2021-11-27:给定一个数组arr,长度为N,做出一个结构,可以高效的做如下的查询:
int querySum(L,R) : 查询arr[L…R]上的累加和;int queryAim(L,R) : 查询arr[L…R]上的目标值,目标值定义如下:假设arr[L…R]上的值为[a,b,c,d],a+b+c+d = s,
目标值为 : (s-a)^2 + (s-b)^2 + (s-c)^2 + (s-d)^2;int queryMax(L,R) : 查询arr[L…R]上的最大值.
要求:初始化该结构的时间复杂度不能超过O(N*logN);三个查询的时间复杂度不能超过O(logN);查询时,认为arr的下标从1开始,比如 :
arr = [ 1, 1, 2, 3 ];
querySum(1, 3) -> 4;
queryAim(2, 4) -> 50;
queryMax(1, 4) -> 3。
来自美团。
答案2021-11-27:
querySum方法,前缀和。
queryAim方法,前缀和,平方数组的前缀和,线段树。对目标值展开,(N-2)*S平方+a1的平方+a2的平方+…+an的平方。
queryMax方法,线段树。
代码用golang编写。代码如下:
package main
import (
"fmt"
"math"
)
func main() {
q := NewQuery([]int{1, 2, 3, 4, 5})
ret := 0
ret = q.querySum(1, 3)
fmt.Println(ret)
ret = q.queryAim(1, 3)
fmt.Println(ret)
ret = q.queryMax(1, 3)
fmt.Println(ret)
}
type SegmentTree struct {
max []int
change []int
update []bool
}
func NewSegmentTree(N int) *SegmentTree {
ret := &SegmentTree{}
ret.max = make([]int, N<<2)
ret.change = make([]int, N<<2)
ret.update = make([]bool, N<<2)
for i := 0; i < len(ret.max); i++ {
ret.max[i] = math.MinInt64
return ret
}
return ret
}
func (this *SegmentTree) pushUp(rt int) {
this.max[rt] = getMax(this.max[rt<<1], this.max[rt<<1|1])
}
func getMax(a int, b int) int {
if a > b {
return a
} else {
return b
}
}
// ln表示左子树元素结点个数,rn表示右子树结点个数
func (this *SegmentTree) pushDown(rt, ln, rn int) {
if this.update[rt] {
this.update[rt<<1] = true
this.update[rt<<1|1] = true
this.change[rt<<1] = this.change[rt]
this.change[rt<<1|1] = this.change[rt]
this.max[rt<<1] = this.change[rt]
this.max[rt<<1|1] = this.change[rt]
this.update[rt] = false
}
}
func (this *SegmentTree) update0(L, R, C, l, r, rt int) {
if L <= l && r <= R {
this.update[rt] = true
this.change[rt] = C
this.max[rt] = C
return
}
mid := (l + r) >> 1
this.pushDown(rt, mid-l+1, r-mid)
if L <= mid {
this.update0(L, R, C, l, mid, rt<<1)
}
if R > mid {
this.update0(L, R, C, mid+1, r, rt<<1|1)
}
this.pushUp(rt)
}
func (this *SegmentTree) query(L, R, l, r, rt int) int {
if L <= l && r <= R {
return this.max[rt]
}
mid := (l + r) >> 1
this.pushDown(rt, mid-l+1, r-mid)
left := 0
right := 0
if L <= mid {
left = this.query(L, R, l, mid, rt<<1)
}
if R > mid {
right = this.query(L, R, mid+1, r, rt<<1|1)
}
return getMax(left, right)
}
type Query struct {
sum1 []int
sum2 []int
st *SegmentTree
m int
}
func NewQuery(arr []int) *Query {
ret := &Query{}
n := len(arr)
ret.m = len(arr) + 1
ret.sum1 = make([]int, ret.m)
ret.sum2 = make([]int, ret.m)
ret.st = NewSegmentTree(ret.m)
for i := 0; i < n; i++ {
ret.sum1[i+1] = ret.sum1[i] + arr[i]
ret.sum2[i+1] = ret.sum2[i] + arr[i]*arr[i]
ret.st.update0(i+1, i+1, arr[i], 1, ret.m, 1)
}
return ret
}
func (this *Query) querySum(L, R int) int {
return this.sum1[R] - this.sum1[L-1]
}
func (this *Query) queryAim(L, R int) int {
sumPower2 := this.querySum(L, R)
sumPower2 *= sumPower2
return this.sum2[R] - this.sum2[L-1] + (R-L-1)*sumPower2
}
func (this *Query) queryMax(L, R int) int {
return this.st.query(L, R, 1, this.m, 1)
}
执行结果如下:
左神java代码
欢迎分享,转载请注明来源:内存溢出
评论列表(0条)