关于质心的K均值聚类问题

Question

所以我在下面的脚本时遇到了麻烦，代码可以运行和编译，但是有一些问题。我不确定我的问题是计算问题还是格式问题。我什至不确定自己是否步入正轨，我感觉自己像是我，但我是python的新手，我非常愿意提供任何建议或建议！坚持以下几点：

1）K_means类根本不将质心分配给群集

2）K_means类不执行任何重新计算以使重心居中

下面是我的代码，我只是粘贴了函数，因为我认为这是问题所在：

class K_means():
    def __init__(self,k, lines, **kwargs):
        self.k = k
        self.lines = lines 
        self.index = []
        self.distance = []
        self.elements = []
        self.points = []
        for keys, values in kwargs.items():
            setattr(self, keys, values)

    def determineDistance(self, centroids, new_dict):

        for keys, points in new_dict.items():
            self.points.append(points)
            for elements in range(len(centroids)):
                self.elements.append(elements)
                distance = abs(points - centroids[elements])
                self.distance.append(distance)


    def assign_to_clusters(self, lines, clusters, centroids, 
        point_assignments, distance, new_dict):

        for index, points in new_dict.items():
            closest_index = (self.k + 1)
            for elements in range(len(centroids)):
                distance = self.determineDistance(centroids, new_dict)
                if distance < closest_index:
                    closest_index = distance
                    index = elements
            clusters[index].append(points)
            point_assignments[index] = closest_index
        return point_assignments

    def update_location(self, clusters, centroids):
        for keys, values in clusters.items():
            new_centroids ={keys:sum(values)/float(len(values))} 
            centroids.update(new_centroids)
        return centroids