使用 Fortran 正确查找标准差和总体方差时出现问题

发布于 2024-10-30 10:13:39 字数 546 浏览 10 评论 0原文

当我在 fortran 95 上搜索标准差和总体方差时，我找不到太多东西。所以我想知道是否有人可以帮助我？感谢您抽出时间。

这是我所做的，它编译得很好，但结果给出了错误的值：

    SUBROUTINE ComputeStats(num, sum, avg, variance, sdv)

        REAL, INTENT(IN):: num(10)
        REAL, INTENT(OUT):: sum
        REAL, INTENT(OUT):: avg, variance, sdv

        DO i=1, 10
            sum=sum+num(i)
        END DO

        avg=sum/10

        DO i=1, 10
            variance=(((num(i)-avg)**2.0)/10)
            variance=variance+i
            sdv=variance**0.5
        END DO

    END SUBROUTINE

原文

I couldn't find much when I searched for standard deviations and population variances on fortran 95. So I'm wondering if someone could help me? Thanks for your time.

Here is what I did, it compiled fine but it turned out giving wrong values:

    SUBROUTINE ComputeStats(num, sum, avg, variance, sdv)

        REAL, INTENT(IN):: num(10)
        REAL, INTENT(OUT):: sum
        REAL, INTENT(OUT):: avg, variance, sdv

        DO i=1, 10
            sum=sum+num(i)
        END DO

        avg=sum/10

        DO i=1, 10
            variance=(((num(i)-avg)**2.0)/10)
            variance=variance+i
            sdv=variance**0.5
        END DO

    END SUBROUTINE

分享到QQ

分享到微博

如果你对这篇内容有疑问，欢迎到本站社区发帖提问参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。

发布评论

需要登录才能够评论，你可以免费注册一个本站的账号。

昨迟人 2024-11-06 10:13:39

您没有正确计算方差。

       variance = 0
       DO i=1, 10
         variance = variance + (((num(i)-avg)**2.0)/10)
       END DO

这为您提供了正确的方差，例如 1,1,1,1,1,1,1,1,1,1，即 0，而不是 10。

You don't calculate the variance correctly.

       variance = 0
       DO i=1, 10
         variance = variance + (((num(i)-avg)**2.0)/10)
       END DO

This gives you the right variance for e.g. 1,1,1,1,1,1,1,1,1,1 which is 0, not 10.

回复收藏 0 原文

溺孤伤于心 2024-11-06 10:13:39

用于查找数组标准差的简单子例程

subroutine find_std(n,arr,std_dev)
integer,intent(in)::n
real,intent(in),dimension(n)::arr
real,intent(out)::std_dev
real::variance, avg
integer::i
avg=sum(arr)/n
variance=0.
do i=1,n
  variance=variance+(arr(i)-avg))**2
end do
variance=variance/n
std_dev=sqrt(variance)
end subroutine

数组的大小在此处作为第一个参数 (n) 传递。
我们可以通过使用假定形状数组并将子例程放在显式接口（如模块或接口块）下来规避此问题。

subroutine find_std(arr,std_dev)
real,intent(in),dimension(:)::arr
real,intent(out)::std_dev
real::variance, avg
integer::n,i
n=size(arr)
avg=sum(arr)/n
variance=0.
do i=1,n
  variance=variance+(arr(i)-avg)**2
end do
variance=variance/n
std_dev=sqrt(variance)
end subroutine

A simple subroutine for finding the standard deviation of an array

subroutine find_std(n,arr,std_dev)
integer,intent(in)::n
real,intent(in),dimension(n)::arr
real,intent(out)::std_dev
real::variance, avg
integer::i
avg=sum(arr)/n
variance=0.
do i=1,n
  variance=variance+(arr(i)-avg))**2
end do
variance=variance/n
std_dev=sqrt(variance)
end subroutine

The size of the array is passed here as the first argument (n).
We can circumvent this by using an assumed-shape array and putting the subroutine under an explicit interface like module or an interface block.

subroutine find_std(arr,std_dev)
real,intent(in),dimension(:)::arr
real,intent(out)::std_dev
real::variance, avg
integer::n,i
n=size(arr)
avg=sum(arr)/n
variance=0.
do i=1,n
  variance=variance+(arr(i)-avg)**2
end do
variance=variance/n
std_dev=sqrt(variance)
end subroutine

回复收藏 0 原文

~没有更多了~