Problem with array bounds

[Jorge_Alonso]

Dear all,

I used to compile and execute my code under earlier versions of ifort 11.x in ubuntu and it worked fine. I also tested the same code under compaq visual fortran and it worked. Right now I am using a Mac OS Lion and last version of ifort.

When I run my code I get the following error

"forrtl: severe (408): fort: (2): Subscript #1 of the array DK has value 40001 which is greater than the upper bound of 40000

Image PC Routine Line Source

Stack trace terminated abnormally."

After some experimenting I isolated the subroutine where my program collapse. I define in a module the parameter "np=40000". In the following subroutine this dimension is indexed by "it." I do not understand why the array index goes to 40001

Next I paste the shell compilation code and the subroutine is causing trouble... I appreciate in advance any suggestion

P.D. I include all my code in a zip file

****************************SHELL SCRIPT****************************************

## SHELL PROGRAM TO COMPILE AND EXECUTE THE BENCHMARK MODEL OF "SOCIAL SECURITY

## AND RETIREMENT ACROSS THE OECD"

ifort -debug all -warn all -noaltparam -fast -xHost -unroll -assume protect_constants -assume noaccuracy_sensitive -check bounds -fltconsistency -traceback -c nrtype.f90 nrutil.f90 \\

modvar.f90 norm_cdf.f90 norm_inv.f90 tauchen.f90 q1da.f90 q1dax.f90 idamax.f gl15t.f90 uni.f90 zbrent.f90 aime.f90 \\

golden_srch.f90 oecd_simulation.f90 pia.f90 sdist3.f90 slcprofiles.f90 soptlc2b.f90 soptvalg2.f90 sprice1c2.f90 \\

stauchen_adda_cooper.f90 sub_cumsum.f90 sub_grid_generation.f90 sub_linint2d.f90 sval_end2.f90 swelfare.f90

ifort -debug all -warn all -noaltparam -fast -xHost -unroll -assume protect_constants -assume noaccuracy_sensitive -check bounds -fltconsistency -traceback -o myprog nrtype.o nrutil.o\\

modvar.o norm_cdf.o norm_inv.o tauchen.o q1da.o q1dax.o idamax.o gl15t.o uni.o zbrent.o aime.o golden_srch.o \\

oecd_simulation.o pia.o sdist3.o slcprofiles.o soptlc2b.o soptvalg2.o sprice1c2.o stauchen_adda_cooper.o \\

sub_cumsum.o sub_grid_generation.o sub_linint2d.o sval_end2.o swelfare.o

rm *.o

./myprog

**************************SUBROUTINE****************************************

SUBROUTINE sdist3(dv,dc,dh,dk,dz,de,dai,dpn,vfcn,cfcn,hfcn,kpfcn,w,me,b)

USE modvar

IMPLICIT NONE

REAL(DP), INTENT(OUT), DIMENSION(np,na) :: dv,dc,dh,dk,de,dai,dpn

INTEGER, INTENT(OUT), DIMENSION(np,na) :: dz

REAL(DP), INTENT(IN), DIMENSION(na,nk,ne,nz) :: vfcn,cfcn,hfcn,kpfcn

REAL(DP), INTENT(IN) :: w,me,b

REAL(DP) :: rd(np,na),smss(nz),aime,pia,hint

INTEGER :: it,ia,iz

CALL RANDOM_SEED(PUT=(/13567879/))

CALL RANDOM_NUMBER(rd)

dk(:,1)=b

dai(:,1)=0.0

dpn(:,1)=0.0

DO it=1,np

CALL sub_cumsum(smss,mss,nz)

DO iz=1,nz

IF (rd(it,1) .le. smss(iz)) THEN

dz(it,1)=iz

EXIT

END IF

END DO

dh(it,1)=hfcn(1,1,1,dz(it,1))

dv(it,1)=vfcn(1,1,1,dz(it,1))

dc(it,1)=cfcn(1,1,1,dz(it,1))

de(it,1)=ga(1)*gz(dz(it,1))*dh(it,1)

END DO

DO ia=2,na

DO it=1,np

CALL sub_cumsum(smss,m(dz(it,ia-1),:),nz)

DO iz=1,nz

IF (rd(it,ia) .le. smss(iz)) THEN

dz(it,ia)=iz

EXIT

END IF

END DO

CALL sub_linint2d(nk,ne,dk(it,ia),gk,ge,kpfcn(ia-1,:,:,dz(it,ia-1)),dk(it,ia-1),dai(it,ia-1))

! dk(it,ia)=dmax1(dk(it,ia),gkmax)

dai(it,ia)=aime(dai(it,ia-1),w*de(it,ia-1),ia-1)

CALL sub_linint2d(nk,ne,hint,gk,ge,hfcn(ia,:,:,dz(it,ia)),dk(it,ia),dai(it,ia))

IF (hint .ge. ft/2.0) THEN

dh(it,ia)=ft

ELSE

dh(it,ia)=nw

END IF

CALL sub_linint2d(nk,ne,dv(it,ia),gk,ge,vfcn(ia,:,:,dz(it,ia)),dk(it,ia),dai(it,ia))

CALL sub_linint2d(nk,ne,dc(it,ia),gk,ge,cfcn(ia,:,:,dz(it,ia)),dk(it,ia),dai(it,ia))

de(it,ia)=ga(ia)*gz(dz(it,ia))*dh(it,ia)

dpn(it,ia)=dmax1(0.0,generosity*pia(dai(it,ia),ia,me)*(1.0-itax*(1.0-(ft-dh(it,ia))/ft)))

END DO

! print '(I2,3F10.5)', ia,dk(it,ia),dai(it,ia),dz(it,ia)

END DO

END SUBROUTINE sdist3

[mecej4]

You did not provide the input files for units 16, 17 and 18, without which the program did not run far enough to display the subscript error that you reported.