c++ - ctime 有问题，计算函数运行时间

Question

我无法计算出我的两个 maxsubarray 函数的运行时间。 (就在代码的底部)它给我的输出:

 Inputsize: 101 Time using Brute Force:0 Time Using DivandCon: 12

我第二次使用 clock() 时是正确的，但对于第一个区别 diff1 它只给了我 0 我不确定为什么？

编辑:修订代码。

Edit2:添加输出。

#include <iostream>
#include <cmath>
#include <cstdlib>
#include <ctime>
#include <limits.h>

using namespace std;

int Kedane(int a[], int size)
{
int max_so_far = 0, max_ending_here = 0;
int i;
for(i = 0; i < size; i++)
{
    max_ending_here = max_ending_here + a[i];
    if(max_ending_here < 0)
        max_ending_here = 0;
    if(max_so_far < max_ending_here)
        max_so_far = max_ending_here;
}
return max_so_far;
}

int BruteForce(int array[],int n)
{
int sum,ret=0;

for(int j=-1;j<=n-2;j++)
{
    sum=0;
    for(int k=j+1;k<+n-1;k++)
    {
        sum+=array[k];

        if(sum>ret)
        {
            ret=sum;
        }
    }
}
return ret;
}
//------------------------------------------------------
// FUNCTION WHICH FINDS MAX OF 2 INTS
int max(int a, int b) { return (a > b)? a : b; }

// FUNCTION WHICH FINDS MAX OF 3 NUMBERS
// CALL MAX FUNCT FOR 2 VARIS TWICE!
int max(int a, int b, int c) { return max(max(a, b), c); }

// WORKS OUT FROM MIDDLE+1->RIGHT THE MAX SUM &
// THE MAX SUM FROM MIDDLE->LEFT + RETURNS SUM OF THESE
int maxCrossingSum(int arr[], int l, int m, int h)
{
int sum = 0;            // LEFT OF MID
int LEFTsum = INT_MIN;  // INITIALLISES SUM TO LOWEST POSSIBLE INT
for (int i = m; i >= l; i--)
{
    sum = sum + arr[i];
    if (sum > LEFTsum)
        LEFTsum = sum;
}

sum = 0;                // RIGHT OF MID
int RIGHTsum = INT_MIN;
for (int i = m+1; i <= h; i++)
{
    sum = sum + arr[i];
    if (sum > RIGHTsum)
        RIGHTsum = sum;
}

// RETURN SUM OF BOTH LEFT AND RIGHT SIDE MAX'S
return LEFTsum + RIGHTsum;
}
// Returns sum of maxium sum subarray in aa[l..h]
int maxSubArraySum(int arr[], int l, int h)
{
// Base Case: Only one element
if (l == h)
    return arr[l];

// Find middle point
int m = (l + h)/2;

/* Return maximum of following three possible cases
 a) Maximum subarray sum in left half
 b) Maximum subarray sum in right half
 c) Maximum subarray sum such that the subarray crosses the midpoint */
return max(maxSubArraySum(arr, l, m),
           maxSubArraySum(arr, m+1, h),
           maxCrossingSum(arr, l, m, h));
}

// DRIVER
int main(void)
{
std::srand (time(NULL));

// CODE TO FILL ARRAY WITH RANDOMS [-50;50]
int size=30000;
int array[size];

for(int i=0;i<=size;i++)
{
    array[i]=(std::rand() % 100) -50;
}

// TIMING VARI'S
clock_t t1,t2;
clock_t A,B;
clock_t K1,K2;
volatile int  mb, md, qq;

//VARYING ELEMENTS IN THE ARRAY
for(int  n=101;n<size;n=n+100)
{
    t1=clock();
    mb=BruteForce(array,n);
    t2=clock();

    A=clock();
    md=maxSubArraySum(array, 0, n-1) ;
    B=clock();

    K1=clock();
    qq=Kedane(array, n);
    K2=clock();


 cout<< n << "," << (double)t2-(double)t1 << ","<<(double)B-(double)A << ","<<(double)K2-(double)K1<<endl;

}

return 0;

101,0,0,0
201,0,0,0
301,1,0,0
401,0,0,0
501,0,0,0
601,0,0,0
701,0,0,0
801,1,0,0
901,1,0,0
1001,0,0,0
1101,1,0,0
1201,1,0,0
1301,0,0,0
1401,1,0,0
1501,1,0,0
1601,2,0,0
1701,1,0,0
1801,2,0,0
1901,1,1,0
2001,1,0,0
2101,2,0,0
2201,3,0,0
2301,2,0,0
2401,3,0,0
2501,3,0,0
2601,3,0,0
2701,4,0,0
2801,4,0,0
2901,4,0,0
3001,4,0,0
3101,4,0,0
3201,5,0,0
3301,5,0,0
3401,6,0,0
3501,5,0,0
3601,6,0,0
3701,6,0,0
3801,8,0,0
3901,7,0,0
4001,8,0,0
4101,7,0,0
4201,10,1,0
4301,9,0,0
4401,8,0,0
4501,9,0,0
4601,10,0,0
4701,11,0,0
4801,11,0,0
4901,11,0,0
5001,12,0,1
5101,11,1,0
5201,13,0,0
5301,13,0,0
5401,15,0,0
5501,14,0,0
5601,16,0,0
5701,15,0,0
5801,15,1,0
5901,16,0,0
6001,17,0,0
6101,18,0,0
6201,18,0,0
6301,19,0,0
6401,21,0,0
6501,19,0,0
6601,21,1,0
6701,20,0,0
6801,22,0,0
6901,23,0,0
7001,22,0,0
7101,24,0,0
7201,26,0,0
7301,26,0,0
7401,24,1,0
7501,26,0,0
7601,27,0,0
7701,28,0,0
7801,28,0,0
7901,30,0,0
8001,29,0,0
8101,31,0,0
8201,31,1,0
8301,35,0,0
8401,33,0,0
8501,35,0,0
8601,35,1,0
8701,35,0,0
8801,36,1,0
8901,37,0,0
9001,38,0,0
9101,39,0,0
9201,41,1,0
9301,40,0,0
9401,41,0,0
9501,42,0,0
9601,45,0,0
9701,45,0,0
9801,44,0,0
9901,47,0,0
10001,47,0,0
10101,48,0,0
10201,50,0,0
10301,51,0,0
10401,50,0,0
10501,51,0,0
10601,53,0,0
10701,55,0,0
10801,54,0,0
10901,56,0,0
11001,57,0,0
11101,56,0,0
11201,60,0,0
11301,60,0,0
11401,61,1,0
11501,61,1,0
11601,63,0,0
11701,62,1,0
11801,66,1,0
11901,65,0,0
12001,68,1,0
12101,68,0,0
12201,70,0,0
12301,71,0,0
12401,72,0,0
12501,73,1,0
12601,73,1,0
12701,76,0,0
12801,77,0,0
12901,78,1,0
13001,79,1,0
13101,80,0,0
13201,83,0,0
13301,82,0,0
13401,86,0,0
13501,85,1,0
13601,86,0,0
13701,89,0,0
13801,90,0,1
13901,90,0,0
14001,91,0,0
14101,97,0,0
14201,93,0,0
14301,96,0,0
14401,99,0,0
14501,100,0,0
14601,101,0,0
14701,101,0,0
14801,103,1,0
14901,104,0,0
15001,107,0,0
15101,108,0,0
15201,109,0,0
15301,109,0,0
15401,114,0,0
15501,114,0,0
15601,115,0,0
15701,116,0,0
15801,119,0,0
15901,118,0,0
16001,124,0,0
16101,123,1,0
16201,123,1,0
16301,125,0,0
16401,127,1,0
16501,128,1,0
16601,131,0,0
16701,132,0,0
16801,134,0,0
16901,134,1,0
17001,135,1,0
17101,139,0,0
17201,139,0,0
17301,140,1,0
17401,143,0,0
17501,145,0,0
17601,147,0,0
17701,147,0,0
17801,150,1,0
17901,152,1,0
18001,153,0,0
18101,155,0,0
18201,157,0,0
18301,157,1,0
18401,160,0,0
18501,160,1,0
18601,163,1,0
18701,165,0,0
18801,169,0,0
18901,171,0,1
19001,170,1,0
19101,173,1,0
19201,178,0,0
19301,175,1,0
19401,176,1,0
19501,180,0,0
19601,180,1,0
19701,182,1,0
19801,184,0,0
19901,187,1,0
20001,188,1,0
20101,191,0,0
20201,192,1,0
20301,193,1,0
20401,195,0,0
20501,199,0,0
20601,200,0,0
20701,201,0,0
20801,209,1,0
20901,210,0,0
21001,206,0,0
21101,210,0,0
21201,210,0,0
21301,213,0,0
21401,215,1,0
21501,217,1,0
21601,218,1,0
21701,221,1,0
21801,222,1,0
21901,226,1,0
22001,225,1,0
22101,229,0,0
22201,232,0,0
22301,233,1,0
22401,234,1,0
22501,237,1,0
22601,238,0,1
22701,243,0,0
22801,242,1,0
22901,246,1,0
23001,246,0,0
23101,250,1,0
23201,250,1,0
23301,254,1,0
23401,254,0,0
23501,259,0,1
23601,260,1,0
23701,263,1,0
23801,268,0,0
23901,266,1,0
24001,271,0,0
24101,272,1,0
24201,274,1,0
24301,280,0,1
24401,279,0,0
24501,281,0,0
24601,285,0,0
24701,288,0,0
24801,289,0,0
24901,293,0,0
25001,295,1,0
25101,299,1,0
25201,299,1,0
25301,302,0,0
25401,305,1,0
25501,307,0,0
25601,310,1,0
25701,315,0,0
25801,312,1,0
25901,315,0,0
26001,320,1,0
26101,320,0,0
26201,322,0,0
26301,327,1,0
26401,329,0,0
26501,332,1,0
26601,339,1,0
26701,334,1,0
26801,337,0,0
26901,340,0,0
27001,341,1,0
27101,342,1,0
27201,347,0,0
27301,348,1,0
27401,351,1,0
27501,353,0,0
27601,356,1,0
27701,360,0,1
27801,361,1,0
27901,362,1,0
28001,366,1,0
28101,370,0,1
28201,372,0,0
28301,375,1,0
28401,377,1,0
28501,380,0,0
28601,384,1,0
28701,384,0,0
28801,388,1,0
28901,391,1,0
29001,392,1,0
29101,399,1,0
29201,399,0,0
29301,404,1,0
29401,405,0,0
29501,409,1,0
29601,412,2,0
29701,412,1,0
29801,422,1,0
29901,419,1,0

Answer 1

从未使用 BruteForce 和 maxSubArraySum 的返回值，这为编译器提供了很大的优化空间。

例如，在我的机器上，使用 clang -O3 将对 BruteForce 的调用减少为 vector 拷贝，仅此而已。

强制计算这些函数的一种方法是将它们的结果写入可变变量:

volatile int mb, md;
// ...
    mb = BruteForce(array, n);
    // ...
    md = maxSubArraySum(array, 0, n-1);

由于变量是可变的，因此必须存储赋值右侧给出的值，尽管没有任何其他副作用，这会阻止编译器优化计算。