Both Kruskal and Prim algorithm work fine because they have the same time complexity O(E.log(V)).

Warning: for any Minimum Spanning Tree problems, use graph as list of [u,v,w] instead of adjacency list (hash table). MST requires graph sorting in ascending order of weights, so it's a suicide to use adjacency list. For example: use graph = [[1,2,16], [1, 3, 12]] instead of graph = {1: {2: 16}, 1: {3: 12}}

My Python Kruskal implementation here

Prim is the best suited as graph has too many edges. However Kruskal is also fine ;) 100 points are 100 points, no matter how you get it :P

For me, optimised Prim's with min priority queue worked.

Wrong answer on test cases 5 and 6 with two different implementations of the Prim's algorithm using Java 8: https://github.com/veniva/algo-minimal-network

why are comments blocked on problem 106?