You are viewing a single comment's thread. Return to all comments →
There is no performance improvement in unfolding the 2d matrix in to 1d. Thats what you did here.
Well, he doesn't. Actually he does not care about "matrixes"(vectors of vectors), he simply looks in each simple vector by time, and pick two values (he already knows what values to pick, no verifying is needed), without relating one vector to another in a major matrix. Capisce?
It is a clever solution, dude!