Solving the dynamic programming problems

About DP

  • Dynamic programming is a very important algorithm, and mastery of it should be the basic skills for computer science students.

  • Let’s explain as much as possible about the understanding of this type of algorithm.

  • The key is state definition and state transition equation


Longest Increasing Subsequence

Given a sequence of K integers { N1, N2, …, NK }, any contiguous subsequence can be expressed as { Ni, Ni+1, …, Nj }, where 1 <= i <= j < = K. The largest contiguous subsequence is the element and the largest one of all consecutive subsequences, such as the given sequence { -2, 11, -4, 13, -5, -2 }, whose largest contiguous subsequence is { 11, -4, 13 } and the maximum sum is 20.

  • Define sum[i] as the maximum value of contiguous subsequence with A[i] as the last end

  • The state transition equation is: sum[i]=max(sum[i-1]+a[i], a[i])

  • T he actual solution is, as long as sum>0, then adding a[i] is still possible to increase max; but if sum<0, it should be discarded immediately, starting from 0 to calculate the next

“starthighlight” a = {4, 8, -12, 3, 7, 9} n = len(a) sum = 0 max = 0 for i in range(n): sum += a[i] if sum > max: max = sum if sum < 0: sum = a[i] return sum “endhighlight”

{% highlight Java %}

public int LIS(int[] arr) { int i, j, max = 0; int n = arr.length; int[] list = new int[n]; // length Arrays.fill(list, 1); int[] index = new int[n]; // distance Arrays.fill(index, -1);

    for (i = 1; i < n; i++)
        for (j = 0; j < i; j++) {
            if (arr[i] > arr[j] && list[i] < list[j] + 1) {
                list[i] = list[j] + 1;
                index[i] = j;
            }
        }

    // choose the max
    int max_index = 0;

    for (i = 0; i < n; i++)
        if (list[i] > max) {
            max = list[i];
            max_index = i;
        }


    StringBuilder builder = new StringBuilder();
    builder.insert(0, arr[max_index]);
    int next_index = index[max_index];
    while (next_index != -1) {
        builder.insert(0, arr[next_index] + " ");
        next_index = index[next_index];
    }
    System.out.println(builder.toString()); // print the result
    return max;

“endhighlight”


Number tower problem

img

From the top layer to the bottom layer, each layer can only go to the adjacent node, and calculate the sum of the numbers

  • Define the state equation: max[i,j] represents the sum of the largest numbers passed by [i,j] as the starting point. Then max[1,1] is our target

  • Define the state transition equation: max[i,j]=num[i,j]+max(max(i+1,j),max(i+1,j+1))

  • Then you can choose to solve it from top to bottom or from bottom to top

{% highlight C %} #include
#include
using namespace std;
#define maxnum 1000
int num[maxnum][maxnum];
int d[maxnum][maxnum];
int main(void)
{
int i,j,k;
int n,m;
cin»n;
while(n–){
cin»m;
for(i=1;i<=m;i++)
for(j=1;j<=i;j++)
cin»num[i][j];
for(j=1;j<=m;j++) d[m][j]=num[m][j];

    for(i=m-1;i>=1;i--)  
        for(j=1;j<=i;j++)  
        d[i][j]=num[i][j]+max(d[i+1][j+1],d[i+1][j]);  
    cout<<d[1][1]<<endl;  
    memset(d,0,sizeof(d));  
}  

}
“endhighlight”


Backpack Problem

  • There are different versions of backpack problem.Let me illustrate the basic one.

There are N items and a backpack with a capacity of V. The cost of the i-th item is c[i] and the value is w[i]. Decide which items are loaded into the backpack so the sum of the values can be max.

  • Solve it with Java.There are two types of answer,see the comment.

{% highlight Java %} public int Knapsack1(int[] value, int[] weight, int capacity, int number) { if (capacity <= 0 || number == 0) return 0; if (weight[number - 1] > capacity) return Knapsack1(value, weight, capacity, number - 1); else return max(value[number - 1] + Knapsack1(value, weight, capacity - weight[number - 1], number - 1), Knapsack1(value, weight, capacity, number - 1));

}

public int Knapsack2(int[] value, int[] weight, int capacity, int index) {
    if (capacity <= 0 || index >= value.length)
        return 0;
    if (weight[index] > capacity)
        return Knapsack2(value, weight, capacity, index + 1);
    else
        return max(value[index] + Knapsack2(value, weight, capacity - weight[index], index + 1),
                Knapsack2(value, weight, capacity, index + 1));
}

    // two ways to solve it
    int[] value = {60, 100, 120};
    int[] weight = {10, 20, 30};
    int capacity = 50;
    int number = value.length;
    System.out.println(dp.Knapsack1(value, weight, capacity, number)); // final result is 220
    int index = 0;
    System.out.println(dp.Knapsack2(value, weight, capacity, index));  // final result is also 220

“endhighlight”


303 Range Sum Quwey - Immutable

  • Link

  • About the meaning:Give an array that returns the value of any two range of ranges

  • Idea: The rpblem content says many calls to function, so the traversal solution will definitely be TLE(Time Limit Exceeded); and sum[i,j]=sum[j]-sum[i-1] is obvious, so all the sums are found in one pass. After the value is done, subtraction is fine. Note you may need to use global variables

  • code:

{% highlight C %} // written in c++ // Using vector to add status will be better class NumArray { int dp[100000]; public: NumArray(vector &nums) { if (nums.empty()) return ; int length=nums.size(); memset(dp,0,sizeof(dp)); dp[0]=nums[0]; for(int i=1;i<length;i++) dp[i]=dp[i-1]+nums[i]; }

int sumRange(int i, int j) {
    if (i==0)
        return dp[j];
    else
        return dp[j]-dp[i-1];    
}

}; “endhighlight”

Writen in Python:

“starthighlight” class NumArray(object): def init(self, nums):

    self.dp = nums
    for i in range(1,len(nums)):
        self.dp[i] += self.dp[i-1]

def sumRange(self, i, j):
    return self.dp[j] - (self.dp[i-1] if i > 0 else 0)

“endhighlight”


70 climbStatirs
  • link

  • About the meaning:Ascend a staircase, you can take 1 step or take two steps.if you go to n steps, how many solutions will you have?

  • Idea: Use dp[n] to indicate the number of methods to go to n steps. For dp[n-1], you can only choose to take 1 step; for dp[n-2], If you choose 1+1, it will overlap with dp[n-1], you can only choose 2 steps.

  • Code:

{% highlight C %} Written in C++ class Solution { public: int climbStairs(int n) { int dp[n]={0}; dp[0]=0; dp[1]=1; dp[2]=2; for(int i=3;i<=n;i++) dp[i]=(dp[i-1]+dp[i-2]); return dp[n]; } }; “endhighlight”

“starthighlight”

written in Python

class Solution(object): def init(self): self.dp={}

def climbStairs(self, n):
    self.dp[1]=1
    self.dp[2]=2
    for i in range(3,n+1):
        self.dp[i]=self.dp[i-1]+self.dp[i-2]
    return self.dp[n]  

“endhighlight”


64 Minimum Path Sum
  • link

  • About the meaning: m*n is a matrix of all positive numbers, which can be moved downwards to the right,find the sum of the distances from the upper left to the lower right.

  • Idea: Dynamic programming.using dp[i][j] to represent the shortest number of steps taken with i, j as the last square. Then dp[i][j]=max(dp[i-1][j]+grid[i][j],dp[i][j-1]+gird[i][j],we just need to implement the equation.

  • Code:

{% highlight C %} // written in C++ class Solution { public: int minPathSum(vector<vector>& grid) { if (grid.empty()) return 0; int m=grid.size(); int n=grid[0].size(); int dp[m][n]={0}; int i,j,k; dp[0][0]=grid[0][0]; for(i=1;i<m;i++) dp[i][0]=(dp[i-1][0]+grid[i][0]); for(i=1;i<n;i++) dp[0][i]=(dp[0][i-1]+grid[0][i]); for(i=1;i<m;i++) for(j=1;j<n;j++) dp[i][j]=min(dp[i-1][j]+grid[i][j],dp[i][j-1]+grid[i][j]); return dp[m-1][n-1]; } }; “endhighlight”