下面是O(n^2)算法的Scala实现:

object Solve {
  def longestIncrSubseq[T](xs: List[T])(implicit ord: Ordering[T]) = {
    xs.foldLeft(List[(Int, List[T])]()) {
      (sofar, x) =>
        if (sofar.isEmpty) List((1, List(x)))
        else {
          val resIfEndsAtCurr = (sofar, xs).zipped map {
            (tp, y) =>
              val len = tp._1
              val seq = tp._2
              if (ord.lteq(y, x)) {
                (len + 1, x :: seq) // reversely recorded to avoid O(n)
              } else {
                (1, List(x))
              }
          }
          sofar :+ resIfEndsAtCurr.maxBy(_._1)
        }
    }.maxBy(_._1)._2.reverse
  }

  def main(args: Array[String]) = {
    println(longestIncrSubseq(List(
      0, 8, 4, 12, 2, 10, 6, 14, 1, 9, 5, 13, 3, 11, 7, 15)))
  }
}

下面是从动态规划的角度评估问题的三个步骤:

递归定义:maxLength(i) == 1 + maxLength(j) where 0 < j < i and array[i] > array[j] 递归参数边界:可能有0到i - 1个子序列作为参数传递 求值顺序:由于是递增子序列，所以要从0求值到n

如果我们以序列{0,8,2,3,7,9}为例，at index:

我们会得到子序列{0}作为基本情况 [1]有一个新的子序列{0,8} [2]试图评估两个新的序列{0,8,2}和{0,2}通过添加元素在索引2到现有的子序列-只有一个是有效的，所以添加第三个可能的序列{0,2}只到参数列表 .．.

下面是c++ 11的工作代码:

#include <iostream>
#include <vector>

int getLongestIncSub(const std::vector<int> &sequence, size_t index, std::vector<std::vector<int>> &sub) {
    if(index == 0) {
        sub.push_back(std::vector<int>{sequence[0]});
        return 1;
    }

    size_t longestSubSeq = getLongestIncSub(sequence, index - 1, sub);
    std::vector<std::vector<int>> tmpSubSeq;
    for(std::vector<int> &subSeq : sub) {
        if(subSeq[subSeq.size() - 1] < sequence[index]) {
            std::vector<int> newSeq(subSeq);
            newSeq.push_back(sequence[index]);
            longestSubSeq = std::max(longestSubSeq, newSeq.size());
            tmpSubSeq.push_back(newSeq);
        }
    }
    std::copy(tmpSubSeq.begin(), tmpSubSeq.end(),
              std::back_insert_iterator<std::vector<std::vector<int>>>(sub));

    return longestSubSeq;
}

int getLongestIncSub(const std::vector<int> &sequence) {
    std::vector<std::vector<int>> sub;
    return getLongestIncSub(sequence, sequence.size() - 1, sub);
}

int main()
{
    std::vector<int> seq{0, 8, 2, 3, 7, 9};
    std::cout << getLongestIncSub(seq);
    return 0;
}

c++中最简单的LIS解决方案，具有O(nlog(n))时间复杂度

#include <iostream>
#include "vector"
using namespace std;

// binary search (If value not found then it will return the index where the value should be inserted)
int ceilBinarySearch(vector<int> &a,int beg,int end,int value)
{
    if(beg<=end)
    {
        int mid = (beg+end)/2;
        if(a[mid] == value)
            return mid;
        else if(value < a[mid])
            return ceilBinarySearch(a,beg,mid-1,value);
        else
            return ceilBinarySearch(a,mid+1,end,value);

    return 0;
    }

    return beg;

}
int lis(vector<int> arr)
{
    vector<int> dp(arr.size(),0);
    int len = 0;
    for(int i = 0;i<arr.size();i++)
    {
        int j = ceilBinarySearch(dp,0,len-1,arr[i]);
        dp[j] = arr[i];
        if(j == len)
            len++;

    }
    return len;
}

int main()
{
    vector<int> arr  {2, 5,-1,0,6,1,2};
    cout<<lis(arr);
    return 0;
}

输出: 4

这里是java O(nlogn)的实现

import java.util.Scanner;

public class LongestIncreasingSeq {


    private static int binarySearch(int table[],int a,int len){

        int end = len-1;
        int beg = 0;
        int mid = 0;
        int result = -1;
        while(beg <= end){
            mid = (end + beg) / 2;
            if(table[mid] < a){
                beg=mid+1;
                result = mid;
            }else if(table[mid] == a){
                return len-1;
            }else{
                end = mid-1;
            }
        }
        return result;
    }
    
    public static void main(String[] args) {        
        
//        int[] t = {1, 2, 5,9,16};
//        System.out.println(binarySearch(t , 9, 5));
        Scanner in = new Scanner(System.in);
        int size = in.nextInt();//4;
        
        int A[] = new int[size];
        int table[] = new int[A.length]; 
        int k = 0;
        while(k<size){
            A[k++] = in.nextInt();
            if(k<size-1)
                in.nextLine();
        }        
        table[0] = A[0];
        int len = 1; 
        for (int i = 1; i < A.length; i++) {
            if(table[0] > A[i]){
                table[0] = A[i];
            }else if(table[len-1]<A[i]){
                table[len++]=A[i];
            }else{
                table[binarySearch(table, A[i],len)+1] = A[i];
            }            
        }
        System.out.println(len);
    }    
}

//可以使用TreeSet

说到DP solution，我发现很奇怪的是没有人提到LIS可以简化为LCS。你所需要做的就是对原始序列的副本进行排序，删除所有重复的副本，然后对它们进行LCS。在伪代码中是:

def LIS(S):
    T = sort(S)
    T = removeDuplicates(T)
    return LCS(S, T)

以及用Go语言编写的完整实现。如果你不需要重构解，你就不需要维护整个n^2 DP矩阵。

func lcs(arr1 []int) int {
    arr2 := make([]int, len(arr1))
    for i, v := range arr1 {
        arr2[i] = v
    }
    sort.Ints(arr1)
    arr3 := []int{}
    prev := arr1[0] - 1
    for _, v := range arr1 {
        if v != prev {
            prev = v
            arr3 = append(arr3, v)
        }
    }

    n1, n2 := len(arr1), len(arr3)

    M := make([][]int, n2 + 1)
    e := make([]int, (n1 + 1) * (n2 + 1))
    for i := range M {
        M[i] = e[i * (n1 + 1):(i + 1) * (n1 + 1)]
    }

    for i := 1; i <= n2; i++ {
        for j := 1; j <= n1; j++ {
            if arr2[j - 1] == arr3[i - 1] {
                M[i][j] = M[i - 1][j - 1] + 1
            } else if M[i - 1][j] > M[i][j - 1] {
                M[i][j] = M[i - 1][j]
            } else {
                M[i][j] = M[i][j - 1]
            }
        }
    }

    return M[n2][n1]
}

求最长递增子序列(LIS)的O(NLog(N))递归DP方法

解释

该算法涉及创建节点格式为(a,b)的树。

A表示到目前为止我们考虑添加到有效子序列的下一个元素。

B表示剩余子数组的起始索引，如果a被添加到目前为止我们所拥有的子数组的末尾，则下一个决策将从该子数组开始。

算法

We start with an invalid root (INT_MIN,0), pointing at index zero of the array since subsequence is empty at this point, i.e. b = 0. Base Case: return 1 if b >= array.length. Loop through all the elements in the array from the b index to the end of the array, i.e i = b ... array.length-1. i) If an element, array[i] is greater than the current a, it is qualified to be considered as one of the elements to be appended to the subsequence we have so far. ii) Recurse into the node (array[i],b+1), where a is the element we encountered in 2(i) which is qualified to be appended to the subsequence we have so far. And b+1 is the next index of the array to be considered. iii) Return the max length obtained by looping through i = b ... array.length. In a case where a is bigger than any other element from i = b to array.length, return 1. Compute the level of the tree built as level. Finally, level - 1 is the desired LIS. That is the number of edges in the longest path of the tree.

注意:算法的记忆部分被省略了，因为它是从树中清除的。

随便举个例子 标记为x的节点从DB内存值中获取。

Java实现

public int lengthOfLIS(int[] nums) {
            return LIS(nums,Integer.MIN_VALUE, 0,new HashMap<>()) -1;
    }
    public int LIS(int[] arr, int value, int nextIndex, Map<String,Integer> memo){
        if(memo.containsKey(value+","+nextIndex))return memo.get(value+","+nextIndex);
        if(nextIndex >= arr.length)return 1;

        int max = Integer.MIN_VALUE;
        for(int i=nextIndex; i<arr.length; i++){
            if(arr[i] > value){
                max = Math.max(max,LIS(arr,arr[i],i+1,memo));
            }
        }
        if(max == Integer.MIN_VALUE)return 1;
        max++;
        memo.put(value+","+nextIndex,max);
        return max;
    }