Homework Assignment #3. 1 Demonstrate how mergesort works when sorting the following list of numbers:

Transcription

1 CISC 5835 Algorithms for Big Data Fall, 2018 Homework Assignment #3 1 Demonstrate how mergesort works when sorting the following list of numbers: Given the following array (list), follows the following pseudocode to partition the array using the first element, 6. For your reference, here is an article on quicksort with python code: // partition list A[startI...endI] using // A[startI] (first element in the sublist) Partition (A, starti, endi) 1. pivot = A[startI] //set first element as pivot value 2. left = starti+1; //left index 3. right = endi; //right index 4. do { 5. // increment left until we find a value larger than pivot or 5. // until left==right 5. while (a[left]<=pivot && left!=right) 6. left // decrement right until we find a value smaller than pivot or until 9. // lef==right 10. while (a[right]>=pivot && left!=right) 11. right-- 12.

2 13. if (left<right) 14. swap (a[left], a[right]) 15. } while (left<=right) if (right!=starti) 18. swap (a[right], a[starti]) return right //the index of pivot value in the array (1) Show the array after each time swap (line 14 above) is executed. (2). Read the book section on Selection (page 60), and comment on where the 2nd smallest element should be, i.e., the left partition or right parittion of the above partition result? 2

3 3 A positive integer N is a power if it is of the form q k, where q, k are positive integers and k > 1. (a) Give an efficient algorithm that takes as input a number N and deterimines whether it is a square, that is, whether it can be written as q 2 for some positive integer q. What s the running time of your algorithm? (b) Show that if N = q k (with N, q, and k all positive integers), then either k log 2 N or N = 1. (c) Give an efficient algorithm for determining whether a postive integer N is a power. Analyze its running time. 3

4 4 Suppose we want to evaluate the polynomial p(x) = a 0 + a 1 x + a 2 x a n x n at point x. Write a pseudocode of an efficient algorithms for this problem: given the coefficients a 0, a 1,..., a n and x, calculate p(x). Express the number additions and multiplications steps executed in the above function in terms of n, i.e., T (n). 4

5 5 Suppose a list P [1...n] gives the daily stock price of a certain company for a duration of n days, we want to find an optimal day d 1 to buy the stock and then a later day d 2 to sell the stock, so that the profit is maximized (i.e., P [d 2 ] P [d 1 ] is maximized. For example, if P [1...6] = 20, 10, 30, 50, 5, 14 Then we should buy in at day 2, and sell at day 4 to earn profit of per share. (Simply buying at the lowest point and selling at the highest point will not necessarily work, as one have to buy before sell.) (a) Think about how you solve this problem algorithmically, and write down the pseudocode and analyze the running time in terms of input size n (the length of the stock price data). Hint: A good starting point to any problem is try brute-force approach first, i.e., try all possible pairs of buy/sell days, and see which one yields largest profit. (b) Think about how to solve this problem more efficiently using divide-and-conquer paradigm, and describe your algorithm in pseudocode. Hint: Consider the following analysis of the problem, and translate this into a pseudocode. For an input of length n (i.e., prices of the stock for n consecutive days), the optimal way to make profit is the best among the following three possibilities: i. either the optimal buy/sell days are both in the first half of the n days (i.e., left half of the array). ii. or the optimal buy/sell days are both in the second half of the n days (i.e., left half of the array). iii. or the optimal days is buy in the first half of the n days, and sell in the second half of the n days. Note that for possibility 1 and 2 above, they are just solutions to the two subproblems (obtained by cutting the array into two halves), while the solution to the third possibility can be found by searching for the lowest price among first half of the array, and highest price among second half of the array. 5

6 (c) Implement both algorithms above, and test them using the the following data given in CVS file (downloaded from Yahoo finance), using the Open (second column) price for each data to calculate the optimal buy/sell dates. 6