Re: How to use template to improve efficiency?
shuisheng wrote:
I have the following example and try to improve its efficiency.
#include<iostream>
using namespace std;
double a[] = {1.1, 2, 3, 4, 5, 6, 7, 8, 9.2};
template<double p[], int n>
inline
double Sum()
{
double result = 0;
for (int i = 0; i < n; ++i)
{
result += a[i];
}
return result;
}
int main()
{
cout << Sum<a, 9>();
return 0;
}
I hope the sum result can be calculated at compilation. But it isn't.
I look into the diassembly code, I see several fadd operations (the
loop is unrolled by the compiler). Any way can make the final sum
result is calculated at the compilation.
A way to make it all be done at compilation time is to create a
recursive template. You know, you can basically do the addition in a
recursive function:
int sum(int array[], int n)
{
if (n > 1)
return array[n-1] + sum(array, n-1);
else
return array[0];
}
Now, put it in the form of a template, by means of defining a
specialization (no 'if's allowed). Remember that
template<int K> int blah();
can be specialized for any particular K, like so
template<> int blah<42>() { ... }
The idiomatic form of calculating the factorial is
template<int N> int factorial() {
return N*factorial<N-1>();
}
// along with the specialization
template<> int factorial<0> {
return 1;
}
Enough to go for you homework? Call us back if you get stuck again.
V
--
Please remove capital 'A's when replying by e-mail
I do not respond to top-posted replies, please don't ask