K-Means Centroid Update

Recompute k-means centroids from cluster labels. labels[i] in $\{0,\dots ,K-1\}\backslash \{0,\backslash dots,K-1\backslash \}$ gives the cluster of row X[i]; each centroid is the mean of its cluster's points. A cluster with no points gets the zero vector.

Implement solve(X,labels,K).

Returns. Return a (K, d) float NumPy array of centroids.

Examples

Example 1

Input:

solve([[0.0, 0.0], [2.0, 0.0], [10.0, 10.0]], [0, 0, 1], 2)

Output:

[[1.0, 0.0], [10.0, 10.0]]

Example 2

Input:

solve([[1.0, 1.0], [3.0, 3.0]], [0, 0], 3)

Output:

[[2.0, 2.0], [0.0, 0.0], [0.0, 0.0]]