1、题目
Given an array A of integers, return the length of the longest arithmetic subsequence in A.
Recall that a subsequence of A is a list A[i_1], A[i_2], …, A[i_k] with 0 <= i_1 < i_2 < … < i_k <= A.length - 1, and that a sequence B is arithmetic if B[i+1] - B[i] are all the same value (for 0 <= i < B.length - 1).
Example 1:
Input: [3,6,9,12]
Output: 4
Explanation:
The whole array is an arithmetic sequence with steps of length = 3.
Example 2:
Input: [9,4,7,2,10]
Output: 3
Explanation:
The longest arithmetic subsequence is [4,7,10].
Example 3:
Input: [20,1,15,3,10,5,8]
Output: 4
Explanation:
The longest arithmetic subsequence is [20,15,10,5].
Note:
2 <= A.length <= 2000
0 <= A[i] <= 10000
2、思路
这是一道坐标型动态规划的题,让求最长算术子序列,所谓算术子序列就是序列中相邻元素之间的差值都相等,我们可以按照下面四步来分析:
2.1、确定状态
我们从最后一个元素入手,假设它是一个最长算术子序列中最后一个元素a[k]
,则有:
a[k]-a[k-1] = a[k-1]-a[k-2]
但是diff=a[k]-a[k-1]
差值是多少我们不知道,对于所给列表中的最后一位元素,我们需要知道它前面哪个元素和他的差值是最长算术子序列的差值diff
.
例如:[1,3,5,4,7]
对于7
,和前面四个元素的差值分别是 3、2、4、6
,哪个差值才是最长子序列的差值diff
?显然我们需要枚举某个元素a[k]
与前面k-1
个元素的差值,并找到以该差值为标准的最长算术子序列长。
这里我们需要枚举a[k]
与其前k-1
个元素之间的差值,题中给出0=<A[i]<=10000
,所以a[k]-a[k-1]
的范围就是[-10000, 10000]
,于是我们可以开一个n*20001
的数组来存放差值对应的最长算术子序列长度
状态:
设f[i][j]=
以元素a[i]
结尾且以差值j为标准的的最长算术子序列长度
2.2、转移方程
f[i][j] = 2
f[i][j] = max{f[i-1][j], f[i-2][j],...f[0][j]}
2.3、 初始条件和边界情况
无初始条件和边界情况
2.4、 计算顺序
从f[1][diff], f[2][diff], f[3][diff]...f[n-1][diff]
结果就是找到最大的f[i][diff]
时间复杂度O(n^2)
时间复杂度O(n*20001)
3、实现
public int longestArithSeqLength(int[] A) {
int n = A.length;
int[][] f = new int[n][20000+1];
int max = 2;
for (int i = 1; i < n; i++){
for (int j = 0; j < i; j++){
int diff = A[i]-A[j]+10000;
// 说明前面没有差值相同的元素
if (f[j][diff] == 0){
f[i][diff] = 2;
continue;
}
f[i][diff] = Math.max(f[i][diff], f[j][diff]+1);
max = Math.max(max, f[i][diff]);
}
}
return max;
}
4、优化
我们直接开了一个n*20001的二维数组,导致空间浪费,我们可以找到所给数组中的最大值和最小值,再决定开多大的数组。
public int longestArithSeqLength(int[] A) {
int n = A.length;
int max = Integer.MIN_VALUE;
int min = Integer.MAX_VALUE;
for (int a : A){
if (a > max) max = a;
if (a < min) min = a;
}
int maxDiff = max - min;
int[][] f = new int[n][maxDiff*2+1];
int ans = 2;
for (int i = 1; i < n; i++){
for (int j = 0; j < i; j++){
int diff = A[i] - A[j] + maxDiff;
if (f[j][diff] == 0){
f[i][diff] = 2;
}
f[i][diff] = Math.max(f[j][diff] + 1, f[i][diff]);
if (f[i][diff] > ans){
ans = f[i][diff];
}
}
}
return ans;
}