📜  最长递增连续子序列

📅  最后修改于: 2021-10-27 07:16:58             🧑  作者: Mango

给定 N 个元素,编写一个程序,打印其相邻元素差为 1 的最长递增子序列的长度。
例子:

朴素的方法:正常的方法是迭代每个元素并找出最长的递增子序列。对于任何特定元素,找出从该元素开始的子序列的长度。打印由此形成的子序列的最长长度。这种方法的时间复杂度为 O(n 2 )。
动态规划方法:让 DP[i] 存储以 A[i] 结尾的最长子序列的长度。对于每个 A[i],如果 A[i]-1 存在于第 i 个索引之前的数组中,则 A[i] 将添加到具有 A[i]-1 的递增子序列。因此, DP[i] = DP[ index(A[i]-1) ] + 1 。如果数组中第 i 个索引之前不存在 A[i]-1,则 DP[i]=1,因为 A[i] 元素形成以 A[i] 开头的子序列。因此,DP[i] 的关系是:

下面给出了上述方法的说明:

C++
// CPP program to find length of the
// longest increasing subsequence
// whose adjacent element differ by 1
#include 
using namespace std;
 
// function that returns the length of the
// longest increasing subsequence
// whose adjacent element differ by 1
int longestSubsequence(int a[], int n)
{
    // stores the index of elements
    unordered_map mp;
 
    // stores the length of the longest
    // subsequence that ends with a[i]
    int dp[n];
    memset(dp, 0, sizeof(dp));
 
    int maximum = INT_MIN;
 
    // iterate for all element
    for (int i = 0; i < n; i++) {
 
        // if a[i]-1 is present before i-th index
        if (mp.find(a[i] - 1) != mp.end()) {
 
            // last index of a[i]-1
            int lastIndex = mp[a[i] - 1] - 1;
 
            // relation
            dp[i] = 1 + dp[lastIndex];
        }
        else
            dp[i] = 1;
 
        // stores the index as 1-index as we need to
        // check for occurrence, hence 0-th index
        // will not be possible to check
        mp[a[i]] = i + 1;
 
        // stores the longest length
        maximum = max(maximum, dp[i]);
    }
 
    return maximum;
}
 
// Driver Code
int main()
{
    int a[] = { 3, 10, 3, 11, 4, 5, 6, 7, 8, 12 };
    int n = sizeof(a) / sizeof(a[0]);
    cout << longestSubsequence(a, n);
    return 0;
}


Java
// Java program to find length of the
// longest increasing subsequence
// whose adjacent element differ by 1
 
import java.util.*;
class lics {
    static int LongIncrConseqSubseq(int arr[], int n)
    {
        // create hashmap to save latest consequent
        // number as "key" and its length as "value"
        HashMap map = new HashMap<>();
        
        // put first element as "key" and its length as "value"
        map.put(arr[0], 1);
        for (int i = 1; i < n; i++) {
        
            // check if last consequent of arr[i] exist or not
            if (map.containsKey(arr[i] - 1)) {
        
                // put the updated consequent number
                // and increment its value(length)
                map.put(arr[i], map.get(arr[i] - 1) + 1);
           
                // remove the last consequent number
                map.remove(arr[i] - 1);
            }
 
            // if their is no last consequent of
            // arr[i] then put arr[i]
            else {
                map.put(arr[i], 1);
            }
        }
        return Collections.max(map.values());
    }
 
    // driver code
    public static void main(String args[])
    {
        // Take input from user
        Scanner sc = new Scanner(System.in);
        int n = sc.nextInt();
        int arr[] = new int[n];
        for (int i = 0; i < n; i++)
            arr[i] = sc.nextInt();
        System.out.println(LongIncrConseqSubseq(arr, n));
    }
}
// This code is contributed by CrappyDoctor


Python3
# python program to find length of the
# longest increasing subsequence
# whose adjacent element differ by 1
 
from collections import defaultdict
import sys
 
# function that returns the length of the
# longest increasing subsequence
# whose adjacent element differ by 1
 
def longestSubsequence(a, n):
    mp = defaultdict(lambda:0)
 
    # stores the length of the longest
    # subsequence that ends with a[i]
    dp = [0 for i in range(n)]
    maximum = -sys.maxsize
 
    # iterate for all element
    for i in range(n):
 
        # if a[i]-1 is present before i-th index
        if a[i] - 1 in mp:
 
            # last index of a[i]-1
            lastIndex = mp[a[i] - 1] - 1
 
            # relation
            dp[i] = 1 + dp[lastIndex]
        else:
            dp[i] = 1
 
            # stores the index as 1-index as we need to
            # check for occurrence, hence 0-th index
            # will not be possible to check
        mp[a[i]] = i + 1
 
        # stores the longest length
        maximum = max(maximum, dp[i])
    return maximum
 
 
# Driver Code
a = [3, 10, 3, 11, 4, 5, 6, 7, 8, 12]
n = len(a)
print(longestSubsequence(a, n))
 
# This code is contributed by Shrikant13


C#
// C# program to find length of the
// longest increasing subsequence
// whose adjacent element differ by 1
using System;
using System.Collections.Generic;
class GFG{
     
static int longIncrConseqSubseq(int []arr,
                                int n)
{
  // Create hashmap to save
  // latest consequent number
  // as "key" and its length
  // as "value"
  Dictionary map = new Dictionary();
 
  // Put first element as "key"
  // and its length as "value"
  map.Add(arr[0], 1);
  for (int i = 1; i < n; i++)
  {
    // Check if last consequent
    // of arr[i] exist or not
    if (map.ContainsKey(arr[i] - 1))
    {
      // put the updated consequent number
      // and increment its value(length)
      map.Add(arr[i], map[arr[i] - 1] + 1);
 
      // Remove the last consequent number
      map.Remove(arr[i] - 1);
    }
 
    // If their is no last consequent of
    // arr[i] then put arr[i]
    else
    {
      if(!map.ContainsKey(arr[i]))
        map.Add(arr[i], 1);
    }
  }
   
  int max = int.MinValue;
  foreach(KeyValuePair entry in map)
  {
    if(entry.Value > max)
    {
      max = entry.Value;
    }
  }
  return max;
}
 
// Driver code
public static void Main(String []args)
{
  // Take input from user
  int []arr = {3, 10, 3, 11,
               4, 5, 6, 7, 8, 12};
  int n = arr.Length;
  Console.WriteLine(longIncrConseqSubseq(arr, n));
}
}
 
// This code is contributed by gauravrajput1


Javascript


输出:
6

时间复杂度: O(n)
辅助空间: O(n)

如果您希望与专家一起参加现场课程,请参阅DSA 现场工作专业课程学生竞争性编程现场课程