Please help me find a wrong case for my submission http://www.codechef.com/viewsolution/3937375 .

Its giving right answers for the test cases.

Can someone explain to me what “set-union” mean with respect to this question.

That problem can be solved by union find algorithm.

The algorithm maintains tree structures (using array) which gives you the different connected components.

We maintain an array `parent`

, such that `parent[i]`

stores parent node of `ith`

node, if `ith`

node itself is independent or if `ith`

node is a root of some connected component then `parent[i] = i`

.

We use one more array called `size`

, such that `size[i]`

stores the number of all children of `ith`

node (including ‘ith’ node)

initialize `parent[i] = i`

for all `i`

< `vertices.size()`

and `connected_component = 0`

now algorithm iterates through all edges,

```
for each edge from u to v in graph G:
u = root(u);
v = root(v);
if (u == v)
already connected;
else :
//this condition is used to make sure that our tree doest go too deep(instead we make it
//broad), if it goes too deep then time taken for finding root increases.
if (size[u] > size[v])
parent [v] = u;
else
parent [u] = v;
connected_components++;
```

now `connected_component`

gives number of independent graphs in a given graph, if we want you want to find the content of independent components then run a `dfs algorithm`

or `bfs algorithm`

using `parent`

array.

for more detailed explanation refer: [Alogorithms] [1]

[1]: http://algs4.cs.princeton.edu/15uf/