Date: Wed, 22 May 2002 08:15:06 -0700
From: David Case
Subject: Re: again about sander_classic Amber 6.0

On Wed, May 22, 2002, Nikolai Smolin wrote:

> I set
> ntwv=100, ntwvm=0,
> ntwx=100, ntwxm=0,
> ntwe=100, ntwem=0,
> And after simulation I have *.mdcrd and *.mden with all information
> about trajectory and energy, but emty *.mdvel.
>

Make the following change to runmd.f:

*** runmd.f Tue Aug 3 11:13:18 1999
--- runmd.f.new Wed May 22 08:11:34 2002
***************
*** 1114,1120 ****
c For simplicity for now, an allgatherv is done...
c
if ( (plevel .eq. 1 .and. NTP .gt. 0) .or.
! . (NTWV.GT.0 .AND. NTWV.LE.NTWVM) .or.
. (NSTEP+1 .ge. NSTLIM) .or.
. (mod(nstep+1,nscm).eq.0) ) then
# ifdef CSPP
--- 1114,1120 ----
c For simplicity for now, an allgatherv is done...
c
if ( (plevel .eq. 1 .and. NTP .gt. 0) .or.
! . (NTWV.GT.0 .AND. NTWV.GE.NTWVM) .or.
. (NSTEP+1 .ge. NSTLIM) .or.
. (mod(nstep+1,nscm).eq.0) ) then
# ifdef CSPP
***************
*** 1359,1365 ****
C
c IF (NTWV.GT.0 .AND. NTWV.LE.NTWVM) THEN changed 9/95 -- dap
IF (NTWV.GT.0 .and. nstep.ge.ntwvm) THEN
! IF (NTWV.LE.NTWVM .AND. MOD(NSTEP,NTWV).EQ.0)
+ CALL CORPAC(V,NRX0,NRX,13,LOUTFM)
ENDIF
C
--- 1359,1365 ----
C
c IF (NTWV.GT.0 .AND. NTWV.LE.NTWVM) THEN changed 9/95 -- dap
IF (NTWV.GT.0 .and. nstep.ge.ntwvm) THEN
! IF (MOD(NSTEP,NTWV).EQ.0)
+ CALL CORPAC(V,NRX0,NRX,13,LOUTFM)
ENDIF
C



(In the seven years since dap made this change, this is the first time this
bug has been reported. I think that shows how often people save
velocities...I know I never have.)

...thanks for the report....dac