Dear all,
During my simulation, I wanted to get the snapshot of a particular timestep and therefore I used a very high frequency for dump command, ensuring that dumping occurs only at that instant and for last timestep.
This worked for the serial executable, but when I did this in parallel, it quits the execution signaling “Invalid dump frequency”. Could anyone of you help me let me know the maximum limit of dump frequency, if there is one such value.
Thank you
Jeeno Jose
Dear all,
During my simulation, I wanted to get the snapshot of a particular timestep
and therefore I used a very high frequency for dump command, ensuring that
dumping occurs only at that instant and for last timestep.This worked for the serial executable, but when I did this in parallel, it
quits the execution signaling "Invalid dump frequency". Could anyone of you
help me let me know the maximum limit of dump frequency, if there is one
such value.
this doesn't make sense. there is only one location in the source code
where this specific error message is printed, which is executed
regardless whether you run in serial or parallel and directly
evaluates the input. that message/error is only triggered when the
dump frequency is given as zero or negative, i.e. when it indeed is
invalid. so the cause has to be a different input or some other
difference in how you run your input.
axel.
Dear all,
True that it doesn’t make any sense, but it happened, I am unable to figure why it happened,
let me put the source code and log_restart (in serial(works fine) and parallel(where error is signaled)) of a similar code, but which produces same problem.
source code:
# VARIABLES
variable fname index data.susiminimlammps
variable simname index datarestart_susiminimlammps
# Initialization
units real
boundary p p p
atom_style molecular
log log.{simname}.txt*
*read_data {fname}
neighbor 3 bin
neigh_modify page 10000000 one 1000000 every 1 delay 0 check no
# dielectric 1.0
# potentials
pair_style lj/cut 20.045 # global cut off 5*sigma sigma for rminij 4.5
pair_coeff 1 1 0.0936 4.009 9.2207 # 2.3 sigma as suggested by doxasrakis
pair_coeff 1 2 0.1015 3.7925 8.72278 #
pair_coeff 1 3 0.1015 3.7925 8.72278 #
pair_coeff 1 4 0.0936 4.009 9.2207 #
pair_coeff 1 5 0.1456 4.009 9.2207 #
pair_coeff 1 6 0.1456 4.009 9.2207 #
pair_coeff 1 7 0.1456 4.009 9.2207 #
pair_coeff 2 2 0.1 3.3854 7.78642 #
pair_coeff 2 3 0.1 3.3854 7.78642 #
pair_coeff 2 4 0.1015 3.7925 8.72278 #
pair_coeff 2 5 0.1505 3.7925 8.72278 #
pair_coeff 2 6 0.1505 3.7925 8.72278 #
pair_coeff 2 7 0.1505 3.7925 8.72278 #
pair_coeff 3 3 0.1 3.3854 7.78642 #
pair_coeff 3 4 0.1015 3.7925 8.72278 #
pair_coeff 3 5 0.1505 3.7925 8.72278 #
pair_coeff 3 6 0.1505 3.7925 8.72278 #
pair_coeff 3 7 0.1505 3.7925 8.72278 #
pair_coeff 4 4 0.0936 4.009 9.2207 #
pair_coeff 4 5 0.1456 4.009 9.2207 #
pair_coeff 4 6 0.1456 4.009 9.2207 #
pair_coeff 4 7 0.1456 4.009 9.2207 #
pair_coeff 5 5 0.2264 4.009 9.2207 #
pair_coeff 5 6 0.2264 4.009 9.2207 #
pair_coeff 5 7 0.2264 4.009 9.2207 #
pair_coeff 6 6 0.2264 4.009 9.2207 #
pair_coeff 6 7 0.2264 4.009 9.2207 #
pair_coeff 7 7 0.2264 4.009 9.2207 #
bond_style harmonic # CVFF http://redmine.scorec.rpi.edu/anonsvn/lammps-cuda/tools/msi2lmp/biosym_frc_files/cvff.frc
bond_coeff 1 322.8 1.51
bond_coeff 2 655.2 1.34
bond_coeff 3 322.8 1.51
bond_coeff 4 322.8 1.51
bond_coeff 5 322.8 1.51
bond_coeff 6 322.7 1.53
bond_coeff 7 322.8 1.51
angle_style harmonic # CGenFF
angle_coeff 1 45 125.0
angle_coeff 2 45 125.0
angle_coeff 3 45 126.0
angle_coeff 4 45 125.0
angle_coeff 5 57 112.0
angle_coeff 6 57 112.0
angle_coeff 7 45 125.0
dihedral_style table spline 10000
dihedral_coeff 1 dihedrallistminimisationsusi.table TORSION13
dihedral_coeff 2 dihedrallistminimisationsusi.table TORSION2
dihedral_coeff 3 dihedrallistminimisationsusi.table TORSION13
improper_style harmonic # a high value
improper_coeff 1 500 0.0
improper_coeff 2 500 0.0
improper_coeff 3 500 0.0
improper_coeff 4 500 0.0
improper_coeff 5 500 0.0
special_bonds lj 0.0 0.0 0.05
# dump 1 all dcd 5 susiminimlammps.dcd
# dump_modify 1 unwrap yes
dump img0 all image 1000000000000000000000000000000 configu.0*.jpg mol type axes yes 1.25 0.01 box no 0.0 view 90 0 size 4096 4096 # center s 0.2 0.5 0.5
dump_modify img0 backcolor white
dump_modify img0 amap min max ca 0.0 10 min cyan 2 blueviolet 3 aquamarine 4 olive 5 tomato 6 burlywood 7 rosybrown 8 cadetblue 9 chartreuse max indigo
# Minimization
thermo 400
thermo_style custom step temp etotal pe ke evdwl eangle ebond edihed eimp lx ly lz density vol press # c_inter c_inter[1] c_inter[2] c_inter[3] c_poly[1] c_poly[2] c_poly[3] c_grap[1] c_grap[2] c_grap[3]
min_style cg
minimize 0.0 0.0 5000000 1000000
write_dump all custom dump.minimisedsusifinalcoordinates id type x y z ix iy iz modify sort id
write_data data.susiminimfinalstate pair ij
log_restart (serial)
read_data ${fname}
read_data data.susiminimlammps
orthogonal box = (0 0 0) to (30 30 30)
1 by 1 by 1 MPI processor grid
reading atoms …
1200 atoms
scanning bonds …
2 = max bonds/atom
scanning angles …
2 = max angles/atom
scanning dihedrals …
1 = max dihedrals/atom
scanning impropers …
1 = max impropers/atom
reading bonds …
1190 bonds
reading angles …
1180 angles
reading dihedrals …
690 dihedrals
reading impropers …
480 impropers
3 = max # of 1-2 neighbors
3 = max # of 1-3 neighbors
6 = max # of 1-4 neighbors
8 = max # of special neighbors
neighbor 3 bin
neigh_modify page 10000000 one 1000000 every 1 delay 0 check no
# dielectric 1.0
# potentials
pair_style lj/cut 20.045 # global cut off 5*sigma sigma for rminij 4.5
pair_coeff 1 1 0.0936 4.009 9.2207 # 2.3 sigma as suggested by doxasrakis
pair_coeff 1 2 0.1015 3.7925 8.72278 #
pair_coeff 1 3 0.1015 3.7925 8.72278 #
pair_coeff 1 4 0.0936 4.009 9.2207 #
pair_coeff 1 5 0.1456 4.009 9.2207 #
pair_coeff 1 6 0.1456 4.009 9.2207 #
pair_coeff 1 7 0.1456 4.009 9.2207 #
pair_coeff 2 2 0.1 3.3854 7.78642 #
pair_coeff 2 3 0.1 3.3854 7.78642 #
pair_coeff 2 4 0.1015 3.7925 8.72278 #
pair_coeff 2 5 0.1505 3.7925 8.72278 #
pair_coeff 2 6 0.1505 3.7925 8.72278 #
pair_coeff 2 7 0.1505 3.7925 8.72278 #
pair_coeff 3 3 0.1 3.3854 7.78642 #
pair_coeff 3 4 0.1015 3.7925 8.72278 #
pair_coeff 3 5 0.1505 3.7925 8.72278 #
pair_coeff 3 6 0.1505 3.7925 8.72278 #
pair_coeff 3 7 0.1505 3.7925 8.72278 #
pair_coeff 4 4 0.0936 4.009 9.2207 #
pair_coeff 4 5 0.1456 4.009 9.2207 #
pair_coeff 4 6 0.1456 4.009 9.2207 #
pair_coeff 4 7 0.1456 4.009 9.2207 #
pair_coeff 5 5 0.2264 4.009 9.2207 #
pair_coeff 5 6 0.2264 4.009 9.2207 #
pair_coeff 5 7 0.2264 4.009 9.2207 #
pair_coeff 6 6 0.2264 4.009 9.2207 #
pair_coeff 6 7 0.2264 4.009 9.2207 #
pair_coeff 7 7 0.2264 4.009 9.2207 #
bond_style harmonic # CVFF http://redmine.scorec.rpi.edu/anonsvn/lammps-cuda/tools/msi2lmp/biosym_frc_files/cvff.frc
bond_coeff 1 322.8 1.51
bond_coeff 2 655.2 1.34
bond_coeff 3 322.8 1.51
bond_coeff 4 322.8 1.51
bond_coeff 5 322.8 1.51
bond_coeff 6 322.7 1.53
bond_coeff 7 322.8 1.51
angle_style harmonic # CGenFF
angle_coeff 1 45 125.0
angle_coeff 2 45 125.0
angle_coeff 3 45 126.0
angle_coeff 4 45 125.0
angle_coeff 5 57 112.0
angle_coeff 6 57 112.0
angle_coeff 7 45 125.0
dihedral_style table spline 10000
dihedral_coeff 1 dihedrallistminimisationsusi.table TORSION13
dihedral_coeff 2 dihedrallistminimisationsusi.table TORSION2
dihedral_coeff 3 dihedrallistminimisationsusi.table TORSION13
improper_style harmonic # a high value
improper_coeff 1 500 0.0
improper_coeff 2 500 0.0
improper_coeff 3 500 0.0
improper_coeff 4 500 0.0
improper_coeff 5 500 0.0
special_bonds lj 0.0 0.0 0.05
3 = max # of 1-2 neighbors
3 = max # of 1-3 neighbors
6 = max # of 1-4 neighbors
8 = max # of special neighbors
# dump 1 all dcd 5 susiminimlammps.dcd
# dump_modify 1 unwrap yes
dump img0 all image 1000000000000000000000000000000 configu.0*.jpg mol type axes yes 1.25 0.01 box no 0.0 view 90 0 size 4096 4096 # center s 0.2 0.5 0.5
dump_modify img0 backcolor white
# give the mol id color, number of entries to be put, always max should be gold (graphene)
dump_modify img0 amap min max ca 0.0 10 min cyan 2 blueviolet 3 aquamarine 4 olive 5 tomato 6 burlywood 7 rosybrown 8 cadetblue 9 chartreuse max indigo
# aliceblue silver (dark alice blue, grey)
# aqua(blue) cyan (bright aqua) blue blueviolet plum (light violet less fluorescent bright) indigo violet(pale than blue violet) navy (dark blue)
# chartreuse(fluorescent gren) cadetblue(less bright green) aquamarine(green) limegreen olive(green and yellow but good)
# orangered (saffron) rosybrown(nearly brown and grey shade) burlywood(bet brown and gold) tomato (dark saffron)
# gold
# dump_modify img0 adiam 1 2
# dump_modify img0 adiam 2 1.0
# dump_modify img0 adiam 3 1.04
# dump_modify img0 adiam 4 2
# dump_modify img0 adiam 5*7 3
# Minimization
thermo 400
thermo_style custom step temp etotal pe ke evdwl eangle ebond edihed eimp lx ly lz density vol press # c_inter c_inter[1] c_inter[2] c_inter[3] c_poly[1] c_poly[2] c_poly[3] c_grap[1] c_grap[2] c_grap[3]
min_style cg
minimize 0.0 0.0 5000000 1000000
WARNING: Resetting reneighboring criteria during minimization (…/min.cpp:168)
Neighbor list info …
1 neighbor list requests
update every 1 steps, delay 0 steps, check yes
master list distance cutoff = 12.2207
Memory usage per processor = 8.29696 Mbytes
Step Temp TotEng PotEng KinEng E_vdwl E_angle E_bond E_dihed E_impro Lx Ly Lz Density Volume Press
0 0 16467.913 16467.913 0 16977.674 1.163775e-008 8.1013297e-008 -509.76073 1.1496271e-007 30 30 30 1.0066705 27000 206538.13
400 0 -1205.1656 -1205.1656 0 -993.17783 266.23306 9.644699 -493.04958 5.1840239 30 30 30 1.0066705 27000 11388.054
800 0 -1454.6536 -1454.6536 0 -1149.6196 223.46972 6.8122171 -538.26824 2.9522211 30 30 30 1.0066705 27000 8749.2918
1183 0 -1485.6243 -1485.6243 0 -1171.0098 218.33523 6.6568161 -542.38799 2.7814595 30 30 30 1.0066705 27000 8327.7541
Loop time of 7.23447 on 1 procs for 1183 steps with 1200 atoms
99.8% CPU use with 1 MPI tasks x 1 OpenMP threads
Minimization stats:
Stopping criterion = linesearch alpha is zero
Energy initial, next-to-last, final =
16467.9130724 -1485.62431204 -1485.62431204
Force two-norm initial, final = 7173.84 5.14589
Force max component initial, final = 1203.71 1.49403
Final line search alpha, max atom move = 3.77729e-010 5.64339e-010
Iterations, force evaluations = 1183 2417
MPI task timings breakdown:
Section | min time | avg time | max time |%varavg| %total
Dear all,
True that it doesn't make any sense, but it happened, I am unable to figure
why it happened,
well, look at the meaningless input that you are using:
using a dump frequency of 1000000000000000000000000000000 is
completely ridiculous. even on the fastest existing supercomputer and
assuming perfect scaling, you would not get a single image output even
if you waited for as long the age of the current universe.
this number cannot be correctly represented in a regular integer, and
now it depends on the libc implementation of your specific platform
and your compiler settings whether it will produce an positive or a
negative integer from wrapping around the capacity of an integer
multiple times.
if you don't want the command to have effect, why not simply comment it out?
axel.
You should avoid this " dump img0 all image 1000000000000000000000000000000 "…
If you want to dump only on specific steps, use “dump_modify every” together with a file containing listed timesteps. Like for instance:
Arthur
another choice might be to just use the write_dump http://lammps.sandia.gov/doc/write_dump.html
Thank you very much for suggestions.
Pablo: its about dumping images.
Thank you very much for suggestions.
Pablo: its about dumping images.
so what?
internally, the write_dump command is just a minimal wrapper around
the dump/undump/dump_modify commands and thus works for images just as
well as for any other dump file format.
axel.