|
| 1 | +""" |
| 2 | +Problem Statement |
| 3 | +================= |
| 4 | +
|
| 5 | +Given a sorted array keys[0.. n-1] of search keys and an array freq[0.. n-1] of frequency counts, where freq[i] is the |
| 6 | +number of searches to keys[i]. Construct a binary search tree of all keys such that the total cost of all the searches |
| 7 | +is as small as possible. |
| 8 | +
|
| 9 | +Video |
| 10 | +----- |
| 11 | +* https://youtu.be/hgA4xxlVvfQ |
| 12 | +
|
| 13 | +Analysis |
| 14 | +-------- |
| 15 | +* Recursive: Exponential O(n^n) |
| 16 | +* Dynamic Programming: O(n^3) |
| 17 | +
|
| 18 | +Reference |
| 19 | +--------- |
| 20 | +* http://www.geeksforgeeks.org/dynamic-programming-set-24-optimal-binary-search-tree/ |
| 21 | +""" |
| 22 | + |
| 23 | + |
| 24 | +def min_cost_bst(input_array, freq): |
| 25 | + size = rows = cols = len(input_array) |
| 26 | + T = [[0 for _ in range(cols)] for _ in range(rows)] |
| 27 | + |
| 28 | + for idx in range(rows): |
| 29 | + T[idx][idx] = freq[idx] |
| 30 | + |
| 31 | + for sub_tree_size in range(2, size + 1): |
| 32 | + for start in range(size + 1 - sub_tree_size): |
| 33 | + end = start + sub_tree_size - 1 |
| 34 | + T[start][end] = float("inf") |
| 35 | + total = sum(freq[start:end + 1]) |
| 36 | + for k in range(start, end + 1): |
| 37 | + val = total + (0 if k - 1 < 0 else T[start][k - 1]) + (0 if k + 1 > end else T[k + 1][end]) |
| 38 | + T[start][end] = min(val, T[start][end]) |
| 39 | + |
| 40 | + return T[0][-1] |
| 41 | + |
| 42 | + |
| 43 | +def min_cost_bst_recursive_helper(input_array, freq, low_index, high_index, level): |
| 44 | + if low_index > high_index: |
| 45 | + return 0 |
| 46 | + min_value = float("inf") |
| 47 | + |
| 48 | + for index in range(low_index, high_index + 1): |
| 49 | + val = (min_cost_bst_recursive_helper(input_array, freq, low_index, index - 1, level + 1) # left tree |
| 50 | + + level * freq[index] # value at level |
| 51 | + + min_cost_bst_recursive_helper(input_array, freq, index + 1, high_index, level + 1)) # right tree |
| 52 | + min_value = min(val, min_value) |
| 53 | + |
| 54 | + return min_value |
| 55 | + |
| 56 | + |
| 57 | +def min_cost_bst_recursive(input_array, freq): |
| 58 | + return min_cost_bst_recursive_helper(input_array, freq, 0, len(input_array) - 1, 1) |
| 59 | + |
| 60 | + |
| 61 | +if __name__ == '__main__': |
| 62 | + input_array = [10, 12, 16, 21] |
| 63 | + freq = [4, 2, 6, 3] |
| 64 | + expected = 26 |
| 65 | + assert expected == min_cost_bst(input_array, freq) |
| 66 | + assert expected == min_cost_bst_recursive(input_array, freq) |
| 67 | + input_array = [10, 12, 20, 35, 46] |
| 68 | + freq = [34, 8, 50, 21, 16] |
| 69 | + expected = 232 |
| 70 | + assert expected == min_cost_bst(input_array, freq) |
| 71 | + assert expected == min_cost_bst_recursive(input_array, freq) |
0 commit comments