11_k_means_neighborhood.py

from sklearn.metrics import silhouette_score

la_grouped_clustering = la_grouped.drop('Neighbourhood', 1)

for n_cluster in range(2, 12):
    kmeans = KMeans(n_clusters=n_cluster).fit(la_grouped_clustering)
    label = kmeans.labels_
    sil_coeff = silhouette_score(la_grouped_clustering, label, metric='euclidean')
    print("For n_clusters={}, The Silhouette Coefficient is {}".format(n_cluster, sil_coeff))
    
    
    # set number of clusters
kclusters = 4

la_grouped_clustering = la_grouped.drop('Neighbourhood', 1)

# run k-means clustering
kmeans = KMeans(n_clusters=kclusters, random_state=0).fit(la_grouped_clustering)

# check cluster labels generated for each row in the data frame
kmeans.labels_

# add clustering labels
Neighbourhoods_venues_sorted.insert(0, 'Cluster Label', kmeans.labels_.astype(int))
# Neighbourhoods_venues_sorted['Cluster Label']=kmeans.labels_.astype(int)
la_merged = nhoods

# merge la_grouped with nhoods to add latitude/longitude for each Neighbourhood
la_merged = la_merged.join(Neighbourhoods_venues_sorted.set_index('Neighbourhood'), on='Neighbourhood')
la_merged.dropna(inplace=True)
la_merged['Cluster Label'] = la_merged['Cluster Label'].astype(int)
la_merged.head()