Consider a kd-tree, just like as kd-tree definition in Mark de. Berg's computational geometry textbook. There are two different analyses of cache-complexity of query answering in the attached pdf file? Which analysis is the right one? Which one is more accurate?