Clustering performance evaluation(Davies-Bouldin Index) error

Question

I am trying to figure out the optimum number of clusters using various clustering performance evaluation methods. I put my data through a loop and ran DBA k-means. I am getting the elbow and silhouette results but the dunn index is showing error. Below is the code:

inertias = []
silhouette = []
davies_bouldin = []
clusters_range = range(1, 10)
    
for K in clusters_range:
        dba_km = TimeSeriesKMeans(n_clusters=K,
                          n_init=2,
                          metric="dtw",
                          verbose=True,
                          max_iter_barycenter=10,
                          random_state=seed)
        y_pred = dba_km.fit_predict(scaled_ts)
        
        
        
        inertias.append(dba_km.inertia_)
        
        if K > 1:
            silhouette.append(silhouette_score(scaled_ts, dba_km.labels_))
            
            davies_bouldin.append(davies_bouldin_score(scaled_ts, dba_km.labels_))

The error is showing on the davies_bouldin.append line:

TypeError: 'list' object is not callable.

Answer 1

this is my first post here: I'm still quite an amateur in python, so I came here looking for some answers myself and found your post. Turns out I've used your code to do something similar, with just a couple of minor adjustments and it worked perfectly! let me show you my adaptation:

wcss = []
sil_score = []
db_score = []
clusters_range = range(2, 16)

for K in clusters_range:
    kmeans = KMeans(n_clusters = K, max_iter = 300)
    kmeans.fit_predict(X2_pca_df)
    wcss.append(kmeans.inertia_)
    sil_score.append(silhouette_score(X2_pca_df, kmeans.labels_))
    db_score.append(davies_bouldin_score(X2_pca_df, kmeans.labels_))

Ps. Not sure when you posted this one, but hope it's not to late to give a hand. Cheers!