In some computer graphics and computer gaming applications, in order to save space, we might like to store a set of two-dimensional points in a single data structure that can be used for both nearest-neighbor queries and range searching. We have already discussed above how a k-d tree can be used to answer nearestneighbor queries with good expected-time behavior. Show that a (round-robin) k-d tree defined on n two-dimensional points can also be used to answer rangesearch queries in O(√n + s) time, where s is the number of points output by the query.