Matrix 分层聚类输出比输入少一个

所以，我的聚类输入有分类数据，这就是为什么我计算gowers距离并使用矩阵作为距离数组的输入，我用它来计算聚类。问题是，在我的原始数据和矩阵中，我有707个实例，但输出只有706个。我做错了什么

data_gower = gower.gower_matrix(original_data)

distArray = ssd.squareform(data_gower)

dendrogram = sch.dendrogram(sch.linkage(distArray, method  = "ward"))

clusters_hier = scipy.cluster.hierarchy.linkage(distArray, method='single',  metric='euclidean')
cluster = AgglomerativeClustering(n_clusters=7, affinity='euclidean', linkage='ward')  
clusters_hierarchical = cluster.fit_predict(clusters_hier) 


# Output
# AgglomerativeClustering(affinity='euclidean', compute_full_tree='auto',
                        connectivity=None, distance_threshold=None,
                        linkage='ward', memory=None, n_clusters=7,
                        pooling_func='deprecated')
# Out[143]:
# {0: 163, 1: 96, 2: 80, 3: 81, 4: 58, 5: 106, 6: 122} # instances per clusters
# 706 rows