Hi all,
I have been doing some simulations involving dense periodic salt crystals
and have found that I am overfilling the f14.4 format for the energy
report. Because the system is so dense I believe that the energy should be
very large and that it is appropriate to simply change the output format in
the mdout file. After poking around in the sander source I think that the
output happens in dynlib.F90. I have tried modifying
$AMBERHOME/AmberTools/src/sander/dynlib.F90 in two different ways and I am
not seeing any change in either the mdout file or the mdinfo file.
I first tried editing around line 400:
write(6,9018) nstep,time,temp,press
write(6,9028) etot,ektot,epot
write(6,*) "Etot",etot,"EKtot",ektot,"EPtot",epot !added by John
write(6,9038) ebond,eangle,edihed
write(6,*) "Ebond",ebond,"Eangle",eangle,"Edihed",edihed !added by John
and then tried:
write(6,9018) nstep,time,temp,press
write(6,9028) etot,ektot,epot
write(6,29) etot,ektot,epot !added by John
29 format (1x,'Etot = ',f24.4,2x,'EKtot = ',f24.4,2x, &
'EPtot = ',f24.4)
write(6,9038) ebond,eangle,edihed
write(6,39) ebond,eangle,edihed !added by John
39 format (1x,'BOND = ',f24.4,2x,'ANGLE = ',f24.4,2x, &
'DIHED = ',f24.4)
After each of those changes I did a make clean and a make install in the
sander directory. I reran my calculations after each of those changes and
did not see any change in the mdout or mdinfo files. I was wondering what's
going on. Is dynlib.F90 not the correct file to change in order to reformat
the output? I am running Amber14 (Amber14+AmberTools14) on a CentOS 5.6
host with the Intel compiler suite Version 12.0.0.084
Thanks,
John Dood
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Jun 30 2015 - 10:30:02 PDT
