Union-find expressed as a social network

Ok to solve this exercise I did the assumption that the log file will look something like this:

0 1 2015-08-14 18:00:00
1 9 2015-08-14 18:01:00
0 2 2015-08-14 18:02:00
0 3 2015-08-14 18:04:00
0 4 2015-08-14 18:06:00
0 5 2015-08-14 18:08:00
0 6 2015-08-14 18:10:00
0 7 2015-08-14 18:12:00
0 8 2015-08-14 18:14:00
1 2 2015-08-14 18:16:00
1 3 2015-08-14 18:18:00
1 4 2015-08-14 18:20:00
1 5 2015-08-14 18:22:00
2 1 2015-08-14 18:24:00
2 3 2015-08-14 18:26:00
2 4 2015-08-14 18:28:00
5 5 2015-08-14 18:30:00

---

Where the 2 first numbers are the members who formed the friendship follow by the timestamp.

Another important thing to call out is that the exercise mention that the the file is sorted, so I have decide to sort it on a ascending order.

With this information, you can use the WeightedQuickUnionFind data structure provided in the class and simple process the file performing union operation on the members, once that you make the union you can ask for how many components are in the structure, if there is only one that means all members have equivalent relation.

Here is the code I did:

public static void main(String[] args) {

        int n = StdIn.readInt();
        WeightedQuickUnion uf = new WeightedQuickUnion(n);
        String date, time;
        //timestamps are sorted ascending
        while (!StdIn.isEmpty()) {

            int p = StdIn.readInt();
            int q = StdIn.readInt();
            date = StdIn.readString();
            time = StdIn.readString();


            uf.union(p, q);

            StdOut.println("["+p+","+q+"]");

            if(uf.getComponents() == 1){
                StdOut.println("All members were connected at: " + date + time);
                break;
            }

        }

---

The performance will be M lg N because you are iterating M times (the amount of lines in the log file) and the union operations takes: lg n.

When you add a friendship to the union-find datastructure, you can note if it results in two graph components being joined. Simply keep adding edges until N-1 of these merging events have happened.

In pseudo-code form:

G := UnionFind(1..N)
count := 0
for timestamp, p1, p2 in friendships {
    if G.Find(p1) != G.Find(p2) {
        G.Union(p1, p2)
        count++
        if count == N-1 {
            return timestamp
        }
    }
}
return +infinity