csv")
@@ -99,6 +101,10 @@ clusters = kmeans.fit_predict(X_scaled)
 # Lisätään klusterit dataan
 data['Cluster'] = clusters
 
+# Lasketaan Silhouette Score uusien ominaisuuksien klusteroinnille
+silhouette_new_features = silhouette_score(X_scaled, clusters)
+print(f"Uusilla ominaisuuksilla klusteroidun datan Silhouette Score: {silhouette_new_features:.3f}")
+
 # PCA
 pca = PCA(n_components=2)
 X_pca = pca.fit_transform(X_scaled)
@@ -121,3 +127,11 @@ legend1 = plt.legend(*scatter.legend_elements(), title="Klusterit")
 plt.gca().add_artist(legend1)
 
 plt.show()
+
+anova_features = features  # käytetään samoja kuin klusteroinnissa
+
+print("\nANOVA-testit uusien ominaisuuksien klusteroinnille:")
+for feature in anova_features:
+    groups = [data[data['Cluster'] == cluster][feature] for cluster in range(4)]  # Klusterit 0–3
+    f_stat, p_value = f_oneway(*groups)
+    print(f"{feature}: p-arvo = {p_value:.2e}")
\ No newline at end of file