求助大家：

同样的一段代码，一个写入module的function，并在主程序中调用，一个直接放在主程序中，出来的结果却不一样

两者的差别主要是，后者为0的地方，前者为NaN或者极小、极大的值

怎样才能通过调用function得到正确的值呢？

谢谢...

提供代码：主要是fppb这个函数...

module trustregion
Implicit none

integer, parameter :: n = 300
integer, parameter :: p = 7, pn = 22, df = 3 
real, parameter :: lambda = 2.0 

integer :: id_begin(p), id_end(p), id(df)
data id_begin /2,5,8,11,14,17,20/
data id_end /4,7,10,13,16,19,22/
integer :: i
real :: x(n,pn), y(n)
real :: beta(pn) = (/(0.0,i = 1,pn)/)
real :: temp_beta(pn)

integer :: ite1 = 1, ite2 = 1
real :: tol = 1D-5
real :: delta0 = 0.5d0
integer :: itemax = 200

real :: fd(df), sd(df,df)

complex :: eigen_vector(df,df), eigen_value(df)

contains

!--------------------------------------------------------
!---------------- read data ----------------------------
subroutine loaddata()
  integer :: i
  open(100,file = "train_splice.txt")
  do i = 1, n
    read(100,*) y(i),x(i,:)
  end do
end subroutine loaddata

!--------------------------------------------------------
!---------------- probability ----------------------------
function fprob(bt)
  real :: bt(pn)
  real :: fprob(n)
  integer :: i
  do i = 1,n
    fprob(i) = exp(sum(x(i,:) * bt))/ (1 + exp(sum(x(i,:) * bt)))
  end do
end function fprob

!--------------------------------------------------------
!----------------  beta  first derivative----------------
function fpb(bt)
  real :: bt(pn)
  real :: fpb(pn)
  integer :: j
  do j = 1,pn
    fpb(j) = - sum(x(:,j) * (y - fprob(bt)))
  end do
end function fpb

!--------------------------------------------------------
!----------------  beta  second derivative---------------
function fppb(bt)
  real :: bt(pn), pr(n)
  real :: fppb(pn,pn),tppb(pn,pn)
  integer :: i,j,k
  pr = fprob(bt)


  do i = 1, n
      do j = 1, pn
        do k = 1, pn
        tppb(j,k) = x(i,j) * x(i,k) * pr(i) * (1-pr(i))
        end do
       end do
     fppb = fppb + tppb
   end do
end function fppb

end module
!-----------------------------------------------
! main program
!-----------------------------------------------
program main
use trustregion
implicit none


real :: ppb1(pn,pn),ppb2(pn,pn)


  real :: bt(pn), pr(n)
  real :: tppb(pn,pn)
  integer :: j,k
  
  call loaddata()
  ppb1 = fppb(beta)
  write(*,*) "ppb1",ppb1


  pr = fprob(bt)


  do i = 1, n
      do j = 1, pn
        do k = 1, pn
        tppb(j,k) = x(i,j) * x(i,k) * pr(i) * (1-pr(i))
        end do
       end do
     ppb2 = ppb2 + tppb
   end do
   write(*,*) "ppb2",ppb2


end program main