两个一维稀疏矩阵之间的距离计算

在sklearn中没有找到能够直接针对稀疏矩阵进行聚类的方法，自己先写了个距离计算

def cal_distance(coo1,coo2):
    '''计算两个COO格式稀疏向量欧式距离'''
    common = set(arr1.col)&set(arr2.col)
    unique1 = set(arr1.col)-set(arr2.col)
    unique2 = set(arr2.col)-set(arr1.col)
    distance = 0
    for i,c in enumerate(arr1.col):
        if c in common:
            distance += (arr1.data[i]-arr1.data[arr1.col==c][0])**2
        elif c in unique1:
            distance += (arr1.data[i])**2
    for i,c in enumerate(arr2.col):
        if c in unique2:
            distance += (arr2.data[i])**2
        else:
            continue
    return distance