Fortran 90：变量的值似乎是随机变化的。内存分配？

我试图在一个较大的代码中找到一个bug，并将问题隔离到下面的代码片段中，我将其放入一个单独的程序中进行故障排除。究竟为什么矩阵“Ga”的一个条目（而且只有一个）会通过用零初始化HEB而改变？如果删除此初始化，两个“Ga”输出将相同。请向下滚动至底部，查看问题显示的位置。非常感谢你，这让我快发疯了。我使用ifort进行编译

program AC
 implicit none
  integer, parameter :: dp = selected_real_kind(15, 307)
  integer :: n, ndim, k, j, i, l, m, o, steps
  real(dp) :: emax, omega, pi, EFermi, auev
  complex(dp) :: Grs,Gas, ACCond, tinyc, cunit, czero, cone

  complex(dp), allocatable :: GammaL(:,:)     
  complex(dp), allocatable :: GammaL_EB(:,:)  
  complex(dp), allocatable :: Gr(:,:)     
  complex(dp), allocatable :: Ga(:,:)  
  complex(dp), allocatable :: GrInv(:,:)    
  complex(dp), allocatable :: GaInv(:,:)  
  complex(dp), allocatable :: GrDiag(:,:)     
  complex(dp), allocatable :: GaDiag(:,:)  
  complex(dp), allocatable :: GrInvDiag(:,:)   
  complex(dp), allocatable :: GaInvDiag(:,:)  
  complex(dp), allocatable :: GammaR(:,:)    
  complex(dp), allocatable :: GammaR_EB(:,:)     
  complex(dp), allocatable :: R(:,:) 
  complex(dp), allocatable :: Yc(:,:)         
  complex(dp), allocatable :: Yd(:,:)        
  complex(dp), allocatable :: AnaInt(:,:)     
  complex(dp), allocatable :: H(:,:)         
  complex(dp), allocatable :: HEB(:,:)          
  complex(dp), allocatable :: HamEff(:,:)    
  complex(dp), allocatable :: EigVec(:,:)     
  complex(dp), allocatable :: InvEigVec(:,:)  
  complex(dp), allocatable :: EigVal(:)       
  complex(dp), allocatable :: ctemp(:,:)     
  complex(dp), allocatable :: S(:,:)          
  complex(dp), allocatable :: SelfL(:,:)     
  complex(dp), allocatable :: SelfR(:,:)      
  complex(dp), allocatable :: SHalf(:,:)     
  complex(dp), allocatable :: InvSHalf(:,:)   
  complex(dp), allocatable :: Abba(:,:)   
  complex(dp), allocatable :: Integrand(:,:)
  complex(dp), allocatable :: HDiag(:,:)
  complex(dp), allocatable :: unity(:,:)

!Lapack arrays and variables
  integer :: info, lwork
  complex(dp), allocatable :: work(:)       
  real(dp), allocatable :: rwork(:)    
  integer,allocatable :: ipiv(:)

!########################################################################

!Constants
    pi = 3.14159265359
    cunit = (0,1)
    czero = (0,0)
    cone = (1,0)
    tinyc = (0.0, 0.000000000001)


!System and calculation parameters
    ndim = 3 !Dimension of the Hamiltonian
    lwork = ndim*ndim
    EFermi = 0.0 
    emax = 5.0 !Energy in eV to which the admittance is to be calculated
    steps = 1000 !Number of energy steps for the admittance calculation


    allocate(Integrand(ndim,ndim))
    allocate(H(ndim,ndim))
    allocate(Yc(ndim,ndim))
    allocate(Yd(ndim,ndim))
    allocate(S(ndim,ndim))
    allocate(SelfL(ndim,ndim))
    allocate(SelfR(ndim,ndim))
    allocate(HamEff(ndim,ndim))
    allocate(GammaR(ndim,ndim))
    allocate(GammaL(ndim,ndim))
    allocate(AnaInt(ndim,ndim))
    allocate(EigVec(ndim,ndim))
    allocate(EigVal(ndim))
    allocate(InvEigVec(ndim,ndim))
    allocate(R(ndim,ndim))
    allocate(GammaL_EB(ndim,ndim))
    allocate(GammaR_EB(ndim,ndim))
    allocate(HEB(ndim,ndim))
    allocate(Ga(ndim,ndim))
    allocate(Gr(ndim,ndim))
    allocate(GaInv(ndim,ndim))
    allocate(GrInv(ndim,ndim))
    allocate(GaDiag(ndim,ndim))
    allocate(GrDiag(ndim,ndim))
    allocate(GaInvDiag(ndim,ndim))
    allocate(GrInvDiag(ndim,ndim))
    allocate(Abba(ndim,ndim))
    allocate(HDiag(ndim,ndim))
    allocate(unity(ndim,ndim))


!################################################


!Definition of the trial system

    H = reshape((/ -0.3, -0.8, -0.2, -0.8, -0.5, -0.14, -0.2, -0.14, -0.24 /), shape(H))
    SelfL = reshape((/ -0.1*cunit, -0.3*cunit, -0.0*cunit, -0.3*cunit, -0.0*cunit, -0.0*cunit, -0.0*cunit, -0.0*cunit, -0.0*cunit /), shape(SelfL))
    SelfR = reshape((/ -0.0*cunit, -0.0*cunit, -0.0*cunit, -0.0*cunit, -0.0*cunit, -0.6*cunit, -0.0*cunit, -0.6*cunit, -0.3*cunit /), shape(SelfR))

    HamEff(:,:)=(H(:,:) + SelfL(:,:) + SelfR(:,:))   
    GammaL(:,:)=(-2.0 * cmplx(aimag(SelfL(:,:)),0.0_dp,dp)) 
    GammaR(:,:)=(-2.0 * cmplx(aimag(SelfR(:,:)),0.0_dp,dp))

unity(:,:) = 0.0_dp
do i=1,ndim
unity(i,i) = cone
end do



!Calculate Eigenvalues and Eigenvectors of the Hamiltonian (Checked! Yields correct Eigenvalues for trial matrix).
!In the PDF: B = EigVec, B^(-1) = InvEigVec, Hk = EigVal

    allocate(ctemp(ndim,ndim))
    allocate(work(lwork),rwork(2*ndim))
    ctemp(:,:) = HamEff(:,:)
    call zgeev('N', 'V', ndim, ctemp, ndim, EigVal, InvEigVec, ndim, EigVec, ndim, work, lwork, rwork, info)
    deallocate(work,rwork)
    deallocate(ctemp) 

    InvEigVec(:,:)=EigVec(:,:)
    lwork = 3*ndim
    allocate(ipiv(ndim))
    allocate(work(lwork))
    call zgetrf(ndim,ndim,InvEigVec,ndim,ipiv,info)
    call zgetri(ndim,InvEigVec,ndim,ipiv,work,lwork,info)
    deallocate(work)
    deallocate(ipiv)


!Green Function at E=2
    GrInv(i,j) = 0.0_dp
    GaInv(i,j) = 0.0_dp
    Gr(i,j) = 0.0_dp
    Ga(i,j) = 0.0_dp

    do i=1,ndim
       do j=1,ndim
          GrInv(i,j) = (3.0_dp + tinyc)*unity(i,j) - HamEff(i,j)
       end do
    end do

    do i=1,ndim
       do j=1,ndim
          GaInv(i,j) = (3.0_dp - tinyc)*unity(i,j) - conjg(HamEff(j,i))
       end do
    end do


!Matrix Inversions

    Ga(:,:)=GaInv(:,:)
    lwork = 3*ndim
    allocate(ipiv(ndim))
    allocate(work(lwork))
    call zgetrf(ndim,ndim,Ga,ndim,ipiv,info)
    call zgetri(ndim,Ga,ndim,ipiv,work,lwork,info)
    deallocate(work)
    deallocate(ipiv)

    Gr(:,:)=GrInv(:,:)
    lwork = 3*ndim
    allocate(ipiv(ndim))
    allocate(work(lwork))
    call zgetrf(ndim,ndim,Gr,ndim,ipiv,info)
    call zgetri(ndim,Gr,ndim,ipiv,work,lwork,info)
    deallocate(work)
    deallocate(ipiv)




    GrInvDiag(i,j) = 0.0_dp
    GaInvDiag(i,j) = 0.0_dp
    GrDiag(i,j) = 0.0_dp
    GaDiag(i,j) = 0.0_dp
    do i=1,ndim
       do j=1,ndim
          GrInvDiag(i,j) = (3.0_dp + tinyc)*unity(i,j) - HDiag(i,j)
       end do
    end do

    do i=1,ndim
       do j=1,ndim
          GaInvDiag(i,j) = (3.0_dp - tinyc)*unity(i,j) - conjg(HDiag(j,i))
       end do
    end do



!Matrix Inversions

    GaDiag(:,:)=GaInvDiag(:,:)
    lwork = 3*ndim
    allocate(ipiv(ndim))
    allocate(work(lwork))
    call zgetrf(ndim,ndim,GaDiag,ndim,ipiv,info)
    call zgetri(ndim,GaDiag,ndim,ipiv,work,lwork,info)
    deallocate(work)
    deallocate(ipiv)

    GrDiag(:,:)=GrInvDiag(:,:)
    lwork = 3*ndim
    allocate(ipiv(ndim))
    allocate(work(lwork))
    call zgetrf(ndim,ndim,GrDiag,ndim,ipiv,info)
    call zgetri(ndim,GrDiag,ndim,ipiv,work,lwork,info)
    deallocate(work)
    deallocate(ipiv)



!The problem occurs here. Why does an entry of Ga get deleted by initialization of a new array HEB?

    write(*,*) Ga
    write(*,*) "--------------------"


     HEB(i,j) = 0.0_dp


    write(*,*) Ga




!#############################################

    deallocate(Integrand)
    deallocate(HEB)
    deallocate(Yc)
    deallocate(Yd)
    deallocate(HamEff)
    deallocate(GammaR)
    deallocate(GammaL)
    deallocate(Ga)
    deallocate(Gr)
    deallocate(GaInv)
    deallocate(GrInv)
    deallocate(GaDiag)
    deallocate(GrDiag)
    deallocate(GaInvDiag)
    deallocate(GrInvDiag)
    deallocate(AnaInt)
    deallocate(EigVec)
    deallocate(EigVal)
    deallocate(InvEigVec)
    deallocate(H)
    deallocate(S)
    deallocate(SelfL)
    deallocate(SelfR)
    deallocate(R)
    deallocate(GammaL_EB)
    deallocate(GammaR_EB)
    deallocate(Abba)
    deallocate(HDiag)
    deallocate(unity)
end program AC

---

正如ShinTakezou指出的，在

DO

循环之外使用

i

和

j

似乎有问题。除非重新初始化，否则用于

DO

循环的变量比其上一个值大1（即，如果要将其打印到屏幕上，您将发现

i=j=4

）。显然，内存地址

HEB（4,4）

与

GA（3,1）

的内存地址对齐

---

解决这个问题的方法是使用

HEB=0.\u dp

或

HEB（：，：）=0.\u dp

在print语句前后都会得到相同的答案。

当你在任何“do”之外写GrInv（i，j）=0.0dp等时，i和j的值应该是什么？也就是说，似乎有一些var的用法不包含正确的值，特别是你在之前的do循环中使用I来初始化unity-可能你剥离了太多，注入了明显的错误？是的，这是我一定忽略了的。我将所有出现的问题更正为“GaDiag（：，：）=0.0_dp”类型的初始化，问题仍然存在。编辑：没关系！我忘了HEB（I，j）本身！谢谢，已经够了。是的，谢谢。是否有任何编译器或编译器设置显示这样的错误？我花了好几天的时间寻找这个，因为我认为这个错误可能是在数学/物理方面code@user464293当然使用ifort，您可以将

-check

添加到编译选项中。它仍然会编译成功，但会给您一个

运行时检查失败

错误。如果编译器标志中包含

-g-traceback

，它将显示错误的行号（在您的情况下，第一个出现在

GrInv（i，j）=0.0dp

）。