[Leetcode] Binary search, DP--300. Longest Increasing Subsequence

Given an unsorted array of integers, find the length of longest increasing subsequence.

For example,
Given [10, 9, 2, 5, 3, 7, 101, 18],
The longest increasing subsequence is [2, 3, 7, 101], therefore the length is 4. Note that there may be more than one LIS combination, it is only necessary for you to return the length.

 

Solution:

#1. naive method; time complexity o(n^2). two layers iteration. 1) first iterate each element nums[i] in the nums, , 2) second iteration to find the bigger num[j] to add as each list 3) then find the maximum length list in the lists. length[i] += 1 if nums[j] > nums[i]
#or reversely, smaller nums[j] to update length[i]. i to 0 to len(nums), j = 0 to i; so length[i] = max(length[i], length[j]+1)

#2n d use binary search, try to select and insert into the increasing sequence
#(1) maintain a result list ans = [nums[0]]
#(2) iterate nums from second element num, compare num with the last element of ans:
# a. if num < ans[-1]
# insert num into ans
# else binary search in the ans the left insertion position for num (i.e. the smallest number that is bigger than num), and replace it

   def binarySearch(lst, ele):
            if len(lst) == 1:
                return 0
            l = 0
            h = len(lst) - 1
            while (l <= h):
                mid = (l+h)/2
                if lst[mid] == ele:
                    return mid
                elif lst[mid] < ele:
                    l = mid + 1
                else:
                    h = mid - 1
            if l >= len(lst):
                return -1
            return l
        
        if len(nums) == 0:
            return 0
        ansLst = []
        ansLst.append(nums[0])
        for i in range(1, len(nums)):
            if nums[i] > ansLst[-1]:
                ansLst.append(nums[i])
            else:
                #binary search
                pos = binarySearch(ansLst, nums[i])
                #print ('pos: ', len(ansLst), pos)
                ansLst[pos]
        return len(ansLst)

　　

 

#note it is for length of longest increasing sequence, the final ansLst may not be the real longest increasing sequence



#3rd use Dynamic programming
#use DP[i] indicate the length of longest increasing sequence at position i so far.
#it has optimal substructure: every sublist has the optimal solution for the longest increasing sequence
#overlapping subproblem: the large sublist problem is affected by the previous smaller sublist :
#the transition equation: DP[i] = max(DP[i], DP[j] + 1) ; i = 1 to len(nums), j = 0 to i
#intialize all DP element as 1

if len(nums) == 0:
            return 0
        
        dp = [1] * len(nums)
        for i in range(1, len(nums)):
            for j in range(0, i):
                if nums[j] < nums[i]:
                    dp[i] = max(dp[i], dp[j] + 1)
        return max(dp)

转载于:https://www.cnblogs.com/anxin6699/p/6962154.html